THE EFFECTS OF MOTIVATION AND ANXIETY ON COLLEGE STUDENTS' USE OF INSTRUCTOR COMMENTS

by

Eric M. Wiltse

A dissertation submitted to the Department of Adult Learning and Technology and The Graduate School of The University of Wyoming in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

In

EDUCATION

Laramie, Wyoming

December, 2000

A primary objective of journalism and communication curriculums involves improvement of students' writing skills. To that end, instructors provide written feedback on students' first drafts to point out strengths and weaknesses of their stories, press releases, term papers and other work. The comments not only serve as an evaluation of student writing, but as an opportunity for the instructor to fill the role of an audience -- a reader of a product intended for a mass audience. The practice of instructor acting as a reader responding to student writing is increasingly used in English composition (Fey, 1993), but writing for an audience is one factor that differentiates journalistic writing from other types (Riffe & Stacks, 1988, 1992). Even in journalism classes, however, the instructor often is the only audience for a student's stories and becomes, in effect, a reader responding to a text. Rather than receiving "real world" phone calls and letters to the editor from pleased and (more often than not) disgruntled readers of their stories, journalism students receive only comments from their instructor.

Instructor feedback on first drafts can serve as an effective instructional method to help students improve their writing, not just in journalism but all types of writing courses. In their work on cognitive learning theory, Glover, Ronning and Bruning (1990) stated, "One direct and simple way to improve students� writing is to have teachers ... give feedback on their efforts" (p. 274). One of the tenets of the 1990s' writing-across-the-curriculum movement is that writing is improved by the opportunity to get feedback on a first draft before turning in a paper for a grade (McKeachie, 1999). Revising a first draft can produce other benefits for students, noted Mohr (1984), including building confidence and a sense of self-worth. Even with the advent of computer writing laboratories, spell- and grammar-checking software, and electronic communication networks, instructor feedback still plays an essential role in journalism education. Journalism Professor R. Thomas Berner (1992), who requires students to file stories electronically and then evaluates them on his computer, noted, "I make more comments than I would with pencil" (p. xiii).

The act of writing a news story is a complex cognitive process in which the journalist uses an integrated strategy in which he or she plans, writes and edits several sentences at a time (Pitts, 1989). Instructor feedback also affects the cognitive process of writing (Flower & Hayes, 1981). For instance, feedback from the instructor can help define or clarify the writing task or problem so the student can resolve it. Feedback may also provide cues that allow the writer to retrieve useful knowledge from long-term memory, such as grammatical knowledge. Feedback can cause the writer�s schemata to grow and change as new information is either assimilated or accommodated.

Some argue that writing cannot be viewed as purely a cognitive act because it takes place in a social, cultural and political context (Flower, 1989). A student may receive written comments from a teacher that conflict with feedback from a classmate. In revising the story, the student must weigh the politics of accepting or rejecting the instructor comments, the social pressure of a peer�s opinions, and the overall context of acting in an academic culture with grades, degrees and careers at stake.

Furthermore, the process of an instructor writing feedback and a student reading the instructor's comments can lead to dialogue in which the teacher and student negotiate the meaning of the written text (Flower, 1994). When instructors provide individual feedback to student writers, a new learning environment can evolve in which students and teachers become co-learners, sharing knowledge, and holding a dialogue (Kinkead, 1988). In this view, the student plays an active role in deciding the effect feedback will play upon the final written product.

Besides being a cognitive, social, political and cultural agent, instructor feedback also can have a powerful effect on students' emotions, especially writing apprehension, motivation, and their beliefs about writing and their own skills as writers (McLeod, 1987). Feedback generally can encourage or motivate students to greater effort, and can also serve as a reward for effort (Briggs & Hamilton, 1964).

Overall, examining how students perceive the usefulness of instructor feedback may improve writing instruction in general and the quality of teacher comments in particular. Besides considering the cognitive benefits of feedback, instructors also should be conscious of the social, political, cultural and emotional impacts their comments may have upon students. Student writers should consider and understand the role they play in negotiating meaning of their written texts, rather than serving as mere reactors to the instructor's feedback.

Writing researchers frequently note the amounts of time instructors spend providing feedback that students subsequently ignore or gloss over. "Faculty members have concerns about the time required" to provide written feedback on student assignments, McKeachie (1999, p. 137) noted. Instructors become further frustrated when students ignore their comments and suggestions, even going so far as to simply delete problematic passages rather than deal with them (Ziv, 1982). In their defense, students can become "overloaded" with feedback, causing them to ignore some teacher comments (McKeachie, 1999, p. 138). There is little empirical evidence that students understand or use teachers' responses to their writing (Knoblauch & Brannon, 1981). Researchers also do not know whether teacher comments actually help students become better writers (Sommers, 1982). While it is not clear whether feedback can improve student performance, teachers continue to write comments on student texts, indicating there is still a need to examine feedback from both the instructors' and students' perspectives.

Feedback provides students with an idea of what their instructors expect from them. This can lead to students considering the instructor to be their only audience and aiming their writing toward what they perceive the teacher wants rather than what the audience needs (Dohrer, 1991; Pitts, 1989; Thompson, 1994). This passive role of students merely reacting to teacher comments can also result in the writers merely correcting mechanical errors, such as spelling, rather than addressing larger problems with content or organization (Dohrer, 1991; Ferris, 1995; Pitts, 1989; Thompson, 1993, 1994). Straub (1997) categorized these types of feedback issues as global or local. Global matters focus upon a composition�s content, such as the ideas, development and organization, while local issues include mechanical problems, such as wording, sentence structure and correctness.

The literature also suggests that feedback can confuse students. Sommers (1982) noted that many students will not revise anything in their writing that has not been suggested by the teacher, even if they think they should. She suggested that students are afraid to take risks or may be confused by teacher comments. Others (Ziv, 1982) contended that students may simply misinterpret or not understand what an instructor intended to say in written feedback

Most of these studies have examined student reaction to teacher comments without considering individual writing characteristics. Writing apprehension, for example, may cause further problems for students trying to deal with instructor comments. Writer�s apprehension may be so severe for some students that they ignore an instructor�s feedback because of a history of failure on writing assignments (Daly, 1978). Instructor comments may not only cause apprehension in students, but may paralyze their efforts to improve their writing in the future.

On the other hand, the literature generally agrees that instructor feedback can inspire and motivate students to work harder on improving their writing. Sommers (1982) noted that without teacher comments, students would have no incentive to revise their writing. Lackey, Miller and Flanigan (1997) stated that feedback can motivate students to improve writing performance, but that much written feedback lowers motivation instead. Wlodkowski (1985) said that research shows there is a high correlation between motivation and learning. If a planned approach to motivating students is used, the instructor using that approach will be more successful than will be educators not using it, he added. Therefore, it seems important that writing instructors include motivational strategies in their curriculum.

How students react to feedback is a particular concern in journalism classes because instructors tend to note every error, both mechanical and global. As a product for a mass audience, news stories not only must be accurate and thorough in the reporting, but also must have proper spelling, grammar, and other mechanics. "Poorly written and badly organized stories deter readers from gaining information that may be essential to their well-being" (Fox, 1993, p. vii). As a result, many journalism students may be overwhelmed by instructor feedback. In particular, a weak writer who needs the most help, but whose papers receive the most teacher comments, may view his or her graded paper as a "messy autopsy" with the instructor as "coroner" (Grant-Davie & Shapiro, 1987, p. 6).

Considering the close scrutiny that journalism instructors give to their students' stories, it is surprising that so little attention has been paid to the relationships between feedback and students' writing characteristics.

The underlying theoretical base for this study includes the related motivation theories of Bandura and Vroom, cognitive feedback theory, social-cognitive learning theory, and writing apprehension theory.

Bandura�s social cognitive theory (1986) argued that motivation is largely influenced by a person�s self-efficacy beliefs in their ability to successfully perform certain skills and in their ability to successfully complete specific tasks. Furthermore, he emphasized that providing feedback in the form of modeling can increase self-efficacy. Feedback can also serve as a reward or outcome that can motivate students to work harder on improving their writing. Vroom�s expectancy-value theory of motivation (1970) argued that as the amount of the expected outcomes increases, a person�s motivation becomes stronger. Outcomes expectations involve how people rate the importance of a skill, such as writing, for achieving a goal or outcome. Furthermore, setting and meeting goals is another type of outcome that can motivate. Writing is a goal-directed process, according to Flower and Hayes (1981). Writers create their own writing goals, and the ability to set goals can differentiate good and bad writers.

There is also an interaction between student expectations of their performance ability (what Bandura would call their self-efficacy beliefs) and feedback, according to Kulhavy (1977). For instance, students who are confident that their answer on a test is right, but the answer is wrong, tend to pay more attention to instructor feedback. Many times, instructor's comments can create cognitive dissonance in students' minds between their perceptions of what they've written and the instructor's perceptions (Ziv, 1982). On the other hand, instructor feedback may cause students to feel anxious about their writing, noted Daly (1977), who identified the writing apprehension construct.

In their work on cognitive process theory, Flower and Hayes (1981) stated instructor feedback can impact both long-term memory and writing processes. Finally, Flower (1989, 1994) suggested that feedback can turn the writing process into a social, cultural and political act as the forces of instructor, peers and individual writers compete, negotiate and collaborate.

The purpose of this study was to explore students' writing characteristics that affect how they process teacher comments on first drafts of their written assignments. The driving force behind the study stemmed from an interest in how to provide more effective, relevant feedback to students. How do students use the instructor feedback? Do students focus on correcting local issues pointed out by instructors, or do they spend more time addressing global concerns, or are they equally interested in both types when revising a first draft? Do students with different types of writing characteristics process instructor feedback differently? What are the writing characteristics that influence whether students will use local or global feedback in their revisions? This research investigated the claims of previous studies that indicated students tend to use certain types of feedback and ignore others when revising rough drafts. But this investigation considered feedback usefulness in light of individuals' writing characteristics.

Students in mass communication courses completed a questionnaire and several instruments to measure the individual characteristics of writing apprehension and motivation. They also completed a questionnaire evaluating their use of global or local feedback when revising first drafts of news stories.

The following research questions were considered in the study:

1. What are the relationships among students' writing apprehension, writing task self-efficacy, writing skills self-efficacy, and writing outcomes expectations?
2. What are the effects of writing apprehension, writing outcomes expectations, and writing self-efficacy beliefs on students' use of global, local or both kinds of feedback from instructors on first drafts?

Daly (1977) found that people with high levels of writing apprehension tend to be poor writers compared to people with moderate or low levels. Furthermore, according to Daly, poor writers often have a history of receiving negative feedback on their writings. Flower and Hayes (1981) noted that poor writers may concentrate more on spelling or other mechanical tasks than on the content of their compositions. In a case study of a high-apprehensive writer, Selfe (1985) also observed that she seemed more concerned with avoiding mechanical errors than with the thematic soundness of her composition. Therefore, it seems likely that apprehensive writers will concentrate on instructor comments dealing with mechanical errors and ignore other feedback.

On the other hand, McCarthy, Meier, and Rinderer (1985) found a strong relationship between high self-efficacy beliefs in one's writing abilities and writing performance. The researchers also noted that self-efficacy is partly a result of feedback students have received about the quality of their writing. If high self-efficacy writers also tend to be good writers, they may have transferred knowledge about writing mechanics into long-term memory, freeing themselves to deal more with content and structural matters in their writing assignments, according to Flower and Hayes (1981). They may, therefore, find instructor comments on global issues to be more useful.

Researchers (Daly & Wilson, 1983; Onwuegbuzie, 1999; Pajares & Johnson, 1993) have found negative relationships between writing apprehension and various self-esteem constructs, including self-efficacy. Based upon those findings, it seems reasonable to expect that low self-efficacy writers will also report they use local comments, while low writing apprehension writers will report they use global feedback.

An individual�s expectations of how writing can help them accomplish their goals, both intrinsic and extrinsic, also may play a role in how they process instructor feedback. Vroom (1964, 1970) stated that people who expect to receive rewards, or outcomes, as a result of their efforts will expend more effort on the task at hand. People who do not expect much reward from their efforts will not try hard. For writing students, those rewards could include grades, praise, and good jobs. It seems that students with high expectations to receive rewards from their writing will report they use global comments from the instructors in efforts to improve their second drafts, while students with lower expectations will report they use local comments more often in their revisions.

The hypotheses tested were:

• Students with high levels of writing apprehension will report they use instructor feedback on local issues when revising a first draft more than students with low levels of writing apprehension do.
• Students with low writing apprehension will report they use feedback on global matters when they revise a rough draft more than students with high writing apprehension do.
• Students with high self-efficacy beliefs will report they use instructor comments on global issues in their revisions more than students with low self-efficacy beliefs do.
• Students with low self-efficacy beliefs will report they use comments on local matters when revising first drafts more than students with high self-efficacy beliefs do.
• Students with high writing outcomes expectations will report they use comments on global matters when revising first drafts more than students with low writing outcomes expectations do.
• Students with low writing outcomes expectations will report they use comments on local matters when revising first drafts more than students with low writing outcomes expectations do.

• There will be negative relationships between writing apprehension and the three variables of writing tasks self-efficacy, writing skills self-efficacy, and writing outcomes expectations.
• Writing tasks self-efficacy, writing skills self-efficacy, and writing outcomes expectations will be positively correlated with each other.

The study had practical applications, and added to the bodies of knowledge on cognitive feedback theory and social-cognitive writing theory. This study examined what kinds of feedback students with different writing characteristics find useful when they revise stories, so that both students and instructors might make better use of their time regarding feedback. Writing instructors may find ways to provide more useful comments to students that will result in improved writing for subsequent drafts. They may discover that certain comments can motivate low self-efficacy students to work harder toward improvement. Or, they may reduce other types of comments to lower the apprehension some students may feel toward writing. Students may become aware of their writing characteristics that influence their use of instructor feedback, leading them to better utilize teacher comments that previously were overlooked.

The study was particularly relevant to the journalism field, where few studies have been conducted on student characteristics and feedback. Nearly all of the studies in these areas have come from the English composition field rather than from journalism. The findings of those studies may not be generalizable to journalism, because journalistic writing is different from English composition and other writing disciplines (Berner, 1992; Fox, 1993; Riffe & Stacks, 1988, 1992). In addition, most writing studies have been anecdotal rather than empirical. As journalism programs enter a perilous era of mergers with other departments (Medsger, 1996), more knowledge about the distinctive nature of journalistic writing and journalism students' characteristics seems important to protect the integrity of the discipline from intrusive, disruptive influences by other academic areas.

Because the research design was causal-comparative, the study did not address whether writing apprehension, self-efficacy or motivation cause a feedback preference. Nor did it address the reverse situation: whether instructor feedback can cause changes in writing apprehension, self-efficacy or motivation. Future research into these areas using experimental or quasi-experimental designs may be appropriate to investigate causal relationships. However, the study may help develop understanding about the possible cause-and-effect relationships among the variables and suggest directions for future research.

The study also did not attempt to measure feedback effectiveness, but only student perceptions about the usefulness of different kinds of instructor comments. All of the instruments used in the study were self-report, which always raises the possibility that respondents did not answer honestly.

Because a convenience sample was used, and random assignment was not employed in the research design, the results of the study may be limited in generalizability to a larger population. The study was conducted at one university, which again may limit the generalizability of the findings. Student use of certain kinds of instructor comments may differ at smaller four-year colleges and community colleges. However, a complete description of the sample at the university setting may help make the findings generalizable to other campuses with similar attributes.

Several terms should be defined for the purposes of this study:

Motivation, according to Wlodkowski (1985), describes psychological processes that arouse and cause behavior, give direction to behavior, continue to allow behavior to persist, and lead to preferring a particular behavior. For this study, the motivation construct will have an operational definition of students' writing task self-efficacy beliefs, writing skills self-efficacy beliefs, and writing outcomes expectations, as measured by the instruments described below.

Self-efficacy is defined by Bandura (1986, p. 391) as "people�s judgments of their capabilities to organize and execute courses of action required to attain designated types of performances." Self-efficacy will be measured by a composite of two subscales developed by Shell, Murphy, and Bruning (1989).

McCarthy, et al. (1985) defined writing self-efficacy as students' evaluation of their own writing skills. Writing self-efficacy consists of three components -- writing skills, tasks and outcomes -- that will be measured by subscales developed by Shell, et al (1989).

Writing skills self-efficacy beliefs are students' confidence in their ability to successfully perform certain writing mechanics, such as spelling and punctuation.

Writing task self-efficacy beliefs are students� confidence in their ability to successfully complete specific writing problems, such as a letter to a friend or a job resume.

Writing outcomes expectations are how students rate the importance of writing for achieving various life goals, such as getting a job or being financially secure.

Feedback is defined as written or verbal commentary to the student from the instructor or peers concerning various aspects of the student�s writing, including basic skills and content. According to Wlodkowski (1999), feedback is information that learners receive about the quality of their work, such as notes on written assignments and comments about emerging skills.

Use of instructor feedback is operationally defined as whether a second draft is changed based upon the teacher�s written comments on the first draft. Student perceptions of the usefulness of feedback will be measured by a questionnaire based upon an instrument used by Straub (1997).

Global feedback is a comment upon a written composition�s content, such as the ideas, development and organization.

Local feedback is a comment on mechanical writing problems, such as wording, sentence structure and correctness.

Writing apprehension is a construct that attempts to differentiate people who find writing enjoyable and those who experience high levels of apprehension when writing is required (Daly & Miller, 1975a). Faigley, Daly, and Witte (1981) further defined it as "a construct associated with a person's tendencies to approach or avoid situations that require writing accompanied by some amount of evaluation" (p. 16). The Daly-Miller Writing Apprehension Test will be used to measure the construct.

Instructor feedback plays an important role in improving student writing. However, many students ignore, misunderstand, or do not use teacher comments when they revise first drafts. Writing apprehension and writing self-efficacy beliefs are powerful affective influences on student writing performance. This study examined the relationships among writing apprehension, writing self-efficacy, and student use of instructor comments during the revision process. Top

Writing is not only a cognitive process but also an emotional activity, according to McLeod (1987), who called for writing researchers to develop a theory of affect. While much research has examined how people think when they write, fewer studies have been conducted on how people feel during the writing process. The most important emotions that McLeod identified for a theory of affect are writing apprehension, motivation, and students� beliefs about writing and their own skills as writers. Furthermore, Kulhavy and Stock (1989) suggested that the amount of time a learner spends on instructor feedback and the learner�s response to the feedback may be influenced by the learner�s emotional and motivational states. This literature review will examine the relationships among the relevant literature not only on writing apprehension, motivation and student attitudes, but also the literature dealing with social-cognitive and cognitive-feedback theories.

The underlying writing theory for this research is the social cognitive writing model, partly because of the cognitivist emphasis on feedback in the writing process (Glover, et al., 1990). This section of the literature review focuses on primary sources � the work of Bandura, and Flower and Hayes. A secondary source (Pitts, 1989) was identified through an ERIC search and is included because it extends cognitive writing theory into the journalistic domain.

Bandura's work on social-cognition is central to this study because his self-efficacy theory addresses motivation, apprehension, feedback and other behaviors related to writing. His self-efficacy theory (Bandura, 1986) deals with intrinsic rewards, in which successful experiences will lead to a sense of mastery and control over the learning environment. The locus of control can be internal, in which a student perceives that his or her efforts will result in success, or external, in which the student believes that success is determined not by effort but by environmental factors outside of his or her control. Bandura considered human behavior to be largely a result of the interaction of three types of motivation: incentive, vicarious, and self-produced. Incentive motivators are external, environmental incentives such as monetary, social, sensory, activity, status, power, and primary incentives (which provide the basic needs of food and drink). These are largely behavioral influences because actions that bring rewards are generally repeated, while those that bring punishment tend to be discarded. Incentive motivators for journalism students could include grades, praise, publication of their stories, and awards in press competitions.

But Bandura added that people do not learn just from first-hand experience and trial and error. Vicarious motivators allow people to observe how others act, and what benefits and penalties others receive as a consequence of their actions. Seeing others get rewards for their acts increases the likelihood that the observer will act in the same manner. For educators, vicarious motivators include modeling desired behaviors or actions for students. In journalism classes, modeling could include the instructor providing examples of correct sentence structure in written comments on student papers. Bandura contended that modeling behaviors that receive rewards is more effective than only modeling with no consequences. For teachers, that indicates the importance of including a grade with comments on a paper. Vicarious motivators also lead to social comparisons. For instance, if two students perform the same act and one student is praised more than the other is, the student who received less praise will feel punished by the inequitable reward. For this study, the possibility of social comparison points out the need for equitable treatment of students when providing written feedback on their writing.

But external incentive and vicarious motivators do not control people's behavior, Bandura argued. Self-produced motivators, such as self-directedness, self-evaluation and internal standards of behavior, give people some control over their thoughts, emotions and actions. He used the act of writing as an example of behavior that is self-regulated through self-evaluation. Writers, he observed, continually revise their work during the writing process until their work meets the writers� self-set standards. In light of this study, although students receive instructor feedback and modeling of good writing, ultimately the students� own standards of what constitutes good writing and their ability to self-evaluate their writing will regulate those external motivators.

In cognitive process theory, there are three major elements that constitute the act of writing: task environment, the writer�s long-term memory, and writing processes. These elements interact in complex and non-linear fashions (Flower & Hayes, 1981). For the purposes of this study, instructor feedback should be considered as a component of the task environment, and it can impact both long-term memory and writing processes. For instance, feedback from the instructor can help define or clarify the writing task or problem so the student can resolve it. Feedback may also provide cues that allow the writer to retrieve useful knowledge from long-term memory, such as grammatical knowledge. Feedback also can cause the writer�s schemata to grow and change as new information is either assimilated or accommodated.

Cognitive process theory also helps explain the motivation of writers. The theory proposes that the act of writing is goal-directed and that writers create their own goals. It follows that a writer who is capable of setting his or her high-level goals and sub-goals that support the high-level goals is likely to be more motivated to complete the writing task than a writer whose goals are vague or limited. Content goals (what a writer wants to say to an audience) should be strong in journalism students because their stories are fact-driven and the audience is defined. Since goals help guide the writing process and provide a "logic" (p. 379) for it, goals that are sensitive to an audience and that are tied to a topic may be forceful and powerful motivators.

Translating ideas into writing is another part of the cognitive process that seems to interact with the writer�s long-term memory. That interaction can have an affective result for writers, perhaps even leading to writing apprehension. One translation task involves deciding on the mechanics of writing, such as the local issues of spelling and grammar. The other is a global task of deciding content and organization -- what the writer wants to say and how to say it. The multiple acts of translation can lead to frustration for some students. For instance if they decide that translating spelling should take precedence over organization, their text will be poorly planned and incoherent. If they decide to forego spelling and concentrate on the global writing task, the text will be full of mechanical errors. This seems to be a potential problem for students with poor writing skills. According to cognitive learning theory, these students may not have rehearsed sentence structure knowledge in their short-term memory and, therefore, it was never moved to permanent, long-term memory. For skilled writers, retrieving sentence structure knowledge from long-term memory is nearly automatic, freeing them to translate global writing tasks. The frustration that some writers experience while unsuccessfully trying to juggle and balance the translation process could lead to writing apprehension, since a history of failure can cause writing apprehension (Daly, 1978).

However, writing cannot be viewed as purely a cognitive act. Later research has placed cognitive process writing theory into a social, cultural and political context, with which cognition interacts (Flower, 1989). In this scenario, writing is viewed as an act within a discourse community and as a contribution to a larger conversation. For example, journalists have goals of maintaining an informed electorate, informing and entertaining an audience, and chronicling the ongoing events of a community. Certainly, the cognitive aspects of composing a news story are deeply situated within a social, cultural and political community whenever reporters write for their audience. In the classroom, the writing process also becomes a social, cultural and political act as the forces of instructor, peers and individual writers compete, negotiate and collaborate. A journalism student may receive written comments from a teacher that conflict with feedback received from a classmate. In revising the story, the student must weigh the politics of accepting or rejecting the instructor comments, the social pressure of a peer�s opinions, and the overall context of acting in an academic community with grades, degrees and future careers hanging in the balance. In this manner, the context will largely affect the writer�s thinking, acting as a cue to cognitive action.

Social cognitive theory does not merely expand the boundaries of the task environment, though. It acknowledges that while context affects cognition, individual goals, learning styles and prior knowledge create individual differences among writers. Students will interpret instructor and peer feedback and other contextual features according to their individual differences. The classroom becomes a shared community of writers, but individual cognition mediates the context of that community. Even though it is situated in a social, cultural and political context, writing remains a particularly individual act.

A third aspect of social cognitive writing theory is that writers will negotiate and construct meaning out of the various contextual purposes. For instance, students must decide whether to accept or reject the traditional purpose of journalism, which has been defined by the culture of journalists as informing the public about events of public interest. As students recognize the boundaries of journalistic purpose, they must next create an individual purpose that meets their goals and still allows them to operate within the boundaries of journalism. A crucial part of constructing purpose in the classroom involves negotiation with the instructor, through teacher feedback on assignments or other communication.

In another study of social-cognitivism, Flower (1994) also examined how affect, including self-image, emotion, motivation and attitudes, influences student writing when it interacts with cognition and context. Reflecting the powerful role that affect plays, some students described their writing processes as "dilemma-driven action" ( p. 243) rather than problem-solving strategies. Students also reported having little or no sense of control of these feelings, and attributed their success or failure at writing to external factors, such as luck or time. It is important to note, though, that many students did report positive feelings toward writing. The addition of the emotion/affect component to social-cognitive writing theory puts this study in a new light since writing apprehension and motivation apparently play critical roles when they interact with societal forces and cognition.

Using the protocol analysis methodology of Flower and Hayes (1980), Pitts (1989) developed a process model for journalism that identified two main steps in news writing. Reporters first focus on writing the lead, or beginning, of their stories before moving on to write the rest of the text and make revisions. Unique activities in news writing include the lead writing, frequent use of memory recall, the development of only short-term goals and plans, and the integrated use of rereading and editing as part of the writing process. Pitts does not situate the cognitive process of news writing in a social, cultural or political context. However, she does note that journalism students usually write for their teacher as the audience, and that students are less likely to reread their stories for a sense of continuity than are professional journalists. Students tend to look for mechanical errors rather than clarity of meaning. Pitts urged teaching methods that incorporate more interaction between the instructor and students during the writing process.

For journalists and journalism students, the act of writing a story is a cognitive process involving problem solving, decision making and critical thinking (Medsger, 1996). Her description of a reporter�s mental gymnastics echoed Pitts� findings of how reporters continuously reread and then revise their stories in a non-linear process. "The well-trained journalist�s mind inquires, weaves, thinks again, unravels, asks again, corrects, goes back again, weaves again and then � decides what the story is and creates the words and images to tell it" (p. 9).

Most research on writing instruction agrees on the importance of providing some form of feedback to student assignments. Database searches in ERIC, Colorado Alliance of Research Libraries (CARL), PsycInfo and Dissertation Abstracts were conducted on the topic of feedback using the broad descriptors of writing evaluation, feedback and teacher responses. More detailed searches added the descriptor terms of higher education, student attitudes, writing instruction, writing research, cognition, learning, and student reaction.

Many composition instructors are adopting feedback methods that are less critical of students� basic writing skills. Their feedback methods concentrate more on the message of the students� writing, particularly with the advent of computerized writing centers (Coogan, 1995, Hawisher & Moran, 1993). Some research has focused upon the importance of establishing a two-way dialogue about writing between students and teachers (Fey, 1993; Ziv, 1984) so teachers don�t take control of the student�s writing through their written feedback. Establishing an on-going dialogue about student writing may create an andragogical approach to instruction in which student and instructor share knowledge and become co-learners, while leading students to take responsibility for their learning (Kinkead, 1988). While some instructors may still feel compelled to address basic writing problems, such as grammar or spelling, they do it sparingly as in "selective correction" (Walling, 1987).

But English composition classes differ from journalism writing classes in terms of audience. Much composition writing is private, such as keeping journals, while journalism writing is intended for a mass audience. In addition, composition students commonly write essays in which they express their ideas, while journalism students write stories based upon facts (Riffe & Stacks, 1988, 1992). As a product for public consumption, with many critical and alert readers, journalists do not have the liberty of "selective correction." Their stories not only must be factually correct and complete, but also must have proper spelling, punctuation, grammar, sentence structure and word usage. "A good journalist gets the mechanics right because that is what the audience understands" (Berner, 1992, p.2). "Poorly written and badly organized stories deter readers from gaining information that may be essential to their well-being" (Fox, 1993, p. vii).

To achieve this high standard of writing, journalism instructors traditionally have provided feedback to students by using proofreading marks to correct basic errors, along with comments in the margins and at the end of stories. Novice journalism students often find their stories marked with so much red ink from the instructor that the stories have earned their own slang term in journalism schools - "blood sheets." In particular, a weak writer who needs the most help, but whose papers receive the most teacher comments, may view his or her graded paper as a "messy autopsy" with the instructor as "coroner" (Grant-Davie & Shapiro, 1987, p. 6).

The nature of feedback has a long history in learning theory. Feedback�s role in behavioral learning theory is to reinforce desired student responses. Reflecting the importance of information processing in cognitive learning theory, Briggs and Hamilton (1964) defined feedback simply as information to the learner. In cognitive learning, feedback can inform the student why he or she is wrong, suggest how the learner can find out whether the answer is correct, give direction, and help the student decide whether to further study the same material or move on to new subject matter. Briggs and Hamilton alluded to the affective function of feedback, noting how it can encourage or motivate students to greater effort, and can also serve as a reward for effort. For anxious students, including those with a history of academic failure, feedback can reassure and encourage them to continue their academic efforts.

Mory (1996) listed three types of feedback. Feedback serves as a motivator to increase performance. Feedback provides reinforcement connecting responses to prior stimuli, resulting in correct responses. Feedback also provides information learners can use to validate or change previous responses, mainly focusing upon error responses. This study will primarily consider feedback as information, and how the learner cognitively processes information in short-term and long-term memory. However, the study is also interested in whether feedback serves an affective function that changes motivation, self-efficacy beliefs, and writing apprehension levels.

Most communication theory models include interactive feedback loops between the receiver and the transmitter of messages. Feedback represents the receiver�s response to the message sent (Heinich, Molenda, Russell & Smaldino, 1996). In educational settings, feedback from students allows the teacher to determine whether instruction has been successful. Feedback from the teacher allows students to know whether their efforts have been successful, and if not, what needs to be done to improve their work. Instructional feedback can be subject-specific, directions to students on how to proceed, questions, suggestions, or other information. Heinich, et al. emphasized that the instructor must carefully choose a medium to ensure that the student receives the message clearly and accurately. Several writing researchers have noted that students often misunderstand instructor feedback (Sommers, 1982; Williams, 1997; Ziv, 1982).

The Shannon-Weaver communication model includes the component of noise (Black & Bryant, 1995), a distorting factor that can disrupt the message during its transmission through a channel. The Shannon-Weaver model sought to address three communication problems, and in this study the semantic problem and effectiveness problem are of particular interest. Semantically, it is important for the instructor to choose appropriate words to increase the precision of conveying feedback. Regarding effectiveness, the instructor�s feedback should produce desired changes in the student�s conduct in the manner the instructor intended. The semantic and effectiveness problems can be caused by noise. In early transmission models, noise was physical or environmental � glare on a blackboard, static on a radio signal, or background sounds. In this study, noise can be psychological or cognitive in nature. For example, writer�s apprehension may be so severe for some students that they ignore an instructor�s feedback because of a history of failure on writing assignments (Daly, 1978). Students may misinterpret what an instructor intended to say in written feedback (Williams, 1997; Ziv, 1982).

Bandura (1986) stated that people could learn socially by observing other people who had adopted new ways after seeing them modeled. For instance, instructors can model desired writing behavior by providing examples of good writing in their feedback. Wlodkowski (1999) stated that feedback may be most powerful process that teachers can use to influence a learner�s competence, which Bandura (1986) would call the learner's sense of self-efficacy. Ziv (1984) stated that a direct way to affect student writing performance is writing comments on their papers. However, she said one problem is there is no model for instructors to follow for their comments. Furthermore, Ziv pointed out a gap in the research on how teacher responses affect student writing performance.

Wlodkowski (1999) stated that effective feedback is quick, frequent and positive. Feedback enhances motivation because learners can evaluate their progress, correct errors efficiently, receive encouragement from instructors, self-assess their work, and maintain their efforts toward realistic goals. Foster (1992) also emphasized the importance of providing speedy feedback to student writing. His stimulus-reinforcement behavioral model is based upon the premise that the more quickly an instructor provides feedback, the stronger its effects on student writing behavior. Long delays in giving feedback tend to cause students to repeat their errors because other concerns and tasks have arisen that divert their attention. However, feedback theorists (Briggs & Hamilton,1964; Kulhavy, 1977) had much earlier noted that many studies found delaying feedback for up to a day or more resulted in better performance than providing immediate feedback. That phenomenon is called Delayed-Retention Effect or DRE. While the findings on the speed of feedback may be contradictory, it is possible that the timing of feedback will affect writing apprehension and motivation.

The nature of instructor feedback may also affect student motivation and writing apprehension levels. One of the first studies on student reaction to instructors� written comments (Gee, 1972) found that students who received negative comments tended to write less frequently and had significantly more negative attitudes toward writing than did students who were praised by instructors� feedback.

Ziv (1984) recommended establishing an ongoing dialogue between teachers and students on their writing. However, according to Ziv, one problem is that students see those comments as evaluation rather than the response of an adult reader. In four case studies of college freshmen, Ziv found that students responded favorably to instructor comments that were explicit, offering specific suggestions on how to reorganize stories or specific corrections on their writing errors. For example, comments such as "rephrase this sentence" did not result in a better second draft because students did not understand how to rephrase or what to rephrase.

Other research conflicted with Ziv�s findings that students wanted explicit, specific comments from instructors on how to improve their writing. In a study in which freshmen writing students rated teacher comments on an essay, Thompson (1993) found no relationship between different student learning styles and how students rated the comments. He did, however, allude to the situational aspect of writing, noting that findings showed the influence of social context on the way students used teacher comments. Noting that freshmen at military school where the research was conducted are very busy, Thompson reported that students there only wanted teacher comments that would let them immediately correct writing errors, not expand upon the context of their essays. A later study of 400 students by Thompson (1994) indirectly addressed what types of teacher comments affected student motivation. That study found that specific comments on how to change their papers made students feel the most comfortable. He found that students wanted teacher responses that would help them get better grades. He suggested that instructors might apply Maslow�s hierarchy of needs by meeting students� lower-level needs, such as correcting spelling and grammar, before moving on to meeting higher-level needs, such as improving content and organization of papers.

A different student reaction was found by Dickinson (1992), who provided students in a writing course with a lengthy, detailed comment page written on a word-processing program attached to their papers rather than hand-writing comments in the margins of their papers. In course evaluations, her students said that they found written comments to be more helpful than any other type of teacher response. Dickinson also found that revised papers were improved in content, organization and editing over first drafts.

Ziv (1982) conducted case studies on four writing students and found that written comments from the teacher created confusion for students because their perception of their papers did not agree with the instructor�s perceptions. Students reacted to the confusion in three ways: by trying to follow the instructor�s suggestions in the next draft, by defending their papers against the instructor�s criticism, or by avoiding dealing with the comments. Instead, they simply deleted words, phrases and passages that drew comment. Ziv concluded that the findings suggested teacher comments are not helpful to student writers and that further research is needed on the manner of commenting.

In a study of eight college students in a first-semester writing course, Williams (1997) found that even the successful writers were confused by teacher comments that used unfamiliar rhetorical jargon from the composition field. However, both successful and less successful writers correctly interpreted 76 percent of teacher comments on their papers as the instructor intended. Williams concluded that rather than being confused by teacher comments, most students may either lack the ability or motivation to solve problems raised by instructors in their written comments.

Also noting that students often seem confused by teacher evaluations of their writing, Burkland and Grimm (1984) found that 52 percent of the 197 freshman composition students who completed questionnaires said they found teacher comments helpful in writing their next papers. Their study also found that 71 percent of the students considered critical comments from teachers to be the most motivational compared to comments that were praise or compliments. They concluded that students want specific, critical suggestions from instructors, and that students did not find complimentary comments to be useful.

Research into student use of instructor feedback also should consider the next step in the process: revising the first draft. McKeachie (1999) suggested that unless a student has an opportunity to write a second draft after receiving teacher feedback, his or her writing may not improve. Not all student writing needs to be revised, Hillocks (1990) stated, but major assignments should undergo the revision process. The act of revising a written assignment can also create emotive conflicts for students, Mohr (1984) noted. Some may be nervous because revision can be threatening, while others may confidently view revision as a way to gain control over their writing. Some see revision as tedious, while others consider it exhilarating. They also seemed more comfortable when correcting spelling, punctuation and other local problems. But as for global issues, "they were uncertain about what to do with the rest" (p. 7).

A study based upon Bandura�s (1986) self-efficacy theory of motivation found that students will benefit from written teacher feedback when they perceive that acting upon the comments will improve their ability or improve their performance (Lackey, et al., 1997). In their correlational study of 137 students in a second-semester freshman composition course, they found that grades resulted in the highest motivation among different forms of written feedback. Schunk (1991) added that the timing of feedback is important. Providing feedback on a student�s early successes in a class can enhance the student�s learning efficacy. Ability feedback (comments such as "You�re good at this") improves self-efficacy better than does effort feedback (comments such as "You�ve been working hard on this"). Receiving feedback about the quality of their writing helps students develop their writing self-efficacy (McCarthy, et al., 1985). In their study of 137 freshman-writing students, they found the strength of a student�s perceived writing efficacy was a significant predictor of writing performance. The other three predictors tested in the research were not statistically significant.

Grades were the best motivator among the types of written, instructor feedback, concluded Lackey, et al. (1997). Previously, Bell and Price (1982) conducted an experiment to determine the effect of grades on student writing. They withheld grades from some students in a composition course but provided written comments. Other students received both grades and comments on their papers. Their study found no significant difference in writing improvement or lack of improvement between the two groups. They concluded that, while withholding grades raised anxiety in students, they were able to deal effectively with anxiety. Highly motivated students may actually benefit from increased anxiety, they concluded.

Noting that much written feedback lowers motivation and fails to improve student performance, Lackey, et al. (1997) added that, based upon their research, the most effective feedback contained the most task-specific comments, rather than general suggestions, to improve student writing performance.

Based upon research indicating that students react negatively to teacher comments, instructors may be tempted to provide no written comments at all on student papers. But other research suggests that without any form of written comments, students have no motivation to change their writing. Sommers (1982) concluded that teacher comments create the motive for students to revise their next draft of a paper and that without comments, students will see no need for revision. However, revision entails taking a risk, and many students won�t risk changing anything not commented upon by the instructor, even if the student senses that revision is needed. The challenge for teachers is to write comments that will provide the motivation for students to revise. Furthermore, classroom exercises need to be designed to help students overcome their apprehension about revising.

Sommers pointed out that there are two different locations for written comments on papers � interlinear and marginal � and often these two locations result in different kinds of suggestions. The comments between the lines usually ask students to edit for grammar, spelling, and other local issues, while comments in the margins ask students to further develop ideas or to conduct more research. The two types of comments may lead students to see revision as two separate tasks � editing and developing ideas � while revising should be considered as part of a holistic writing process.

Pullman (1999) argued that it is futile to apply writing-process theory or any other theories to such an "unstable entity" (p. 27) as writing because there are too many variables involved. He may have alluded to the miscommunications that can arise between teacher and student when discussing a text as the "effect of multiple interpretative efforts by people who may or may not share contexts or interpretative practices, who may or may not occupy the same context at the moment of the text, who may in fact have the text in common only as a site of combat over other issues ..." (pp. 27-28). In Pullman�s view, instructor feedback could be misinterpreted because a teacher may view comments as pedagogical, while a student views the same comments as attempts by the instructor to seize control over his or her writing. The written product is unstable because of "multiple interpretative efforts," and, therefore, no theory of writing will ever apply to all writing processes.

Straub and Lunsford (1995) categorized editing comments as local in focus, while comments referring to ideas are global in focus. They also identified different modes of comments, referring to how the comment is framed. Each mode assumes a different degree of control over a student�s writing by the instructor. Control refers to how the teacher directs changes to be made in later drafts of the writing, and all comments involve some degree of control, they contended. For instance, comments framed as questions indirectly call upon the student to revise. However, a comment framed as an imperative demands that the student correct an error. They conducted a study in which 12 writing instructors wrote comments on 15 student essays. They found that 55 percent of the instructors� comments were global, 21 percent local, and the rest had other focuses. Fifty-nine percent of the comments were questions and reflective statements, which are the least directive modes. They concluded that marking and commenting upon local errors is a low priority for writing instructors in contemporary composition studies. This study will examine whether certain writing characteristics influence whether students use local and global instructor feedback.

One problem with research on student reaction to teacher comments may be that the types of comments studied are too broadly defined by the researcher, according to Straub (1997). His study of 147 freshman writing students asked them to rate their preferences on 43 types of teacher comments and then to explain the reasons for their choices. His findings indicated that students preferred comments on both their ideas and on editing; they preferred comments that were specific and clear; and they preferred comments that provided direction, were helpful and didn�t insist they revise a certain way. Students didn�t like criticism but did like praise. He suggested that teachers consider whether students are interpreting their comments as the teachers intended.

Dohrer (1991) also found that students often misunderstand teachers� intentions in written comments. Students revised papers mainly to get higher grades by meeting what they perceived to be the expectations of their teachers, based upon written comments. As a result, students become correctors of mistakes pointed out by the teacher rather than writers trying to communicate with readers. He concluded that students were not confident in their abilities to revise papers; in other words, they had low writing tasks self-efficacy beliefs.

There is further evidence that students prefer being correctors rather than becoming more skilled writers. Instead of polishing their skills in creating coherent, thoughtful papers, students may just want to fix the immediate mechanical problems. In a survey of 1,200 students from ninth grade to the second year of college, Land and Evans (1987) found that a majority of respondents wanted written comments to provide explanations of how to "fix mistakes" and how "to make my paper better" (p. 115). Their study also raised questions about the value of written comments as a form of instructor feedback. About 88 percent of the participants reported that individual conferences with teachers about their writing were most helpful.

A meta-analysis of research conducted on the effects of feedback type on writing quality concluded that there was no significant difference between treatment and control groups (Hillocks, 1986). However, he stated that positive comments generally seem more effective than negative feedback in improving writing.

The conflicting evidence may lead one to conclude that, rather than risk increasing a student�s apprehension by writing negative comments on papers, why not take the safe route and only write positive, congratulatory slaps on the back? Fox (1980) said that instructors should not lead students to believe they are better writers than they actually are by writing only "sugar-coated" (p. 23) comments. Besides being deceitful, providing false comments will not help students improve their writing because their mistakes will never be pointed out to them. In the long run, feedback containing only positive comments may hurt students� confidence in their writing ability, lowering their writing self-efficacy. Foster (1992) stated that providing only positive feedback can create "a fool�s paradise of students who love their own writing so much they will be devastated when somebody criticizes it later" (p. 214).

Feedback can also depend upon teacher expectations of students, according to Daly (1979), who researched how elementary and secondary teachers perceive high- and low-writing apprehensive students. The study found that teachers significantly expected high-apprehensives to be less successful in other academic classes and less likely to succeed in the future than they did low-apprehensives. Furthermore, teachers were less likely to give positive recommendations to other teachers about high-apprehensives than they were for low-apprehensives. A gender bias also was apparent. The teachers viewed low-apprehensive females most positively and low-apprehensive males the least favorably. This may not be that surprising considering that the sample was all female (n = 33). But Daly expressed concern that low-apprehensive students who readily accept writing assignments could be evaluated more positively by teachers than high-apprehensives who are hesitant about assignments and also tend to turn in poorer quality writing.

There is also an interaction between student expectations of their performance ability and feedback, according to Kulhavy (1977). Students who are confident that they answered a question correctly, and did, will not pay much attention to feedback. However, students who were highly confident in their answer, but were wrong, will pay the most attention to feedback as they attempt to correct their mistakes. Feedback will probably have minimal effect on students who had low confidence and answered questions incorrectly. These students probably did not understand the material, the question, or both, and need to develop new learning strategies. In light of Bandura�s theory, Kulhavy�s observation that supplying feedback for errors is probably more important than acknowledging correct answers may be applicable to students with high self-efficacy beliefs. But Kulhavy was addressing the effects of feedback on performance, while Bandura also addressed how feedback influences people�s feelings. Although Kulhavy contended that feedback will have minimum effect on the learning performance of low-confidence students, it could affect their apprehension and motivation levels.

Increased time spent reading instructor feedback will improve student performance, a study of 60 undergraduates found (Kulhavy, Yekovich & Dyer, 1976). Students remembered more correct information when they studied longer, and they were more likely to study longer when a response they were sure was right turned out to be wrong. The implications for teaching are that if a student understands the test question, providing feedback will increase subsequent test performance regardless of whether the pretest answer was correct. However, feedback will be ineffective if the material was not comprehended in the first place. Instructors, therefore, should be sure the design and content of test materials are appropriate for students, and then provide feedback after each response. A subsequent study (Kulhavy, Yekovich & Dyer, 1979) reaffirmed and expanded upon those recommendations. The study of 120 undergraduates found that students who had low confidence in their response but guessed the correct answer, also spent considerable time studying both the feedback and the test item to increase their understanding. They concluded that feedback has a powerful influence on studying and test behavior.

While the previous Kulhavy and associates studies examined instructor feedback on multiple choice and other forms of tests, they do suggest how much feedback instructors should provide to written assignments. A 1989 paper (Kulhavy & Stock) concluded that simple feedback ("yes," "good," etc.) is most efficient when there is little discrepancy between a student�s certainty and a correct answer. Those students will probably pay little attention to feedback. However, when there is high discrepancy between a student�s certainty and an incorrect answer, more elaboration in the feedback makes sense because high-confidence errors are most likely to be corrected. Extensive, instructional feedback is needed when a student had low confidence, no matter whether the answer was right or wrong, because the student probably did not comprehend the material.

The advent of computer writing laboratories should also be considered in assessing student perceptions of teacher comments, especially since spelling and grammar checking programs can help students resolve many local or mechanical problems in their compositions. In addition, the use of computer labs may affect the nature of instructor feedback. Berner (1992) said that he now requires students to turn in electronic versions of their stories, and "I make more comments than I would with pencil" (p. xiii). But others contend that instructors have written too much on students� papers in the past, and electronic feedback will lead to more focused discourse between instructor and student (Newbold, 1994).

Smith's 1990 study of student attitudes toward computer-assisted instruction (CAI) compared journalism students who used a computer program that provided feedback to students who received only instructor feedback. Both groups used the feedback to rewrite the leads of their stories. He found that the computer group judged the assignment more interesting, believed they learned more, and thought they received more feedback as well as more useful feedback than did the paper group.

Not all researchers have found such success with CAI. Fischer and Grusin (1993) hypothesized that student performance and satisfaction would be higher among those who used grammar checking software rather than students who received only instructor feedback on their stories. In contrast to Smith's 1990 finding of increased student satisfaction with CAI, their study found that the mean satisfaction score for students who did not use the grammar checkers was significantly higher than the mean for the computer group.

However, a 1993 study by Smith, as well as Fischer and Grusin's 1993 grammar-checking software research, pointed out a flaw in CAI -- that students tended to use the computer programs as shortcuts to finish assignments, rather than as learning strategies to understand the theories and concepts behind writing. Fischer and Grusin (1993) even concluded that "grammar checkers may detract from the learning rather than enhance it" (p. 26). These findings suggest that even with the advantages of the computer writing lab, many students still focus on local issues rather than global concerns in their texts.

Although writing apprehension is recognized by writing experts as a common problem, there has been surprisingly little research conducted on the condition (Salovey & Haar, 1990.) They commented, "It is striking that writing apprehension, as compared to other performance anxieties like test apprehension or the fear of public speaking, has had little research attention paid to it, and yet it is quite prevalent and dominates so many aspects of the sufferer�s life" (p. 524). A brief flurry of studies ensued after writing apprehension was identified in 1975, but only scant research in the last decade, and relatively few of those studies involved undergraduate college students (Onwuegbuzie, 1998).

Searches on PsycINFO, Dissertation Abstracts, ERIC and CARL databases were productive for locating primary and secondary sources on writing apprehension. Because the work of Daly and associates first identified the construct and subsequently explored the many ways writing apprehension affects students, the literature review concentrates on the Daly and associates research as a primary source. The initial broad search used the descriptor terms writing apprehension and writing anxiety. Further searches added descriptor terms including higher education, writing evaluation, writing research, feedback, writing skills and writing ability. Because evaluation anxiety research preceded writing apprehension studies, searches were conducted on that topic, first using the broad descriptors of test anxiety and evaluation anxiety. Other searches included the descriptor terms higher education, college students, and student attitudes.

Writing apprehension is a construct that attempts to differentiate people who find writing enjoyable and those who experience high levels of anxiety when writing is required (Daly & Miller, 1975a). Anxiety is marked by fear, phobia or apprehension. Faigley, et al. (1981) added a feedback component to their definition of writing apprehension, labeling it "a construct associated with a person�s tendencies to approach or avoid situations that require writing accompanied by some amount of evaluation" (p. 16).

There is ample evidence (Daly & Miller, 1975a; Daly & Wilson, 1983; Pajares & Johnson, 1993) that rather than being two ends of a continuum with highly apprehensive at one extreme and highly motivated at the other, motivation to write and writing apprehension are distinct constructs. A study of 164 undergraduate students taking composition and interpersonal communication courses found no significant correlation between achievement motivation and writing apprehension (Daly & Miller, 1975a). That suggests that a student could be very motivated to succeed in a writing class, but also could have a high level of writing apprehension.

Later research found more evidence of the independence of writing apprehension as a construct. Daly and Wilson (1983) conducted 13 comparisons between writing apprehension and other general personality constructs with college writing students. Among the findings relevant to this study, writing apprehension was significantly and negatively related to self-esteem, and positively correlated with oral communication anxiety. It seems reasonable that how people feel about themselves will affect their self-efficacy. Someone with high self-esteem likely will feel more confident about their ability to handle a learning task. Daly and Wilson also suggested that evaluation, which is a form of feedback, plays a part in forming a person�s self-esteem, and that a history of positive evaluations will generally lead to higher self-esteem. Conversely, a history of poor evaluations could result in negative feelings toward oneself.

Another study supported the distinct constructs of apprehension and motivation. The study examined self-efficacy, writing apprehension and writing performance. During a semester, some undergraduate education students received instructor feedback on their writing skills, while others did not. Pajares and Johnson (1993) found the writing apprehension levels of the students did not change even though their performance and sense of self-efficacy improved. They also found no correlation between writing apprehension and writing performance. However, the results of the study may not be generalizable because of the small sample size (n = 30) and the gender inequity (five males, 25 females).

Other studies have examined self-concepts related to self-efficacy. In a 1998 study of 89 graduate students, Onwuegbuzie found a significant, negative correlation between writing apprehension and self-motivation. However, a subsequent multiple regression showed that self-motivation was not a significant predictor of apprehension. He concluded that students who lack self-motivation may tend to procrastinate starting writing projects. Onwuegbuzie (1999) conducted a study with graduate students (n = 97) and found significant, negative correlations between writing apprehension and five self-perception measures: perceived creativity, intellectual ability, scholastic competence, job competence and self-worth. A multiple regression indicated that the level of a student�s perceived scholastic competence was the best predictor of writing apprehension, explaining 22 percent of the variance in writing apprehension. He concluded that students with low perceptions of their own academic abilities tend to have higher writing apprehension.

Adult learners often suffer from various forms of evaluation anxiety. Although more research has been done on test anxiety than on writing apprehension, there are parallels between them since both involve forms of anxiety. A brief examination of test anxiety, also known as evaluation apprehension, seemed appropriate for this study because students with high writing apprehension often express anxiety toward evaluation � instructor feedback, in this case � of their written assignments.

Most research in this area has recognized that anxiety consists of two components: worry and emotionality. Worry is defined as the cognitive concern about performance. For example, a student may lack confidence in his or her ability to write a term paper. Emotionality is defined as the physiological arousal response to test anxiety, such as perspiration and an accelerated heartbeat (Spielberger & Vagg, 1995).

Such anxiety can become a lifelong condition. Sogunro (1998) listed eight causes of test anxiety for adult learners, of which negative feedback, setting of unrealistic goals, unfriendly learning and evaluation environment, impromptu assessment, and inadequate preparation seem related to what is known about the possible causes of writing apprehension.

While no research has been done to gauge the extent of writing apprehension, its relation to test anxiety indicates that it may be widespread. "Test anxiety is a pervasive condition among college students," another study concluded (Hashway, Duke & Hammond, 1992, p. 180). Furthermore, test anxiety that develops early in a person�s schooling may persist into his or her adult learning years (Sogunro, 1998).

Anxiety is not necessarily a negative condition in all learning situations and for all individuals. Sogunro noted that a moderate level of anxiety "can be beneficial to learning" (p. 110) because it increases motivation, heightens alertness and concentration, and thereby improves performance. Most people likely experience normal levels of test anxiety and are able to overcome the negative aspects to pass and even excel on exams. However, other students can be crippled by higher levels, and their exam performances may suffer as a result.

One study partially supported Sogruno�s observations. Although their study involved an experiment with ninth-grade writers rather than with college students, Lynch and May (1977) found that high anxiety conditions improved the writing performance of high-creative writers, but the performance of low-creatives suffered under the same conditions. So it is likely that a certain level of apprehension will actually benefit some writing students, serving more as a motivator than an inhibitor to performance.

Daly (1978) found that students with high levels of writing apprehension consider writing to be unrewarding, and they will avoid classes with writing assignments, if possible. Apprehensive students will choose academic majors that they believe will require less writing, while non-apprehensive students will seek out majors where more writing is required (Daly & Shamo, 1978). The effects of writing apprehension can continue after college. High-apprehensive students tend to enter occupations where less writing is required as part of the job (Daly & Shamo, 1976).

Furthermore, writing apprehension can affect a student�s expectation of success not only during current instruction, but in future writing classes. Daly�s 1978 study of 3,602 undergraduate composition students examined the differences in writing skills between low- and high-apprehension students, and concluded that low-apprehension students tend to have higher writing skills than do high-apprehension students. The lower level of skills furthers the writing apprehension because a student who lacks the necessary skills to write will not experience much success in writing courses. The researcher theorized that the lack of success, compounded with high levels of apprehension, will lead to students avoiding writing practice and ignoring evaluative feedback on their writing.

Daly and Hailey (1984) determined that writing apprehension has two dimensions: dispositional anxiety and situational anxiety. Dispositional apprehension is the most commonly studied form and refers to different levels of apprehension (high, moderate and low) found in individuals. Situational apprehension refers to the phenomenon of how a person placed in different situations might experience different levels of anxiety about writing. For example, two papers by the same writer in different writing situations � say, in a classroom and in the comfort of home -- could be very different in quality because of situational apprehension.

While their study examined five different situational characteristics, two types are important to this research because they involve instructor feedback. One characteristic is the level of perceived evaluation in the writing situation. When people are in highly evaluative situations, their apprehension will increase. The second characteristic is the perceived conspicuousness of the situation. When people believe their writing will be seen by others, they will feel more anxious than when they know nobody will see it. Privacy or inconspicuousness will make people less anxious. In an experiment involving 399 undergraduates, the researchers found that all five characteristics were statistically significant. They recommended that further research could lead to ways writing instructors can change the classroom environment, including instructions, tasks and projects, to reduce situational anxiety.

Although the Daly, et al. studies formed the groundwork for writing apprehension theory, the findings of studies using the Daly-Miller instrument perhaps should be interpreted with some caution. A 1988 study raised questions about prior research using the Daly-Miller Writing Apprehension Test to gather data (Reed, Burton & Vandett). The inclusion of the "uncertain" response as the 3-point in the Likert scale is unnecessary, they contended, and could lead to faulty placement in writing apprehension groups or inaccurate total writing apprehension scores. Because the instrument measures attitude and not factual knowledge, the "uncertain" response is inappropriate. The main problem is that a person could answer 3 on all 26 items and get a 78 total score, placing them in a medium apprehension group. They recommended that the "uncertain" response be removed and a four-point Likert scale replace the five-point scale. However, it seems very unlikely that a subject would answer "uncertain" to all 26 items, so the Reed, et al. study serves more as a warning to future researchers rather than a condemnation of the Daly and associates work.

While the Daly-Miller Writing Apprehension Test (1975a) evolved from a unidimensional measure to a two-dimensional measure (Daly & Hailey, 1984), later research argued that writing apprehension was multidimensional. Riffe and Stacks (1988) developed a Mass Communication Writing Apprehension Measure (MCWAM) that, following testing on 621 mass communication undergraduates, supported seven factors of writing apprehension: General Affect; Blank Page Paralysis; Mechanical Skill Competence; Career and Essential Skills; Evaluation Apprehension; Task Avoidance; and Facts Versus Ideas.

Earlier research by Daly and Miller (1975b) indicated that males experienced writing apprehension more frequently than women, low apprehensive students were more likely to voluntarily take advanced writing courses than were high apprehensives, and high apprehensives reported less success in previous writing courses than did low apprehensives. In their discussion, the researchers speculated that the nature of instructor�s feedback on writing assignments could influence writing apprehension. Riffe and Stacks (1988) found that women had more significantly more positive attitudes toward writing on four out of seven factors of their multidimensional writing apprehension test. Contradicting a previous study (Westmoreland, Starr, Shelton & Pasadeos, 1977) that found males more "suited" for newswriting, Riffe and Stacks found that females enjoyed writing more, were more confident in mechanical skills and were less likely to procrastinate on writing assignments. Journalism majors generally tended to be less apprehensive than broadcasting, public relations and advertising majors. Freshmen were less apprehensive than sophomores in mechanical skills and blank page paralysis.

A subsequent study to validate and refine the MCWAM (Riffe & Stacks, 1992) added an eighth dimension � Audience Salience -- in an attempt to differentiate mass communication majors from non-majors. The study reaffirmed previous results, and also found that mass communication majors� attitudes were significantly different from non-majors� attitudes on Career Skills, Audience Salience and Facts Versus Ideas. Surprisingly, non-majors believed that "writing facts is easier" more than majors did. They concluded that, although further research is necessary to validate their measure, the MCWAM could help instructors identify high-apprehensive students and then develop instructional strategies to help them overcome their writing blocks.

Rechtien and Dizinno (1998) administered the MCWAM first to 419 incoming freshmen and later to 149 students who were at least second semester sophomores. The second sample included 37 students who had taken the MCWAM as freshmen. Comparing the mean scores of those 37 from the first test to the second, they found significant changes on four components (mechanical skill competence, facts vs. ideas, evaluation apprehension, and a component added by the researchers � locus of control), indicating less apprehension. They also found a significant change in task avoidance, indicating an increase in apprehension because students were procrastinating longer. However, because there was no treatment, they concluded that the changes could have been due to maturation effects. The typical college student might be expected to have less writing apprehension as he or she takes more writing classes, becoming more self-confident and skilled. On the other hand, high-apprehensive students might experience greater apprehension over time because of a history of failure in previous writing courses.

Because the sample for this dissertation study may have included English as a Second Language students, it is interesting to note an experiment (Phinney, 1991) in which some writing students used computers in class while others wrote on paper with pen. Phinney found writing apprehension actually increased for English-language students during the semester while there was no change for ESL students. The findings led her to conclude that computer use alone does not reduce writing apprehension.

A case study (Selfe, 1985) of a high-apprehensive student, Bev, demonstrated how Bev�s apprehension affected her writing. Bev procrastinated on writing assignments and normally didn�t start work on them until the day before they were due. Then she rushed through a first draft, using no prewriting strategies such as outlining or diagramming her ideas. She was more concerned with mechanical errors than with the overall structural soundness of her compositions. In her revisions, she also focused on surface repairs, and often just deleted problematic phrases and sentences rather than rewrite them. When discussing her writing, Bev frequently used such words as "fear" and "hate." She blamed her apprehension on a lack of writing instruction and writing assignments during primary and secondary schooling.

There is some evidence that the type of academic writing can affect levels of apprehension. Faigley, et al. (1981) found that high-apprehensives were significantly different from low-apprehensives in their writing performance on personal narrative essays, but not on argumentative essays. Walsh (1989) stated that high-apprehensive students enjoyed personal forms of writing, such as letters, more than did low-apprehensives. She suggested that business writing also could be used to alleviate writing apprehension because of its focus on specific business situations, audiences and purposes. Fox (1980) used business writing instruction in a course and determined that it reduced apprehension for students. Using the same reasoning that real-world writing tasks could reduce apprehension, Olson (1990) conducted an experiment in which one section of freshman composition students received a half-semester of news-writing instruction while other sections received traditional instruction. However, he found no significant difference in self-efficacy, motivation, attitudes, grammar and mechanics, and apprehension between the two groups.

Student awareness of writing for an audience is one factor that differentiates journalism from other writing fields (Riffe & Stacks, 1988, 1992). Studies examining the relationship between audience and apprehension have been inconclusive. A study in which students wrote essays aimed at either a familiar or distant audience found no significant relationship between writing apprehension and the quality of writing between the two essay-type groups (Richardson, 1981). A subsequent study, however, found that high-apprehensive students assigned to write for an unspecified audience were less apprehensive than those assigned to write for a specific audience that would respond to their texts (Hurd, 1985). Overall, though, the unspecified audience essay increased apprehension the most for the 179 undergraduate composition students in the experiment. Hurd�s findings partially supported Walsh�s (1986) observation that an apparent cause of writing apprehension is fear of public exposure and embarrassment, of appearing "less than competent" (p. 4) to others. The findings also seem to support Flower's (1989, 1994) theory that writing is a social as well as emotional process.

More than half of the high-apprehensives in Hurd�s 1985 study reported that they considered the teacher as their audience, which again raises the question of how does instructor feedback relate to student apprehension. In Walsh�s (1989) survey of 202 university students, 54 percent of high-apprehensives reported that critical evaluations from the teacher were the main factor that prevented them from writing. She concluded that while instructors may see their evaluations as constructive and useful, apprehensive students view all but positive comments as discouraging (1986, 1989).

While the literature makes it clear that writing apprehension is a problem for many students, less research has been conducted on how to reduce the apprehension. Test anxiety research has examined biofeedback, relaxation therapy, hypnosis, supportive counseling (Sapp, 1996), and even virtual reality as treatments (Knox, Schacht & Turner, 1993). There is "surprisingly" little research involving cognitive-behavior therapy to treat writing apprehension (Salovey & Haar, 1990, p. 514). They noted that the most rigorous writing studies have examined only the writing process � "how people write rather than what they write" (p. 514) rather than cognitive-behavior treatments. They conducted an experiment that found stress inoculation training combined with writing process instruction significantly reduced writing apprehension and improved writing performance. A treatment group that received only writing process instruction also experienced a significant reduction in apprehension, but no improvement in performance. The study involved 51 undergraduates who responded to a newspaper advertisement seeking people who avoid writing.

Most research into treatment mainly has focused on classroom remedies. Such educational innovations as student-centered workshops, learning-centered writing, journal writing, positive evaluation, increased special attention, and alternative instructional structures have been tried as treatments in studies (Daly, 1985). In his review of the literature concerning instructional methods to treat writing apprehension, Smith (1984) listed peer-group evaluation, clear grading criteria and language study as remedies. Based on his review, Smith found there was a consensus that instructors not grade every assignment nor mark every error. An informal classroom study (Tighe, 1987) successfully treated apprehension by discussion groups, writing workshops, student journals, peer responses, clear evaluation criteria and basing all writing assignments on student experiences. An experiment by Fox (1980) both contradicts and supports parts of the Smith and Tighe studies. His treatment group used group interaction, peer response and student-teacher conferences as instructional methods. The control group received traditional writing instruction, consisting of lectures, discussion, teacher evaluation and writing exercises. Both groups had significant reductions in writing apprehension, but the treatment group�s apprehension was reduced more quickly. Apprehension was reduced even though both groups had compulsory writing assignments, contradicting Smith�s (1984) conclusion that compulsory writing causes apprehension. Since apprehension declined for both groups, Fox suspected that a remedy could involve simply increasing writing opportunities for students, who may not have written often in the past. For these students, just thinking about having to write may create more apprehension than the actual act of writing.

The causes of writing apprehension are also unknown. One problem is that most research designs in writing apprehension studies have been correlational, which only indicate the degree and direction of relationships between variables. Several researchers (Daly, 1985; Hillocks, 1986; Selfe, 1985) raised the issue of whether writing apprehension is the cause of or is caused by writing difficulties. Many of the studies do, indeed, suggest there may be a reciprocal effect. If true, that could mean many apprehensive writers could be caught in a vicious cycle of failure and anxiety. For instance, a student with poor writing skills may become so discouraged by failed attempts to improve his or her writing that apprehension over future writing assignments may develop. Even Daly (1985) eventually suspected an interaction between apprehension and writing problems, concluding it was equally plausible that the apprehension may cause writing deficiencies or the deficiencies might cause the apprehension. In his review of the literature, Smith (1984) summarized the causes, including compulsory writing, a history of adverse responses, and the pressure of perfectionism. The consensus of researchers, he concluded, is that evaluation is a major cause of apprehension. Again, research on test anxiety may be useful in this review since it has identified two main causes (Mealy & Host, 1992). First, students experiencing high test anxiety have inadequate learning or study skills. Selfe�s 1985 case study of a high-apprehensive writer who had insufficient writing instruction in grade school lends anecdotal evidence to that conclusion. Second, highly anxious students have habitually negative thoughts during tests, creating an attitude of learned helplessness. Again, the high-apprehensive student who talked of her "fear" and "hate" of writing serves as one example of that possible cause.

Reflecting on his decade of writing-apprehensive research, Daly (1985) argued that "comparison deficiency," when a writer compares his or her intentions with the actual product, could be a cause of the apprehension. An apprehensive writer often finds the product deficient compared to what he or she wants it to be. Non-apprehensive writers, however, tend to realize that the written word is often different from the idea he or she originally conceived or intended to write. They also have discovered that writing takes a lot of effort to reach their intentions. Daly�s comparison deficiency idea seems markedly similar to the McKeachie, Pintrich, Lin, and Smith (1986) expectancy-value motivation model. A writer�s comparison of intentions to the written product seem related to students� expectations about the outcomes of their efforts and to students� beliefs about their competency to accomplish the learning task.

Writing apprehension is negatively correlated with related self-concept, self-esteem and self-competence measures (McCroskey, Daly, Richmond & Falcione, 1977) that contribute to items in the writing self-efficacy instruments to be used in this study. Riffe and Stacks (1992) commented that fear of writing obviously is related to enjoyment of writing. Daly and Wilson (1983) found writing apprehension was significantly and negatively related to self-esteem.

What inspires someone to learn something? Conversely, what can cause students to lose their desire for learning? Much education research has involved the motivation of learners, and many theories have been developed about the nature of motivation.

Primary sources for this section of the literature review are the work of Bandura (1986), Vroom (1964, 1970), and Maslow (1943, 1970) because their theories seem to help explain how journalists are motivated to write. Searches were conducted in ERIC, CARL, Dissertation Abstracts, and PsycINFO databases using the descriptors motivation, academic motivation, expectancy-value, self-efficacy, higher education, writing, feedback, writing evaluation, writing research, writing skills and writing instruction.

Motivation is such a broad and encompassing topic that the American Psychological Association nearly replaced it as a search term in Psychological Abstracts (Walker & Symons, 1997). Motivation should be of vital interest to adult educators because it enhances learning and achievement (Wlodkowski, 1985). The instructor�s own enthusiasm and values have much to do with producing motivated students. "Simply feeling that you care about their learning is an important motivator for student learning" (McKeachie, 1999, p. 308). Thus, instructor feedback of any sort and in any form may motivate student writers by showing them that the instructor cares about their writing and wants them to improve. McKeachie recommended that feedback provide encouragement and guidance for students doing poorly. For better writers, feedback should give students new challenges to further motivate them.

A few studies have specifically examined the relationship between motivation and writing instruction. In their two-year study of high-school writing instruction, Hill and Boone (1982) speculated that if Maslow (1943, 1970) had taught writing, he would have identified students� needs and goals. Maslow�s motivational theory incorporated need reduction. Man must satisfy his lower-level, largely biological needs before he can achieve higher order needs and, ultimately, self-actualization. Maslow believed that as lower level physiological needs are met, higher order needs will emerge. His hierarchy of needs, in order, is physiological, safety, love, esteem and self-actualization. While not among the five basic needs, cognitive needs are also addressed by Maslow, who said freedom of inquiry and expression -- the desires to know and to understand -- must be met if a person is to satisfy the basic needs. Hill and Boone (1982) believed that Maslow would have written assignments so they met students� needs, offered students choices in instructional method, and evaluated them in different ways to help them advance to higher levels on his hierarchy of needs. Their study suggested that students with safety-level needs are likely to suffer from writing apprehension. Marking every error in their writing will probably increase their apprehension, so evaluation methods that promote rather than inhibit growth are necessary. Since many writing classes are now held in computer laboratories, often with software that corrects mechanical errors in grammar and spelling, a more content-oriented approach by instructors may not only increase motivation but reduce apprehension. However, they noted that the difficulty for the writing instructor is identifying at which Maslowian stage of development students have reached.

Writing apprehension has been found to influence students� expectations concerning instructor evaluation of their writing. For instance, students with high levels of writing apprehension expect negative evaluations on their stories because they have a history of negative feedback (Daly, 1977). According to self-worth motivation theory (Covington, 1993), the danger of students expecting negative evaluation is that they may accept failure and simply stop trying to improve their work. Self-worth theory states that some students won�t try because they are conditioned to accept failure; in essence, they have no expectations of success even if they try.

Certain levels of anxiety and stress are unavoidable in any classroom situation, not just the writing class. As Vroom (1964) noted, certain levels of motivation enhance performance but too much motivation causes performance to decline. So, too, can anxiety. The stress associated with anxiety can lead to either improved performance by students or it can detract from learning. "Stress can be intense enough to keep the adult student from ever succeeding, or it can be intense enough to motivate exceptional academic accomplishments" (McClary, 1990, p. 66), depending upon the student�s ability to cope with stressors. She suggested that stress management and coping strategies be part of the curriculum. Bandura (1986) theorized that people able to cope with stressful situations will be motivated by the pressure to expend more effort and improve their performance.

Expectancy-value theory helps explain motivation for this study because it focuses on the interactions of individual characteristics, such as writing apprehension, and environmental effects on behavior, such as feedback. Expectancy-value theory basically states that students will expect to succeed if they try, and also will expect rewarding outcomes, such as good grades and/or positive feedback, if they try. Expectancy-value models are widely used in studying learning and motivation in educational settings (Keller, 1983). In expectancy-value theory, effort is the major measurable motivational outcome (Small, 1997). For effort to occur, the person must value the task and must believe he or she can succeed at the task. Therefore, in an educational setting, the learning task needs to be presented in a way that is engaging and meaningful to the student, and in a way that promotes positive expectations for the successful achievement of learning objectives.

Vroom (1970) based his expectancy-value theory upon an industry and management perspective. It included many aspects of previous expectancy theories. He listed four variables that determine a worker�s attitude toward his role in the workplace and whether the worker will leave the job: 1). Amount of pay, status, acceptance and influence, which Vroom called outcomes; 2). Strength of the person�s desire for or aversion to the outcomes; 3). Amounts of the outcomes that the worker thinks other workers are receiving; and 4). Amounts of the outcomes that the worker expected to receive. For example, as the amount of the desired outcomes increases, the worker�s motivation becomes stronger, resulting in increased effort, Vroom argued.

Considering Vroom�s ideas from a student�s perspective, outcomes could be grades, feedback from the instructor, or reaction from fellow students. In a journalism course, a student who highly desires and expects to receive positive instructor feedback and good grades would be more motivated to write successfully. Conversely, if students expected to receive better grades or feedback than they actually received, they might be less motivated to write.

While Vroom�s theory is aimed at predicting employee motivation in the workplace, educational researchers have adapted his ideas to the classroom, arguing that the interaction of expectancy and value accurately explains how student motivation is produced (Pintrich, 1988). An education expectancy-value model (McKeachie, et al., 1986) contains two paths. Similar to self-efficacy theories, the expectancy path includes a student�s perceived competence, expectancies, level of test anxiety, perception of the learning task�s difficulty, and beliefs about his or her efficacy, control and outcomes. For instance, a student who highly expects to write an "A" paper will spend more time and effort on the paper than will a student with a low expectation of success.

The value path includes three different types of values. Attainment value represents the student�s perception of the probability that an assignment will be challenging. Interest value reflects a student�s intrinsic interest in an academic area; for example, a student enjoys journalism classes. Utility value considers the student�s perception of how useful or practical a learning task will be toward meeting a specific goal, such as graduating from college. The model�s value path also includes the student�s long-term goals, such as career, and short-term goals, such as passing a class or test. Because journalism is a field in which many students have career goals of working in the media, the value path of the McKeachie, et al. model seems particularly relevant to help explain student motivation for this study.

The implications of the education expectancy-value model for teaching include providing feedback to students that stresses the positive aspects of their work and points out what needs to be done to improve it (Lucas, 1990). Instructors also should connect assignments to students� goals, demonstrating the value of the instruction. According to Walker and Symons (1997), in expectancy theory, motivation to perform an activity is determined by effort-to-performance expectations (you expect to succeed if you try) and performance-to-outcome expectancy (you expect rewarding outcomes if you try).

Keller�s (1983) ARCS model is based upon Vroom�s expectancy-value theory and seems applicable to most writing instruction because of its emphasis on feedback from the instructor to the learner. In the ARCS model, instructor�s feedback creates expectations for success in the learner, and also provides learner satisfaction with the instruction. The ARCS model of motivational design incorporates attention, relevance, confidence and satisfaction strategies into instruction. Attention strategies, such as surprise, are intended to arouse and sustain curiosity and interest. Relevance strategies, such as matching learning objectives to students� needs, should link to students� needs, interests and motives. Confidence strategies include the instructor providing written feedback on the quality of students� performance. These strategies are designed to help students develop a positive expectation for success. Satisfaction strategies provide extrinsic rewards, such as certificates, and intrinsic reinforcement, which encourages enjoyment of the learning experience. Keller urged that feedback be provided to the learner during all four stages of his model. In the journalism classroom, the ARCS model could be employed during every part of the writing process.

Drawing upon the ARCS model, Bohlin, Milheim and Viechnicki (1990) added that learner satisfaction can be provided through informative and corrective feedback. While teachers cannot control the learner�s expectations for success, they can control instructional methods and strategies to promote motivation. That suggests that, even though a high-apprehension student may have expectations of failure, appropriate instructor feedback might also reduce writing apprehension.

Expectancy-value theory is closely related to self-efficacy theory (Bandura, 1986), which deals more with intrinsic rewards. Bandura�s self-efficacy theory of motivation differs from expectancy-value theory in that it emphasizes students� beliefs concerning their capabilities to learn in order to attain a valued outcome. Expectancy-value theory places more emphasis on the value of the outcome or goal to the learner (Schunk, 1991). For instance, a student will judge the likelihood of attaining a goal, but will not be motivated to attempt what he or she sees as an impossible goal. Bandura contended that the expected outcomes of an action depend largely on a person�s judgment of his or her self-efficacy. In other words, perceived rewards or punishments result from how competent people believe themselves to be in performing an act. If a student believes he or she is very capable of covering a news event, they probably expect the outcome will be praise from the instructor and a high grade. A student who doubts his or her ability to cover the event likely expects lesser rewards or even penalties. The doubtful student may not try as hard to interview people or pay attention to detail at the news event. Bandura argued that people�s self-efficacy judgments will predict how they act better than will the anticipated outcomes.

Research comparing self-efficacy theory to expectancy-value theory has indicated that self-efficacy beliefs may better predict writing performance than does outcomes expectancy. Factor analysis by Shell, et al. (1989) found that while self-efficacy was a significant predictor of writing performance (p < .01), outcomes expectations was not significant, supporting Bandura�s theory. However, the sample consisted of education majors, who may be different from the mass communication student sample for this experiment. Future teachers may not consider writing skills to be as important for their careers as do mass communication majors. In addition, there was a large gender inequity in the sample (38 males, 115 females). Therefore, the Shell, et al. findings may not be generalizable to students in other majors. Pajares and Johnson (1993) also found no correlation between students' writing self-efficacy and their outcomes expectations. Students' perceived usefulness of writing was unrelated to their writing confidence. Outcomes expectations were significantly correlated with essay scores but were not predictive of them in a regression model. Results of a multiple regression analysis suggested that the relationship between outcomes expectations and performance is mediated by writing self-efficacy, a result supportive of social cognitive theory.

Other related motivation theories help explain the expectancy portion of Vroom�s theory. In their review of academic motivation theories, Cross and Steadman (1996) concluded that most theories "involve the premise that a lack of self-confidence leads to an unwillingness to try" (p. 79). Most cognitive theories of motivation "hold that if people do not believe they can do something or learn something, they are unwilling to take the risk that trying and failing will pose to their self-esteem" (p. 79). Schwartz and Lanyon (1969) outlined two basic views of motivation. The need reduction view says that a person is motivated to act in order to fulfill such biological needs as eating, drinking, avoiding pain, and minimizing anxiety. The positive striving views says that people are innately motivated to grow and develop their abilities. A person�s belief about his ability to succeed at a task is based on his perception of why he has succeeded in the past, according to attribution theory (Weiner, 1985). Covington�s (1993) self-worth model theorizes that students will deal with the threat of failure in achieving academic goals in different ways. Some enjoy learning for the sake of learning. Others are high achievers in order to prove their abilities to themselves and others. Some students struggle to avoid failure by avoiding risky challenges. A fourth group accepts failure and gives up without trying.

Many adult education researchers who examined motivation agree that there must be a need to learn � that learning have value � but that adults differ in the types of needs they have to learn. Knowles (1970) defined education motivation as "the experiencing of self-induced dissatisfaction with present inadequacies, coupled with the clear sense of direction for self-improvement" (p. 42). An adult will be motivated to learn things that he or she sees the need to learn, Knowles added, identifying six basic needs, or motivating forces for behavior. They are physical, growth, security, new experience, affection and recognition needs. In case studies of 22 urban, adult learners, Houle (1961) identified three types of adult learners by their reasons -- or motivations -- to participate in learning. They are goal-oriented, activity-oriented, and learning-oriented reasons. However, there are three elements that must be in place before an adult will undertake a learning project: She or he must recognize a need for learning, have the will to do something about it, and have the opportunity to do so. The will to do something is a drive or motivation. Other factors in adult learning identified by Houle also are related to motivation: a successful initial learning experience, the chance to continue, and skillful guidance by the instructor. In two-hour interviews with 35 adults, Tough (1979) concluded that "anticipated benefits constitute a significant portion of the person�s total motivation for learning" (p. 45). He identified three benefits that motivate a person to learn � for self-esteem, for the pleasure of learning, and to please or impress others. However, all three benefits depend upon successful or at least improved performance by the learner after instruction.

Researchers generally agree there is an negative relationship between writing anxiety and motivation. As anxiety increases, a person�s perceptions of their ability to succeed in a task decreases. McCroskey, et al. (1977) conducted five studies of the relationship between communication anxiety and self-esteem. The Daly-Miller instrument (1975a) was one measurement of communication anxiety. They found significant relationships ranging from -.52 to -.72 in the five studies. Two samples consisted of college students, two samples of public school teachers, and one sample of federal employees, which suggested that the relationship exists across age groups and occupations.

According to social-cognitive theory, people's self-efficacy beliefs in their skills and ability to complete tasks influence their motivation. The expected outcomes or rewards of succeeding in the task can also serve as motivators, according to expectancy-value motivation theory. Students' self-efficacy beliefs in their writing skills and ability to perform different writing tasks may influence their motivation to write. Journalism students who plan to pursue careers in the mass media should have high outcomes expectations that motivate their writing efforts.

Feedback can affect both self-efficacy and writing apprehension in students, research suggests. Writing apprehension may cause students to ignore instructor comments on their assignments, to avoid classes that require writing, and to pursue careers in which no writing is involved. Writing apprehension seems to be more common in poor writers than in skilled writers, although it is possible that poor writing skills may lead to writing apprehension.

Feedback can help students learn and improve their writing in subsequent drafts of papers, researchers generally agree. However, students often seem to ignore, misunderstand or misinterpret teacher comments. Students with high levels of writing apprehension tend to use instructor comments on local matters to quickly fix mechanical writing problems, although some researchers have concluded that all students generally address local problems more than global or content issues when revising first drafts. Top

Because the sample for this study possessed the attribute variables of self efficacy, writing apprehension, and writing outcomes expectations, a causal-comparative research design was appropriate, according to Ary, Jacobs, and Razavieh (1996). The causal-comparative design examined whether the subjects who differ on self-efficacy, writing apprehension, and writing outcomes expectations also differed on their use of global or local instructor feedback on first drafts of news stories. The causal-comparative design allowed the researcher to examine the relationships among independent and dependent variables. Causal-comparative studies are a useful method for much educational research in which the populations of interest are already different on one or several characteristics, according to Ary, et al.. The design "can supply much information of value in educational decision making," they stated (p. 366).

The following research questions were considered in the study:

1. What are the relationships among students' writing apprehension, writing task self-efficacy, writing skills self-efficacy, and writing outcomes expectations?
2. What are the effects of writing apprehension, writing outcomes expectations, and writing self-efficacy beliefs on students' use of global, local or both kinds of feedback from instructors on first drafts?

An availability sample of about 200 undergraduate mass communication students was utilized for the study. Generalizability is limited by the use of such a sample, but the sample size was large enough to avoid some of the generalizability problems caused by small sample size (n = 30) in the 1993 Pajares and Johnson study of self-efficacy, outcomes expectations, and writing apprehension. Although they stated that correlational studies do not require extremely large samples, Ary, et al. (1996) did not recommend samples with fewer than 30 participants. Because it was a convenience sample, the characteristics of the sample were described in detail in Chapter Four so other researchers can infer the population that it represented. To gather this information, questionnaires were distributed to collect biographic and demographic data on the subjects, including age, major, gender and ethnicity (Appendix A). The students were enrolled in mass communication courses at a land-grant university in the Rocky Mountain region. Students were invited to volunteer for the study. The sample was reasonably homogenous, as recommended by Gall, Borg, and Gall, (1996) to ensure that relationships between variables are not obscured by participants who vary widely from each other. The students primarily were majors in three areas � journalism, broadcasting, and communication � offered by the communication and mass media department. There were also students majoring outside of the department or who had not declared a major field of study yet.

A causal-comparative research design was chosen because the study will explore possible cause-and-effect relationships among variables involving personal characteristics and behaviors that are present in some students and absent or present in lesser degrees in other students. For instance, the causal comparative design is appropriate for studying the causes of feedback use (a behavior) or the effects of writing apprehension (a personal characteristic). Gall, et al. noted that causal-comparative studies are appropriate for many educational situations because researchers may not be able to use an experimental treatment or random assignment. In this study, for example, it was not feasible to randomly assign students to low self-efficacy or high self-efficacy groups. Nor would it be ethical to induce writing apprehension in a treatment group of students. Students are either low or high in self-efficacy, outcomes, or apprehension, making causal-comparative the appropriate research design for this study.

To reduce the possibility of non-volunteers and thereby avoid sampling bias, the researcher followed the recommendations of Rosenthal and Rosnow (1975) to improve the rate of volunteering. Those included making the appeal for volunteers as interesting and non-threatening as possible, emphasizing the theoretical and practical importance of the study, and stressing that by volunteering, participants could help others. Instead of the researcher soliciting volunteers, a graduate student closer in age to the sample introduced the study and passed out participant consent forms (See Appendix B), again to encourage participation. Participants were also informed that they could obtain their results on the instruments, and were provided contact information to request their results.

All classes involved in the study were held at the same campus, during the same semester, and under the same climatic and barometric pressure changes that occurred during the semester to ensure the study had ecological validity.

Although their influential work on validity involved experimental designs, the concepts of Bracht and Glass (1968) are applicable in other types of quantitative research (Gall, et al., 1996). Bracht and Glass stated that the operational definition of each dependent variable is relevant to the generalization of the study�s results. In this study the operational definition of the dependent or criterion variable -- student use of instructor feedback -- was the score on a questionnaire (Appendix C) asking students to rate their use of local or global teacher comments. Feedback use was measured by giving students a story that had been edited by an instructor. The story included five comments for local issues and five comments addressing global issues.

Lauer and Asher (1988) noted that researchers often do not have existing instruments to use to measure the aspects of writing they wish to study, and must develop new instruments for their investigations. The feedback instrument was based upon Straub's 1997 questionnaire, which asked students to rate their preferences for teacher comments written on an essay. Participants were asked to assume that they were getting a rough draft back that they had written and were going to revise. They were told, "You know that the paper is rough and that you're going to need to do substantial revision. What kind of comments would you prefer?" The instrument consisted of the essay with teacher comments written upon it, and a questionnaire in which participants rated each comment. The four choices were 1 - definitely prefer, 2 - prefer, 3 - do not prefer, and 4 - definitely do not prefer. Because Straub was studying student preference for each comment, the average score was calculated for each item. He did not report the instrument's validity or reliability.

The instrument was modified for this study to reflect the type of writing mass communication students do -- journalistic writing. A short news story from The Copy Editor's Handbook (Fellow & Clanin, 1998) was used with 10 teacher comments written on it. Five comments dealt with local issues and five involved global matters. A questionnaire with the rating choices on it listed the 10 comments. Because this study examined the general question of whether students would use local or global comments, or both, when revising their first drafts of a story, the four choices to rate each item were reworded: 1 - definitely will use, 2 - might use, 3 - probably won't use, and 4 - definitely won't use. The average scores for the five global comment use items and for the five local comment use items were each calculated. Low scores indicated that students said they would use the instructor comments when revising their first drafts.

Before it was used in the study, the instrument was first presented to students who would be representative of the sample population. They were asked singly (n = 3) and in a small group (n = 18) to read the instrument and discuss whether any parts in the instructions or in the teacher comments were confusing or needed to be rewritten. As a result, the instructions were revised to further clarify that participants were being asked to rate the instructor comments, not the story itself. Next, six content experts were asked to examine the instrument for content validity. Two were English professors who teach freshman composition. One was the director of the university's writing center. Three were journalism professors who have taught news writing courses. They were asked to determine whether the teacher comments addressed global and local issues. As a result of the content experts' suggestions, all 10 of the teacher comments were rephrased so there were fewer directives and more questions. The rewording seemed appropriate for two reasons. Directive comments, such as "check spelling," might have resulted in skewed responses from students, who probably would have answered definitely would use or might use in response to comments that essentially were commands by the instructor. Comments phrased as questions, such as "did you verify name spelling," would give students more choice over their response. The rephrased comments also conformed with the social-cognitive theoretical basis of the study. Straub and Lunsford (1995) noted that composition instructors increasingly are concerned about taking control over student writing. The way teacher comments are phrased can result in different degrees of control. Directive comments are the most controlling because they essentially command students to revise, while question comments are less controlling in that they indirectly suggest that students revise. Furthermore, question comments such as "should you call them cops" can lead to a dialogue between instructor and student over the appropriateness of word choice, a dialogue that Flower (1994) would call negotiating meaning of a text.

A pilot study of the instrument was conducted with students enrolled in a community journalism internship (n = 20) to determine its validity and reliability. A Cronbach alpha of .76 was calculated, which was considered an acceptable degree of reliability. According to Ary, et al. (1996), reliability coefficients as low as .50 or .60 may be acceptable for research purposes.

Item analysis was conducted to check the internal consistency of the instrument. Each item score was correlated with the total-scale scores. Six items had significant correlations ranging from r = .595 to r = .795. Four were local comments and two were global. Non-significant correlations on the other four items ranged from r = .207 to r = .409. The non-significant items were re-examined to see if the wording was consistent with other comments or whether they should be rephrased. As a result, the two items with the lowest correlations (r = .207, r = .224) were rephrased as questions before the instrument was used in the study. One was global and the other was local. Ary, et al. (1996) recommended that items should correlate at least .25 with the total scores, and that items with lower correlations should be examined to see if they are ambiguous. If so, revision may make those items usable.

Bracht and Glass (1968) recommended describing the independent variables explicitly so other researchers could replicate the study and so readers can estimate how much the findings can be generalized to other settings. In this study, three writing characteristics -- writing apprehension, writing self-efficacy beliefs, and writing outcomes expectations -- were the independent variables. The operational definition of each independent variable will be described as follows:

Two instruments were considered to measure writing apprehension. The Mass Communication Writing Apprehension Measure (MCWAM) is a multi-dimensional instrument and measures student attitudes toward eight situational and dispositional dimensions of writing apprehension: It is a 43-item instrument employing a five-point Likert scale with responses ranging from strongly disagree to strongly agree. Riffe and Stacks (1992) reported that reliability is .931, as measured by the Kaiser-Mayer-Olkin measure of sampling adequacy (MSA), a reliability measure that helps establish the strength of relationships. MSA scores in the .90s are "marvelous" (Norusis, 1985). For construct validity, all 43 items in the instrument correlate at least moderately (r > .30) with at least two other items. The MCWAM has been employed by the authors in two studies (Riffe and Stacks, 1988, 1992), but they concede that the instrument needs a longer program of validation research before it can be used to screen students for their level of writing apprehension.

While the Riffe-Stacks instrument seemed suitable to this study because it is intended for measuring writing apprehension for the sample of mass communication students, the Daly-Miller Writing Apprehension Test (1975a) has undergone a longer program of validation and reliability testing (Appendix D). The authors reported their 26-item instrument is reliable and valid. Reliability has been measured with the Cronbach alpha coefficient ranging from .89 (Daly, 1979) to .94 (Daly & Miller, 1975a). The WAT has been widely used by the authors and other researchers, and has been found to be highly valid. Factor analysis showed positive and above .60 loadings between each factor and at least two items. In a 1988 examination of its reliability, Reed, Burton, and Vandett found the instrument reliable but suggested that the 5-point Likert scale be reduced to four points by removing the uncertain response. That adjustment was not made for this study because Reed, et al. did not report reliability or validity measures for the revised instrument. Scoring of the test followed Daly and Miller's formula: Writing apprehension = 78 + Positive Scores - Negative Scores. Scores may range from a low of 26 to a high of 130 with high scores indicating low levels of writing apprehension.

Two possible instruments were identified to measure writing motivation. In his research concerning expectancy-value motivation, Pintrich and DeGroot (1990) used the Motivated Strategies for Learning Questionnaire (MSLQ), a self-report instrument that includes two subscales. The self-efficacy subscale to measure student expectations had a reliability score of .89, assessed with Cronbach�s alpha, and the intrinsic value subscale�s reliability was .87.

However, the researcher decided to use another instrument (Shell, et al., 1989) to measure student motivation in this study because: 1.) its reliability scores were higher than the MSLQ�s reliability; 2.) the writing outcomes expectations instrument measured both extrinsic and intrinsic interests of students; 3.) it measured writing motivation rather than a general academic motivation construct. The instrument (Appendix E) used two writing self-efficacy subscales developed by the researchers based upon Bandura�s self-efficacy theory, plus a third writing outcomes instrument that seemed appropriate to measure the value-component of Vroom's expectancy-value theory.

Furthermore, Bandura (1986), Shell, et al. (1989), and Pajares and Johnson (1993) argued that self-efficacy is task- and context-specific. That is, a general measure of self-efficacy, such as the MSLQ, cannot reliably be used to measure a student�s self-efficacy in unrelated subjects such as mathematics and writing. Bandura pointed out that an individual�s perceived self-efficacy will differ in level according to the difficulty of a learning domain. For instance, a person may feel competent in addition but not at algebra. Self-efficacy also will differ according to generality. A person may feel competent in only one learning domain or in several, but probably not all. Self-efficacy will vary according to strength. A person might feel highly competent in English, for instance, but only moderately competent in math. Because of the individual variability, specific measures of self-efficacy will have more "explanatory and predictive power" (p. 397) than will global measures.

Partially confirming this theory, Pajares and Johnson found that general self-efficacy was significantly related with writing tasks and writing outcomes expectations, but not with writing skills. However, in related subjects such as reading and writing, self-efficacy and outcomes expectations beliefs have been shown to be generalizable across the domains (Shell, et al., 1989). The domain-specificity of the Shell, et al. writing self-efficacy instrument was another reason for its use in this study (Appendix E).

Two subscales measuring writing tasks and skills self-efficacy and one instrument for writing outcomes expectations were administered. The writing outcomes expectations instrument measured students� attitudes toward such short-term and long-term goals as employment, social activities, family life, education and citizenship.

The first, eight-item subscale measured students' confidence that they possessed various composition, grammar, usage, and mechanical skills. The second subscale of 20 items measured students� beliefs of their ability to write different writing tasks. Participants chose a score from 0 (no chance) to 100 (completely certain) to rate their self-efficacy for each skill or task. Shell, et al. reported reliability scores of .95 (Cronbach alpha) for the writing skills self-efficacy subscale and .92 for the writing tasks self-efficacy subscale. Factor analysis showed positive and above .40 correlations between items and subscale scores. They reported that the correlations indicated that all items discriminated well among subjects and were satisfactory for retention in the measures for final analysis. Self-efficacy scores were calculated by computing the mean for the items in each subscale.

The writing outcomes expectations instrument measured students� judgments of the importance of writing for successfully accomplishing various academic and life goals. The instrument consisted of 20 items using a 7-point Likert scale ranging from extremely unimportant to extremely important. Shell, et al. (1989) reported reliability scores of .93 and positive and above .40 correlations for all items on the instrument. Again, they reported that the correlations indicated that all items discriminated well among subjects and were satisfactory for retention in the measures for final analysis. Writing outcomes scores were calculated by averaging the ratings across all 20 items.

As with all self-report instruments, there is the potential for respondents to fake their responses. However, to reduce the chances of falsified responses, each participant was assured that he or she would remain anonymous (Gall, et al., 1996). They also were assigned code numbers based upon the last four digits of their Social Security numbers. Participants also were assured that under no circumstances would their test data be revealed to anyone other than the researcher.

Following data collection, data were first analyzed to provide descriptive statistics, including mean scores on the apprehension and motivation scales, freshman composition grade point averages, ages, and use of global and local feedback.

Correlations were calculated between independent variables to determine which were statistically significant, and the strength and direction of those relationships. Because the data were measured on an interval scale, the Pearson product moment coefficient of correlation was used to express the correlations.

Finally, the data were analyzed with one-way, between-subjects Analyses of Variance to evaluate the relationships among writing self-efficacy, writing apprehension, writing outcomes and student use of instructor feedback. The dependent variables were student scores on the instrument measuring use of global or local feedback. The dependent variables were labeled Global and Local. The three independent variables were labeled Efficacy (for writing self-efficacy beliefs), Apprehension (for writing apprehension), and Outcomes (for writing outcomes expectations). Each independent variable had two levels, in which participants were placed in low or high groups for each of the three constructs serving as factors in the ANOVA. In the Efficacy and Outcomes factors, participants who scored below the mean were placed in the low categories for self-efficacy and outcomes expectations, while those who scored above the mean were grouped in the high categories. However, participants in the Apprehension factor who scored below the mean were categorized as high writing apprehension, while those scoring above the mean were low apprehension.

A causal-comparative research design was appropriate for this study because participants already possessed high or low levels of writing self-efficacy beliefs, writing apprehension, and writing outcomes expectations. Undergraduate students enrolled in mass communication classes participated in the study. They were administered instruments measuring self-efficacy, writing outcomes expectations, writing apprehension, and use of instructor feedback. Data were analyzed to determine the relationships among the independent variables of self-efficacy, writing apprehension, and outcomes expectations. Data also were analyzed to investigate the effects of self-efficacy, writing outcomes expectations, and writing apprehension on students' use of global or local instructor feedback. Top

Data were analyzed with the SPSS for Windows version 9.0 statistical software. Demographic data concerning the sample were calculated, including frequencies, percentages, means and standard deviations. The mean scores and standard deviations were calculated for the independent and dependent variables. To test the six primary hypotheses, Analyses of Variance and Welch's t� -tests were used. Correlations among the independent variables were calculated to test the two secondary hypotheses. Finally, the reliability of the five instruments used to gather data was calculated and reported with the Cronbach alpha.

The demographic questionnaire and instruments were completed by 188 students in May, 2000. Participants were enrolled in six mass media courses offered by the communication and mass media department: Introduction to Mass Media, Broadcast Production, Investigative Reporting, Graphics of Communication, News Writing and Reporting, and a community journalism internship.

Six surveys were eliminated because participants did not complete one or more pages. Another survey was removed from the study because the participant was a graduate student, and the study was concerned only with undergraduate students. Seven surveys had one incomplete item in the instruments. In those cases, the average score for that item was used. The final sample size for the study was 181, 57.1 percent of whom were females (n = 104) and 42.9 percent of whom were males (n = 78).

The mean age was 21.1 years old (SD = 3.62) with ages ranging from 18 to 57. Caucasians composed the largest ethnic group, 92.3 percent, in the study (n = 167). Two African-American, two Hispanic, one Asian , and four Native American students participated in the study. Five respondents listed "other" under ethnicity.

The majority of students in the study, 52.5 percent, were majoring in the communication and mass media department (n = 95), with 58 listing communication, 27 journalism and 10 broadcasting as their majors. The remaining 47.5 percent (n = 86) majored in other academic areas or were undeclared. Participating in the study were 39 freshmen or 21.5 percent of the sample, 48 sophomores or 26.5 percent, 49 juniors or 27.1 percent, and 45 seniors or 24.9 percent. Most students (n = 128), 70.7 percent, had taken at least two of the required three writing classes at the university. Only 25.4 percent (n = 46) had taken just one required writing course, 88 had taken two, and 40 had taken three. Seven participants did not respond to the question.

The mean grade point average in the required freshman composition course was 3.33 (SD = .68). More than 84 percent of the students (n = 153) reported they received a final grade of A or B. Only 19 students or 10.5 percent reported a C grade, and one student said he or she received a D. Eight respondents did not report a grade.

The mean score on the 100-point writing tasks self-efficacy subscale was 75.91 (SD = 12.16). Scores on the 20-item subscale ranged from 43 to 100. The mean score on the 100-point writing skills self-efficacy subscale was 87.17 (SD = 11.88). Scores on the eight-item subscale ranged from 32.5 to 100. For further data analysis, the 28 items composing the writing skills and writing tasks self-efficacy subscales were averaged into a single Efficacy score. The mean for the full-scale Efficacy scores was 79.13 (SD = 10.53), and scores ranged from 49.29 to 98.93.

The mean score on the seven-point writing outcomes expectations instrument was 4.90 (SD = 1.04). Scores on the 20-item scale ranged from 1.65 to 7.00.

The writing apprehension mean score was 96.02 (SD = 16.39) on a possible scale from 26 to 130. Scores on the 26-item instrument ranged from 50 to 130.

Mean scores for local and global feedback were recoded so high scores indicated that students reported high use of instructor feedback. The global feedback mean score was 3.23 (SD = .46) on a scale from 1.0 to 4.0. Scores ranged from 1.8 to 4.0. The local feedback mean score was 3.53 (SD = .42) on a scale from 1.0 to 4.0. Scores ranged from 2.2 to 4.0.

The primary hypotheses to be tested were:

• Students with high levels of writing apprehension will report they use instructor feedback on local issues when revising a first draft more than students with low levels of writing apprehension do.
• Students with low writing apprehension will report they use feedback on global matters when they revise a rough draft more than students with high writing apprehension do.
• Students with high self-efficacy beliefs will report they use instructor comments on global issues in their revisions more than students with low self-efficacy beliefs do.
• Students with low self-efficacy beliefs will report they use comments on local matters when revising first drafts more than students with high self-efficacy beliefs do.
• Students with high writing outcomes expectations will report they use comments on global matters when revising first drafts more than students with low writing outcomes expectations do.
• Students with low writing outcomes expectations will report they use comments on local matters when revising first drafts more than students with low writing outcomes expectations do.

The means and standard deviations for the Global and Local Feedback scores were calculated for low and high Efficacy groups, low and high Outcomes groups, and low and high Apprehension groups. (See Table 1). Means for all levels of the independent variables were between 1.0 and 1.87, indicating that the sample generally responded that it would definitely use or might use both global and local instructor comments when revising first drafts of the news story included in the feedback instrument. Standard deviations for all levels of the independent variables were less than .5 on the four-point scale used in the instrument.

Table 1

Global Feedback Use

Low High

__________________________________________________

Efficacy 3.13 .43 84 3.32 .48 97

Outcomes 3.15 .44 83 3.30 .48 98

Apprehension 3.33 .46 89 3.13 .45 92

Low High

________________________________________________

Efficacy 3.44 .42 84 3.61 .40 97

Outcomes 3.45 .47 83 3.61 .36 98

Apprehension 3.65 .36 89 3.43 .45 92

Because n�s were unequal for the groups in the three independent variables, Levene�s Tests of Equality of Error Variances were calculated to ensure that the assumption of homogeneity of variance was tenable, and that Analyses of Variance were the appropriate statistical tests. The homogeneity of variance was tenable for four of the ANOVAs. It was not tenable for the ANOVA to test the effects of Apprehension on Local feedback use (p < .05) or for the ANOVA to test the effects of Outcomes on Local feedback use (p < .01). Therefore, the Welch t� -test for unequal n�s and heterogenous variances was used to test the statistical hypotheses for Apprehension and Local and for Outcomes and Local, as recommended by Glass and Hopkins (1996). The Welch t� -tests were one-tailed because the researcher expected high apprehension writers and low writing outcomes expectations writers to use local comments more than low apprehension writers and high writing outcomes expectations writers. The level of significance was set at a .05 alpha for the tests. The effect of Apprehension was statistically significant, t� (179) = 3.675, p < .001, with an effect size of .070. The effect of Outcomes was statistically significant, t� (179) = 2.54, p < .05, with an effect size of .036. In the Outcomes Welch t� -test, the mean for the high writing outcomes expectations group (M = 3.61) was significantly higher than for the low outcomes expectations group (M = 3.45), indicating that high outcomes students reported they will use local instructor comments more than do low outcomes writers. Therefore, the results did not support the hypothesis that students with low writing outcomes expectations would use local feedback more than students with high writing outcomes expectations.

The Apprehension Welch t� -test indicated that low writing apprehension students (M = 3.65) reported they will use local instructor comments more than high writing apprehension students (M = 3.43). The results did not support the hypothesis that high writing apprehension students would use local feedback more than would low apprehension students.

To test the remaining four primary hypotheses regarding the differences between students who use local or global feedback, three one-way ANOVAs were computed for each independent variable (Efficacy, Apprehension, and Outcomes) with global feedback use (Global) as the dependent variable. Then, a one-way ANOVA was calculated for Efficacy as the independent variable with local feedback use (Local) as the dependent variable.

Each of the four ANOVAs was found to be statistically significant at the .05 alpha level of significance. (See Table 2). The ANOVA results supported the three primary hypotheses concerning global feedback use. High self-efficacy students indicated they would use global feedback significantly more than did low self-efficacy writers. Writers with high writing outcomes expectations indicated they would use global feedback significantly more than did students with low writing outcomes expectations. Low writing apprehension students reported they would use global comments significantly more than did high writing apprehension students. However, the hypothesis concerning local feedback use was not supported. Contradicting the hypothesis, high self-efficacy students indicated they would use local feedback significantly more than did low self-efficacy writers.

Table 2

_______________________________________________________

Source df MS F df MS F

_______________________________________________________

Efficacy 1, 179 1.607 7.761** 1, 179 1.325 7.797**

Outcomes 1, 179 .933 4.426*

Apprehension 1, 179 1.930 9.403**

_______________________________________________________

*p < .05. **p < .01.

Effect sizes were calculated to describe the magnitude of the differences between the low and high groups for writing apprehension, writing self-efficacy, and writing outcomes expectations. Effect sizes were found to be small. They ranged from a low of .024 for the Outcomes levels on Global Feedback Use to a high of .070 for the Apprehension levels on Local Feedback Use. The effect sizes indicated that the magnitude of the differences between groups in the three factors was small.

The Relationships Among Writing Apprehension, Writing Tasks Self-efficacy, Writing Skills Self-efficacy, and Writing Outcomes Expectations

• There will be negative relationships between writing apprehension and the three variables of writing tasks self-efficacy, writing skills self-efficacy, and writing outcomes expectations.
• Writing tasks self-efficacy, writing skills self-efficacy, and writing outcomes expectations will be positively correlated with each other.

To test the two secondary hypotheses concerning the relationships among the variables, correlations were calculated to find the directions and strengths of the relationships. Scores are reported using the Pearson r. The scores were then corrected for attenuation because of the restricted range of the data (Hopkins, 1998). Correlations among the variables were found to be positive and statistically significant at a .05 alpha level of significance using two-tailed tests. (See Table 3).

Table 3

Tasks Skills Outcomes Apprehension

1. Tasks--.457**.201*-.487**
2. Skills----.198*-.509**
3. Outcomes-------.261**
4. Apprehension

Note. Correlations are corrected for attenuation. *p < .05. **p < .01.

The correlations supported the secondary hypothesis that writing tasks self-efficacy, writing skills self-efficacy, and writing outcomes expectations would all be positive correlated with one another. The secondary hypothesis that writing apprehension would be negatively correlated with writing tasks self-efficacy, writing skills self-efficacy, and writing outcomes expectations also was supported.

The Cronbach alpha coefficient for reliability was calculated for each instrument used in the study. The feedback instrument had a .64 Cronbach alpha, compared to .76 in the pilot study. Item analysis between each item score and the total score revealed correlations that were all positive and in the low to moderate range, with Pearson r values ranging from .36 to .56.

The reliability of the writing tasks self-efficacy instrument in the study was .88, compared to .92 reported by Shell, et al. (1989). The writing skills self-efficacy subscale used in the study had a .90 Cronbach alpha, compared to .95 reported by Shell, et al. The combined skills and tasks self-efficacy subscales in the study had a .90 Cronbach alpha. Shell, et al. did not report reliability for the combined subscales. Reliability for the writing outcomes instrument in the study was .93, and Shell, et al. reported a .93 Cronbach alpha. The writing apprehension test had a .95 Cronbach alpha, compared to .94 reported by Daly and Miller (1975a).

With global feedback use as the dependent variable, one-way, between-subjects Analyses of Variance indicated that there were statistically significant differences between students with low and high self-efficacy beliefs, low and high writing outcomes expectations, and low and high writing apprehension. With local feedback use as the dependent variable, an ANOVA and Welch t� -tests indicated there were statistically significant differences between students with low and high self-efficacy beliefs, low and high writing outcomes expectations, and low and high writing apprehension. However, the effect sizes of the differences were small.

Scores for writing tasks self-efficacy, writing skills self-efficacy, and writing outcomes expectations were positively correlated with one another. There was a negative relationship between writing apprehension and the other three variables. Correlations were low to moderate in strength, and all correlations were statistically significant at a .05 alpha level of significance.

The implications of the results will be evaluated and interpreted in the following chapter of the dissertation. Top

The statistical tests for differences between low and high students in self-efficacy, writing apprehension and writing outcomes expectations seem to provide evidence that some students use global or local instructor feedback more than do others. The effect sizes indicate that the degree of those differences was small. However, the effect sizes are due to the general lack of variability in the scores on the feedback instrument.

The general direction of the means provides some evidence in support of the hypotheses concerning which students would use global feedback in revising first drafts. In this study, students with low writing apprehension (M = 3.33) report that they would use instructor feedback on global writing issues more than would high-apprehension students (M = 3.13). Students with high writing self-efficacy (M = 3.32) and high writing outcomes expectations (M = 3.30) also report that they would use global comments from instructors more than would low writing self-efficacy (M = 3.13) and low writing outcomes expectations (M = 3.15) students. These general tendencies were anticipated and hypothesized, based both upon the literature and upon the intuition of the researcher, who teaches journalistic-writing courses. Even though the effect sizes of the differences are small, the researcher still suspects there are differences between students with those characteristics and the results of this study do not change those suspicions.

However, the researcher did not anticipate the direction of the means between students in low and high efficacy, outcomes and apprehension groups and whether they use local instructor feedback. The hypotheses regarding these relationships were based upon cognitive learning theory and were not supported by the results. The results suggest that low writing apprehension (M = 3.65), high writing self-efficacy (M = 3.61) and high writing outcomes expectations (M = 3.61) students also may use local comments from instructors more than high writing apprehension (M = 3.43), low writing self-efficacy (M = 3.44), and low writing outcomes expectations (M = 3.45) students will. This unexpected result, although not important according to effect sizes, could provide new directions for future inquiries into student use of instructor comments, as well as for practical applications of instructor feedback in classroom settings.

The results of this investigation lend some support to Straub�s (1997) conclusions that students prefer receiving both local and global feedback from writing instructors. There is a wide gap between what kinds of comments students like to get from their instructors and which comments will actually trigger students to improve their written products on subsequent drafts. However, the high mean scores for feedback use in this study suggest that all students want to receive instructor comments on all writing issues, no matter what the motivation or apprehension levels of the students. According to their mean scores, students in every level of the three independent variables reported that they would definitely or might use both global and local feedback from instructors. Pointing out spelling errors was just as useful as commenting on the relevance of content to the students in this sample, judging by students' scores on the feedback questionnaire.

Of course, even though students self-report that they will definitely or might use instructor comments in revision, many writing researchers have noted that students frequently seem to ignore feedback (McKeachie, 1999; Sommers, 1982; Ziv, 1982) for a variety of reasons. The encouraging student responses on the feedback instrument could be interpreted as another example of journalism students writing for the teacher as their audience, as Pitts (1989) noted. Rather than considering the teacher comments on the story used in the feedback instrument in terms of improving the story for newspaper readers, students may have reported they would use the comments because they perceived the teacher wanted them to use the feedback.

The importance of writing outcomes expectations to students in this study was supportive of the results of a study by Pajares and Johnson (1993), who concluded that students� perceived usefulness of writing was unrelated to their writing confidence.

The results of this study concerning outcomes also seem to agree with Bandura�s social-cognitive theory, and disagree with Vroom�s expectancy-value theory of motivation. Bandura theorized that, although goals play a role in motivation, they are largely a result of a person�s self-efficacy beliefs. Vroom thought outcomes and their perceived value inspired people to greater effort. The mean score on the writing outcomes expectations instrument was 4.90 on a seven-point scale. Although writing outcomes correlated significantly (p < .05) and positively with both writing tasks self-efficacy (r = .201) and writing skills self-efficacy (r = .198), the relationships were weak compared to the relationship between task and skill self-efficacies (r = .457). It would seem that communication and mass media students would be goal-driven and place a higher value on writing for careers in public relations, journalism, and broadcasting. Perhaps students who plan to pursue careers in the media consider other skills, such as public speaking or visual communication, more important than writing to reach their goals.

On the other hand, writing outcomes expectations had a slightly stronger (r = -.241, p < .01) negative correlation with writing apprehension. As students� writing outcomes expectations increased, their level of writing apprehension fell. Conversely, as they placed less importance on writing to meet their goals, their apprehension levels increased. Daly (1978) also found that students with high levels of writing apprehension consider writing to be unrewarding. Although students in this sample did not consider writing to be extremely important to reach their goals, highly apprehensive writers tended to find even less value in writing.

The findings support the conclusions of Daly and Wilson (1983), as well as others (Onwuegbuzie, 1999; Pajares & Johnson, 1993), that writing apprehension is negatively correlated with various self-esteem measures, in this case writing tasks and skills self-efficacy beliefs. As writing apprehension levels rose in this sample, self-efficacy levels went down. It also makes sense intuitively that students with low levels of apprehension about writing also would have generally high levels of self-confidence in their ability to write. Of course, as Daly and others have noted, one cannot state with any confidence that any personal characteristics cause writing apprehension, particularly when researchers employ correlational or causal-comparative methods. High writing apprehension may well cause low self-efficacy beliefs in students.

The researcher hoped to find evidence to apply Kulhavy�s cognitive feedback theory (1977) to the writing-revision process. In his studies of tests, Kulhavy concluded that students who were very confident that they answered a question correctly, but instead answered it incorrectly, would pay the most attention to feedback when they try to correct their mistake. Applying Kulhavy�s theory to writing instruction, it seems that high self-efficacy writers would use global and local instructor comments concerning errors or mistakes to revise their stories more than low efficacy students. High self-efficacy writers also may require feedback that elaborates more in depth upon ways to correct errors in the first drafts, according to Kulhavy and Stock�s (1989) work on the elaboration component of feedback theory. Global feedback on content generally would be more elaborative than feedback to correct mechanical writing errors. The means for high-efficacy students on both local and global feedback scores generally support this argument, but the small effect sizes again fail to show any important differences between low and high efficacy writers on use of feedback. Utilizing a sample with more variability in scores on the feedback instrument could produce larger effect sizes, indicating greater differences between groups.

The hypotheses that were not supported by the results of this study may warrant further consideration. Based upon cognitive learning theory, the researcher expected that students with low self-efficacy beliefs, low writing outcomes expectations, and high writing apprehension would use instructor comments on spelling and other mechanical errors more than other students. The researcher believed they would use local instructor comments to quickly "fix" problems in their first drafts rather than responding to global comments that require higher-level cognitive functioning. The results did not support cognitive learning theory. Perhaps it should not be surprising that low apprehension, high outcomes and high self-efficacy students used both types of instructor comments more than other students. Students with those characteristics probably are better writers who recognize the importance of using instructor comments to revise all problematic areas of first drafts. Using instructor feedback for revisions may have become part of their successful writing strategies. However, according to cognitive writing theory, even the low apprehension, high efficacy, high outcomes writers apparently have not reached the level where mechanical writing information is stored in long-term memory and is automatically retrieved, freeing the writer to spend more time processing feedback concerning ideas and other global matters. Future inquiries may examine whether high skills writers use local, global or both types of instructor feedback more than low skills writers.

There are several indications that the sample for this study is not representative of the general population of undergraduate mass communication students. The sample seems to have a very low level of writing apprehension. The mean writing apprehension test score was high (M = 96.02) compared to the 79.28 mean of Daly�s 1975 sample (n = 164). However, Daly�s sample did not consist of mass communication students, but instead of undergraduates enrolled in basic composition and interpersonal communication classes. A 1978 Daly study using a larger sample (n = 3,602) of composition students yielded a mean again much lower than this sample�s (M = 75.59, SD = 13.35). There is less variability in the scores for this sample, as well. The standard deviation for the sample was 16.39 compared to 18.86 in Daly�s 1975 sample. Daly placed the cut-off score between moderate and low apprehension students at 97, which suggests that the students in this sample experience much less apprehension than most people when faced with a writing assignment. The 1978 Daly study concluded that high writing apprehension can lead to students ignoring instructor feedback. But the students in the current study, who had low writing apprehension, may tend to welcome and use teacher comments more than other undergraduate students do. This tendency by the sample to embrace feedback rather than avoid it could have affected the results. Students in this sample also may be better writers than are other mass communication students. Their mean grade-point-average was 3.33 in freshman composition, which would translate to a B-plus on many campuses. There also was little variability in the grades (SD = .68), with only one D reported and no F grades. It is possible that the self-reported GPA may be inflated. Furthermore, grades probably are not the most accurate measure of students� true writing abilities. On the other hand, one might expect mass communication students to have above-average writing skills. Writing courses are usually required in the college curriculum, and the university where this study was conducted requires three writing courses to graduate. Students seeking mass media careers also could be expected to have above-average communication abilities, including written communication skills. It would make little sense for a student to major in journalism if he or she could not write, if not well, at least competently.

Conducting the study with a more heterogenous sample of students that has more variability in writing apprehension and writing ability could yield more convincing results. However, because the population of interest for this study was undergraduate mass communication students, another sample drawn from the same population might well show similar results. Since nearly 48 percent of the students participating in this study were non-majors, one might expect the sample to be sufficiently heterogenous to show sizable differences between groups.

It is possible that the feedback instrument lacks internal consistency, failed to discriminate adequately among different student characteristics, or simply failed to actually measure whether students would use global or local feedback. Although the researcher followed recommended procedures to make the instrument reliable, the Cronbach's alpha coefficient dropped from .76 in the pilot study to .64 in the actual study. That result could be attributed to the larger sample size in the actual study. Since student use of feedback may be a fairly abstract attitudinal object, increasing the number of items in the 10-item feedback instrument might be necessary to attain satisfactory reliability.

Although the researcher followed recommended procedures to make the instrument valid, it is possible that student use of local or global feedback may not be measurable by a questionnaire-type instrument. Straub and Lunsford (1995) raised the question of whether posing a hypothetical context would affect studies of teacher comments on student writing. They suggested an ethnographic study of how students react to teacher comments. The feedback instrument in this study asked students to assume they had written the first draft of the news story and that the teacher comments were to be used in writing their second drafts. Although students and content experts evaluated the instrument during its development and subsequent revisions, validity may still be questionable. The teacher comments on the instrument and the four-item student response scale may well be valid on their face, but in the actual context of the classroom, students probably would not use many of the teacher comments in revising their first drafts. The best way to determine the validity of the feedback instrument might be to compare the results to a performance-based test using stories students actually wrote. Instructors would write comments similar to those on the instrument, and then see if students made changes suggested by the comments when they handed in the second drafts.

There are other indications that the instrument was reliable. The correlations between each item and the total-scale scores were all above the .25 minimum recommended by Ary, et al. (1996) to be considered useful. Each item in the feedback instrument apparently discriminated among participants in the same way as the total score discriminated. There also is validity evidence that the scale discriminated between groups whose members are known to have different attitudes. The statistically significant ANOVA results indicated that high and low groups in efficacy, apprehension and writing outcomes expectations were different, although the differences were small, according to effect sizes.

Even with a valid instrument to measure feedback use, the causal-comparative research design would not permit one to state with any confidence that self-efficacy, outcomes expectations, or writing apprehension cause student use of different types of feedback. Causal-comparative studies should also consider reversed-causality, Ary, et al. (1996) recommend. It could well be that global feedback causes writing apprehension in students. Those students might be confused about what the instructor wants (Sommers, 1982), or they might be unable to retrieve global information from long-term memory (Flower & Hayes, 1981), or they may view any form of teacher comments as yet another failure on their part (Daly, 1978). Instructor comments that praise students could increase their self-efficacy, while critical feedback could lower self-efficacy beliefs.

Common cause, or another variable that accounts for differences in both the independent and dependent variables, should also be considered by causal-comparative researchers, Ary, et al. (1996) suggested. In this study, the common cause for differences in all the variables could be writing skill. Bandura (1986) noted that competency requires both skills and self-beliefs of one�s efficacy to use those skills. Even though people have high levels of self-efficacy that they have the skills to complete a task, unless they actually possess those skills, they probably will not succeed. Data that measured writing skills were collected for the study because researchers suspect that writing apprehension and poor writing skills may be reciprocal; that is, poor writing may cause writing apprehension or vice versa (Daly, 1985; Hillocks, 1986; Selfe, 1985). Some research has shown that a timed, in-class writing sample is the most valid measure of quality, albeit not perfect (Foster, 1983). Both Foster and Hillocks (1986) recommended that researchers gauge writing quality by use of holistic scoring by expert readers. Because of the size of the sample and time restraints upon data collection, the researcher decided it would not be feasible to collect and analyze writing samples from each participant. Participants were asked in the questionnaire for their grade in the required freshman composition course or on the most recent writing course they had taken, although the researcher acknowledges that there may be more valid and reliable methods of assessing writing skill. Pajares and Johnson (1993) noted that assessing the quality of writing is not "an objective or clearly defined task" (p. 8). Other writing researchers have used grades as an indicator of writing skills. For instance, Daly and Miller (1975b) asked subjects to provide their SAT-Verbal scores for a study on writing apprehension. While there may be questions about the accuracy of self-reported grades in writing classes, Daly and Miller reported that students were surprisingly accurate when reporting their scores.

The mean grade-point average for freshman composition was 3.33 for this sample. The relationships were investigated among writing skills, as measured by GPA, and the independent variables in this study. The correlation between writing skills and self-efficacy skills (r = .17) was statistically significant but low. The correlation between writing skills and writing apprehension (r = -.21) also was significant but low. There was a non-significant, weak, negative correlation between writing skills and writing outcomes expectations. To check whether writing skills might be a common cause in this study, three levels of writing skills were created: A, B and C. The one D grade reported by the sample was folded into the C level to facilitate data analysis. There were no F grades.

Since the groups were of unequal n�s, Levene Tests of Homogeneity of Variances were run to ensure the assumption of homogeneity of variance was met to permit use of ANOVA for testing the differences among groups. The independent variable of freshman composition grade was named Grade. As shown in Table 4, a one-way ANOVA showed statistically significant differences at .05 alpha level of significance among the freshman composition grades with Global as the dependent variable, and significant differences among the three freshman composition grades with Local as the dependent variable. A subsequent Tukey multiple comparison test indicated that students with A grades (M = 3.31) used global instructor comments more than B students (M = 3.12), at the .05 level of significance. The Tukey test also indicated that A students (M = 3.64) used local instructor feedback more than B students (M = 3.44). Effect sizes, however, were low, indicating that the differences were minor.

Table 4

_______________________________________________________

Source df MS F df MS F

_______________________________________________________

Grade 2, 178 1.035 5.033** 2, 178 .852 5.048**

_______________________________________________________

**p < .01.

Although the effects of writing skills on student use of feedback were not hypothesized in this study, the relationship between freshman composition grades and feedback use could be a common cause. Of course, use of a method of measuring writing skills other than self-reported grades could produce different results. Foster (1983) and Hillocks (1986) recommend measuring writing skills by having expert readers use holistic scoring on timed, in-class writing samples from students. This researcher agrees that writing samples from students or use of a standardized writing skill test would probably provide a more valid measure of writing skills for further study.

Another problem with this causal-comparative study is the possibility that uncontrolled variables could be responsible for variation in the student use of feedback. One remedy that Ary, et al. (1996) recommended to provide some control in causal-comparative studies is to use homogenous groups. In this study, efforts were made to ensure the sample was homogenous. Students were all enrolled in mass communication classes, were undergraduates, and mostly were in the same age range. However, random assignment to ensure equality of groups was not possible in this study since students already possessed the characteristics of interest in some degree.

Replication of the study with a different sample that is more heterogenous on writing skills and writing apprehension in particular is recommended. While the results of this study were inconclusive, the researcher still believes that different types of students will use global or local instructor comments differently. The general directions of the mean scores obtained in this study lend support to those beliefs.

The feedback instrument requires a longer period of validation, using different samples under different ecological conditions and in different instructional settings. Reliability also is an issue, because the Cronbach alpha coefficient dropped from the pilot study to the full study. An instrument can be reliable without being valid, but it cannot be valid without being reliable.

Although the writing self-efficacy constructs and writing apprehension seemed to be appropriate causes to investigate, based upon the literature review, other independent variables may be considered, such as test anxiety or communication apprehension. Other instruments to measure the constructs, such as the Motivated Strategies for Learning Questionnaire to measure motivation, could be used. Perhaps they are more valid methods to measure the constructs of interest than the instruments used in this study.

Populations of high and low self-efficacy writers, or students with high and low levels of writing apprehension, can be identified. Samples drawn from those populations could be randomly assigned to treatment and control groups to more accurately assess the causes of student use of feedback. For instance, a feedback pretest could be given to participants, and then a group of high-apprehension writers could be given a treatment to reduce their apprehension, such as biofeedback. A post-test could reveal whether there are differences in feedback use between the treatment and control groups.

Finally, however, future inquiries should reconsider whether students use global or local feedback is the appropriate question to ask. The mean scores in this study suggest that students will use both global and local, no matter what their individual writing characteristics. It may be that the focus of teacher comments � local or global � is not important to students. Instead, perhaps different types of students will use different modes of teacher comments on their first drafts. Straub�s (1997) study of what modes of teacher comments students prefer could be adopted to investigate what mode of teacher comments students will use in revising first drafts. For instance, will highly apprehensive writers use comments framed as imperatives more than low-apprehension students? Or will students with low self-efficacy beliefs use comments phrased as questions more than high efficacy students? Hypotheses about other modes of comments also could be developed.

Gebhardt (1990) noted that writing teachers should consider research on writing to be neither all accurate nor all useful. Rather, he recommended, they should wonder how the conclusions might apply to their teaching, and "to test it against their own writing behaviors and those they observe in their students" (p. 39).

Based upon the findings of this study, the implications for instructional practice include:

• Instructor comments may affect students' writing apprehension and self-efficacy beliefs.
• Teachers should include a feedback "reprocessing" stage in their instruction.
• Instructors should consider the political implications of their feedback.
• Identifying writing apprehension could aid students who have high levels of apprehension.

High writing apprehension, low writing self-efficacy, and low writing outcomes expectations were detected among students in this sample. Three students scored below 60 on the Writing Apprehension Test, which would have put them among Daly�s high apprehension group scoring one standard deviation below his sample�s mean of 78. (Many writing researchers use Daly�s categories of 26-59 as high apprehensives, 60-96 as moderate apprehensives, and 97-130 for low apprehensives). Continuing to use Daly�s practice of placing those one standard deviation below the mean in low self-efficacy and low writing outcomes groups, there were 22 students with low levels of self-efficacy beliefs in their writing abilities, and 34 students who judged writing to be unimportant to meeting goals. The causes of their attitudes toward writing are unclear. For instance, as some researchers have noted, poor writing skills may cause writing apprehension or writing apprehension could lead to poor writing. A similar reciprocal relationship could also be true of instructor feedback. Writing apprehension, self-efficacy beliefs, and writing outcomes expectations might cause students to use or not use different types of instructor feedback. However, the literature strongly suggests that instructor�s comments, both written and verbal, can affect writing apprehension and students� self-efficacy beliefs toward writing. Journalism and other writing teachers need to continue exploring a balance between comments that facilitate learning and those that cause damaging affective reactions in students.

This study further demonstrates how little is known about how students actually process instructors� comments when they revise their written products. Most writing instructors openly express frustration over how much effort they put into crafting constructive criticism on student assignments, and how little effort they see students expend on second drafts despite the feedback. Students in this sample reported that they definitely would use or might use both focuses of instructor comments to fix mechanical and content problems in their second drafts. But, in practice, researcher after researcher note how few changes many students make in subsequent drafts despite receiving extensive instructor comments on earlier drafts. This study raises the question of why students say they would use instructor comments to improve second drafts, but in practice, most do not use them. Since the feedback instrument was self-report, it is possible that respondents did not answer honestly. But giving students the benefit of the doubt, what happens between their intent to use feedback to improve their first drafts and the final draft that often shows few signs that they even read the teacher comments? Are writing instructors neglecting to address a pedagogic stage between getting the first draft handed back from the instructor and then revising the first draft into a final product? Perhaps a required conversation to discuss the feedback could be held in the days immediately after a student gets back a rough draft. Students could explain why they used a certain word, for instance, in case the instructor was confused by word choice. Instructors could clarify their written comments so students do not misinterpret them. The dialogues could be face-to-face, over the telephone, or even via electronic mail to make them convenient and conducive to individual communication and learning styles. In cognitive feedback theory, as discussed by Kulhavy and Stock (1989), "errors are of central importance" (p. 282) because information-processing systems have mechanisms to correct them. In the feedback systems discussed by Kulhavy and Stock, a stage in which the learner processes initial instructor feedback and then reprocesses subsequent feedback could be incorporated. This feedback reprocessing stage would become a mechanism for learner and instructor to correct errors resulting from the initial feedback. To extend cognitive feedback theory into the social-cognitive domain, the communication between instructor and learner becomes a social-interaction process in which they hold a dialogue about the initial feedback and negotiate the meaning of that feedback.

The instructor-learner interaction also takes on political implications. Mohr (1984) noted that students fear losing control of their writing when they are told to revise first drafts. Other students, according to Mohr, consider revising stories a way to gain control over their writing. Straub and Lunsford (1995) expressed concern that certain types of teacher comments usurp control of student writing. However, most instructors feel obligated to provide comments on student writing because feedback can improve learning (Glover, et al., 1990; McKeachie, 1999). The negotiations about feedback between instructors and students can become a struggle for control over the text, perhaps ending in a compromise between the two parties or an impasse that will require further communication. Considering the political and social implications of feedback, the instructor-learner interaction becomes more than a means to simply correct mistakes in the Kulhavy cognitive model.

Finally, McLeod's (1987) call for writing researchers to develop a "theory of affect" remains largely unanswered. If writing instructors can help students understand how their emotional states influence their writing, including their use of teacher comments, some students may seek help for writing apprehension, for instance. Some students could benefit by simply becoming aware that writing apprehension is a recognized psychological construct, that many people are affected by it, and that they are not alone. Others could seek counseling for their apprehension or various forms of treatment, such as biofeedback or relaxation techniques. Since many studies have found a negative relationship between writing apprehension and self-confidence, reducing high levels of writing apprehension might have the beneficial side effect of increasing self-efficacy beliefs in their writing abilities. As students become more self-confident and less anxious about writing, it is possible that they will see the value of writing toward achieving their various life goals, further increasing their motivation to write. As Flower and Hayes (1981) noted, writing is a goal-directed process, and setting and meeting goals is an outcome that can motivate. Top

Ary, D., Jacobs, L.C., & Razavieh, A. (1996). Introduction to research in education (5^th ed.). Fort Worth, TX: Harcourt Brace College Publishers.

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, N.J.: Prentice Hall.

Bell, E., & Price, A. (1982). Some effects of grade anxiety on performance in composition sections. Research in the Teaching of English,16, 3, 285-94.

Berner, R.T. (1992). The process of writing news. Boston: Allyn and Bacon.

Black, J., & Bryant, J. (1995). Introduction to media communication (4^th ed.). Dubuque, IA: Brown & Benchmark.

Bohlin, R., Milheim, W. & and Viechnicki, K. (1990). A model for the

motivational instruction of adults. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 323 918)

Bracht, G.H. & Glass, G.V. (1969). The external validity of experiments. American Educational Research Journal,5, 437-474.

Briggs, L.J., & Hamilton, N.R. (1964) Meaningful learning and retention: Practice and feedback variables. Review of Educational Research,34, 545-558.

Burkland, J., & Grimm, N. (1984). Students� response to our response: Parts I and II. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 245 241)

and Composition, 12, 2, 171-81.

Covington, M.V. (1993). A motivational analysis of academic life in college. In

J.C. Smart (Ed.), Higher education: Handbook of theory and research, v. 9. New York: Agathon Press.

the scholarship of teaching. San Francisco: Josey-Bass.

Daly, J.A. (1977). The effects of writing apprehension on message encoding.

Journalism Quarterly, 55,3, 566-72.

Daly, J.A. (1978). Writing apprehension and writing competency. The Journal

of Educational Research, 72, 10-14.

Daly, J.A. (1979). Writing apprehension in the classroom: Teacher role expectancies of the apprehensive writer. Research in the Teaching of English,13, 37-45.

Daly, J.A. (1985). Writing apprehension. In Rose, M. (Ed.), When a writer can�t write (pp. 43-82). New York: Guilford Press.

Daly, J.A., & Hailey, J.L. (1984) Putting the situation into writing research: State

and disposition as parameters of writing apprehension. In R. Beach & L.S. Bridwell, (Eds.), New directions in composition research (pp. 259-273). New York: Guilford Press.

Daly, J.A., & Miller, M.D. (1975a). The empirical development of an instrument to measure writing apprehension. Research in the Teaching of English, 9, 242-49.

Daly, J.A., & Miller, M.D. (1975b). Further studies on writing apprehension: SAT scores, success expectations, willingness to take advanced courses and sex differences. Research in the Teaching of English, 9, 250-56.

Daly, J.A., & Shamo, W. (1976). Writing apprehension and occupational choice. Journal of Occupational Psychology,49, 55-56.

Daly, J.A., & Shamo, W. (1978). Academic decisions as a function of writing apprehension. Research in the Teaching of English,12, 119-126.

Daly, J.A., & Wilson, D.A. (1983). Writing apprehension, self-esteem and personality. Research in the Teaching of English,17, 327-341.

Dickinson, P.F. (1992). Feedback that works: Using the computer to respond.

East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 356 490)

Dohrer, G. (1991). Do teachers� comments on students� papers help? College

Teaching,39, 2, 48-54.

Faigley, L., Daly, J.A., & Witte, S.P. (1981). The effects of writing apprehension on writing performance and competence. Journal of Educational Research,75, 16-21.

Fellow, A.R. & Clanin, T.N. (1998). Copy Editors� Handbook for Newspapers. Englewood, CO: Morton Publishing Co.

Ferris, D. (1995). Teaching students to self-edit. TESOL Journal,4, 4, 18-22.

Fey, M.H. (1993). Reader response, collaborative writing, and computer networking. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 364 875)

Fischer, R., & Grusin, E.K. (1993). Grammar checkers: Programs that may not

enhance learning. Journalism Educator;47, 4, 20-27.

Flower, L.S. (1989). Cognition, context, and theory building. College Composition and Communication,40, 282-311.

Flower, L. (1994). The construction of negotiated meaning: A social cognitive theory of writing. Carbondale and Edwardsville, IL: Southern Illinois University Press.

Flower, L.S., & Hayes, J.R. (1981). A cognitive process theory. College Composition and Communication,32, 365-387.

Foster, D. (1992). A primer for writing teachers: Theories, theorists, issues, problems (2^nd ed.) Portsmouth, NH: Boynton/Cook.

Fox, R. F. (1980). Treatment of writing apprehension and its effects on

composition. Research in the Teaching of English,14, 1, 39-49.

Fox, W. (1993). Writing the news: A guide for print journalists (2^nd ed.). Ames, IA: Iowa State University Press.

Gall, M.D., Borg, W.R., & Gall, J.P. (1996). Educational research: An introduction (6^th ed.). White Plains, NY: Longman Publishers USA.

Gebhardt, R.C. (1990). Introspection and observation for insight and instruction in the processes of writing. In D.A. Daiker & M. Morenberg (Eds.), The writing teacher as researcher: Essays in the theory and practice of class-based research (pp. 38-51).

Portsmouth, NH: Boynton/Cook.

Gee, T.C. (1972). Students� responses to teacher comments. Research in the Teaching of English,6, 212-221.

Glass, G.V., & Hopkins, K.D. (1996). Statistical methods in education and psychology (3^rd ed.). Boston: Allyn & Bacon.

Glover, J.A., Ronning, R.R., & Bruning, R.H. (1990). Cognitive psychology for

teachers. New York: Macmillan Publishing Co.

Grant-Davie, K., & Shapiro, N. (1987). Curing the nervous tick: Reader-based

response to student writing. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 282 196)

Hashway, R., Duke, I., & Hammond, C. (1992). Test anxiety and reading test performance. College Student Journal, 26, 180-182.

College English, 55,5, 627-43.

Heinich, R., Molenda, M., Russell, J.D., & Smaldino, S.E. (1996). Instructional media and technologies for learning (5^th ed.). Englewood Cliffs, NJ: Prentice-Hall.

Hill, A., & Boone, B. (1982). If Maslow taught writing. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 250 715)

Hillocks, G. (1986). Research on written composition: New directions for teaching. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 265 552)

Hillocks, Jr., G. (1990). Teaching, reflecting, researching. In D.A. Daiker & M. Morenberg (Eds.) The writing teacher as researcher: Essays in the theory and practice of class-based research (pp. 15-29). Portsmouth, NH: Boynton/Cook.

Hopkins, K.D. (1998). Educational and psychological measurement and evaluation (8^th ed.). Boston: Allyn and Bacon.

Houle, C.O. (1961). The inquiring mind. Madison, WI: University of Wisconsin Press.

Hurd, R.N. (1985). The effect of audience specification on writing anxiety, performance, and sensitivity to audience. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 269 767)

for Research on Teacher Learning. (ERIC Document Reproduction Service

No. ED 289 463)

Journal, 77, 7, 39-41.

The state of the art. Freshman English News,10, 1-4.

pedagogy. New York: Association Press.

test anxiety in college students. College Student Journal,27, 394-296.

Research, 47, 1, 211-232.

of response certitude. Educational Psychology Review,1, 4, 279-308.

confidence. Journal of Educational Psychology,68, 522-528.

review in programmed instruction. Contemporary Educational Psychology,4, 91-98.

Lackey, J.R., Miller, R.B., & Flanigan, M.C. (1997). The effects of written feedback on motivation and changes in written performance. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 406 690)

Land, R.E., & Evans, S. (1987). What our students taught us about paper marking. English Journal,76, 2, 113-16.

Lauer, J.M., & Asher, J. W. (1988). Composition research: Empirical designs. New York: Oxford University Press.

Lucas, A.F. (1990). Using psychological models to understand student motivation. In M.D. Svinicki (Ed.), The changing face of college teaching (pp. 103-114). New directions for teaching and learning. San Francisco: Jossey-Bass.

Lynch, M.D., & May, L. (1977). Some effects of heightening anxiety levels on writing performance of students with different levels of creativity and prior anxiety. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 141 814)

(Eds.), Management and motivation (pp. 27-41). Baltimore: Penguin Books Inc.

Maslow, A.H. (1970). Motivation and personality (2^nd ed.). New York: Harper & Row.

McCarthy, P., Meier, S., & Rinderer, R. (1985). Self-efficacy and writing: A different view of self-evaluation. College Composition and Communication,36, 465-471.

M. Rossman (Eds.), Applying adult development strategies (pp. 65-75). San Francisco: Jossey-Bass.

McLeod, S. (1987). Some thoughts about feelings: The affective domain and the writing process. College Composition and Communication,38, 426-435.

McCroskey, J.C., Daly, J.A., Richmond, V.P., & Falcione, R.A. (1977). Studies of the relationship between communication apprehension and self-esteem. Human Communication Research,3, 269-277.

McKeachie, W.J., Pintrich, P.R., Lin, Y., & Smith, D. (1986). Teaching and learning in the college classroom: A review of the research literature. Ann Arbor, MI: National Center for Research to Improve Postsecondary Teaching and Learning, University of Michigan.

college and university teachers (10^th ed.). Boston: Houghton Mifflin.

Mealey, D.L., & Host, T.R. (1992). Coping with test anxiety. College Teaching,40, 4, 147-150.

Medsger, B. (1996). Winds of change: Challenges confronting journalism education. Arlington, VA: The Freedom Forum.

Mohr, M.M. (1984). Revision: The rhythm of meaning. Upper Montclair, NJ: Boynton/Cook.

Mory, E.H. (1996). Feedback research. In D.H. Jonassen (Ed.), Handbook of research for educational communications and technology (pp. 919-956). New York: Macmillan Library Reference USA.

Newbold, W. (1994). Strategies for computer-based distance writing courses.

East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 377 476)

Norusis, M.J. (1985). SPSS-X Advanced Statistics Guide. NY: McGraw-Hill.

Olson, L.D. (1990). Effect of news writing instruction in English composition

classes. Journalism Educator,47, 2, 50-56.

Onwuegbuzie, A.J. (1998). The relationship between writing anxiety and learning styles among graduate students. Journal of College Student Development,39, 6, 589-598.

Onwuegbuzie, A.J. (1999). Writing apprehension among graduate students: Its relationship to self-perceptions. Psychological Reports,84, 3, 1034-1039.

Pajares, M.G., & Johnson, M.J. (1993). Confidence and competence in writing: The role of self-efficacy, outcome expectancy, and apprehension. (ERIC Document Reproduction Service No. ED 358 474)

Phinney, M. (1991). Word processing and writing apprehension in first and second language writers. Computers and Research,9, 1, 65-82.

Pintrich, P.R. (1988). A process-oriented view of student motivation and cognition. In J.S. Stark and L.A. Mets (Eds.), Improving teaching and learning through research (pp. 65-79). New directions for institutional research, no. 57. San Francisco: Jossey-Bass.

Pintrich, P.R., & DeGroot, E.V. (1990). Motivational and self-regulated learning components of classroom academic performance. Journal of Educational Psychology,82, 1, 33-40.

Pitts, B. (1989). Model for news writing process. Journalism Educator,44, 1, 12-19, 59.

Pullman, G. (1999). Stepping yet again into the same current. In T. Kent (Ed.), Post-process theory: Beyond the writing-process paradigm (pp. 16-29). Carbondale, IL: Southern Illinois University Press.

Rechtien, J.G., & Dizinno, G. (1998). Apprehension about writing: Further results. Psychological Reports,82, 221-222.

Reed, W.M., Burton, J.K., & Vandett, N.M. (1988). Daly and Miller�s writing apprehension test and Hunt�s t-unit analyses: Two measurement precautions in writing research. Journal of Research and Development in Education,21, 2, 1-8.

Richardson, E.M. (1981). Research on writing: Apprehension, quality, and audience. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 204813)

Riffe. D., & Stacks, D.W. (1988). Dimensions of writing apprehension among mass communication students. Journalism Quarterly,65, 2, 384-91.

Riffe. D., & Stacks, D.W. (1992). Student characteristics and writing apprehension. Journalism Educator,47, 2, 39-49.

Rosenthal, R., & Rosnow, R.L. (1975). The volunteer subject. New York: Wiley.

Salovey, P., & Haar, M.D. (1990). The efficacy of cognitive-behavior therapy and writing process training for alleviating writing anxiety. Cognitive Therapy and Research, 14, 5, 513-526.

Sapp, M.. (1996). Three treatments for reducing the worry and emotionality components of test anxiety with undergraduate and graduate college students: Cognitive-behavioral hypnosis, relaxation therapy, and supportive counseling. Journal of College Student Development, 37, 1, 79-87.

Schunk, D.H. (1991). Self-efficacy and academic motivation. Educational Psychologist,26, 207-231.

education. Education, 89,2, 267-271.

Selfe, C.L. (1985). An apprehensive writer composes. In M. Rose (Ed.), When a writer can�t write (pp. 83-95). New York: The Guilford Press.

Shell, D. F., Murphy, C. C., & Bruning, R. H. (1989). Self-efficacy and outcome expectancy mechanisms in reading and writing achievement. Journal of Educational Psychology,81, 91-100.

Small, R.V. (1997). Motivation in instructional design. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. EDO-IR-97-06)

Smith, M.W. (1984). Reducing writing apprehension. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 243 112)

Smith, W.E. (1993). Computers vs. paper in delivering news story rewrite advice. Journalism Educator, 1, 52-58.

Smith, W.E. (1990) News writing students favor computer simulations. Journalism Educator, 2, 25-31.

Sogunro, O.A. (1998). Impact of evaluation anxiety on adult learning. Journal of Research and Development in Education, 31, 109-120.

Sommers, N. (1982). Responding to student writing. College Composition and Communication, 33, 2, 148-56.

treatment. Bristol, PA: Taylor and Francis.

Straub, R. (1997). Students� reaction to teacher comments: An exploratory study. Research in the Teaching of English,31, 1, 91-119.

Straub, R., & Lunsford, R.F. (1995). Twelve readers reading: Responding to college student writing. Cresskill, NJ: Hampton Press.

Thompson, T. (1993). Personality preferences and student responses to teacher comments. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 357 392)

Thompson, T. (1994). Giving them what they want: Student views of revision. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 370 139)

Tighe, M.A. (1987). Reducing writing apprehension in English classes. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 281 196)

Tough, A. (1979). The adult's learning projects: A fresh approach to theory and practice in adult learning. Toronto: Ontario Institute for Studies in Education.

Vroom, V.H. (1964). The nature of the relationship between motivation and

performance. In V. Vroom & E. Deci (Eds.), Management and motivation (pp. 229-236). Baltimore: Penguin Books Inc.

(Eds.) Management and motivation (pp. 91-106). Baltimore: Penguin Books Inc.

Bess (Ed.), Teaching well and liking it (pp. 3-18). Baltimore: The Johns Hopkins University Press.

Walling, D.R. (1987). A model for teaching writing: Process and product. Bloomington, IN: Phi Delta Kappa Educational Foundation.

Walsh, S.M. (1986). New directions in research on writing apprehension. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 361 716)

Walsh, S.M. (1989). The subtleties of writing apprehension: Fitting the pieces together. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 335 707)

and emotion. Psychological Review, 92, 4, 548-573.

Westmoreland, R., Starr, D.P., Shelton, K., & Pasadeos, Y. (1977). News writing styles of male and female students. Journalism Quarterly,54, 3, 599-601.

Williams, S.E. (1997). Teachers� written comments and students� responses: A socially constructed interaction. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 408 589)

improving instruction and increasing learning achievement. San Francisco: Jossey-Bass.

Wlodkowski, R.J. (1999). Enhancing adult motivation to learn: A comprehensive guide for teaching all adults (Revised). San Francisco: Jossey-Bass.

freshmen. In R. Beach & L.S. Bridwell (Eds.), New directions in composition research (pp. 362-380). New York: Guildford Press.

written comments. East Lansing, MI: National Center for Research on Teacher Learning. (ERIC Document Reproduction Service No. ED 215 361)

APPENDIX A

Demographic and Biographic Questionnaire

Dear Participant,

Thank you for agreeing to participate this semester in a study about instructor feedback and writing assignments. Your participation in the research will help provide information that could improve instruction in writing courses at the University of Wyoming and other campuses.

Please answer this packet of questionnaires openly and honestly. When you are finished with the questionnaires, you will be asked to read a short news story and respond to the teacher's comments on it.

Your identity will be held in confidence during the study, and you will remain anonymous in any subsequent report about the results of the study. All information about you will be recorded under the last four digits of your Social Security number.

Last four digits of your Social Security No. _______________________

Age: __________ (Note: If you are under 18 years old, you cannot take this survey without parental consent.)

Please circle your selection

1. Gender

2. Race:

3. Major

4. Year

5. Grade you received in Freshman Composition ENGL 1010 (If you are currently enrolled, provide your grade so far in the course).

6. University Studies writing class taken (circle highest level you have completed)

APPENDIX B

Participant Consent Form

Dear UW student,

You are invited to participate this semester in a study about instructor feedback and writing assignments. Your participation in the research will help provide information that could improve instruction in writing courses at the University of Wyoming and other campuses.

As part of the study, you will be asked to complete a survey. Your identity will be held in confidence during the study, and you will remain anonymous in any subsequent report about the results of the study.

Your grade will not be affected whether you agree to participate or not. Participation in the study is voluntary. If at any time during the study, you choose to withdraw, your request will be honored. The study presents minimal or no foreseeable physical or psychological risks to you.

Eric Wiltse, an associate lecturer in the Department of Communication and Mass Media at the University of Wyoming, is the investigator directing the study. If you have any questions or want more information, feel free to write him at P.O. Box 3904, Laramie, WY 82071, send e-mail to ewiltse@uwyo.edu, visit his office in 436 Ross Hall, or call (307) 766-6445.

You also may contact him if you would like your results on the surveys. Please allow two weeks before requesting your results.

Thank you in advance for your help in this research.

If you agree to participate in the study described above, please sign below. Thank you.

___________________________ __________________

Participant signature Date

APPENDIX C

Student Use of Feedback Instrument

BACKGROUND: I am conducting research on instructor feedback and writing assignments. Research indicates that students will use some teacher comments on their rough drafts to make changes in second drafts of their papers. However, students will ignore other teacher comments and not make changes in second drafts.

I want to look at what students think about teacher comments. For this project, I am particularly interested in finding out which kinds of comments you would or would not use for writing a second draft of the following story.

INSTRUCTIONS: Assume that you are getting back a first draft that you have written for a journalism class and are going to revise for a final draft.

Which comments from your instructor would you actually use for revising your story? Which comments would you not use? Read the story and teacher comments, and then rate the comments on the attached questionnaire. Rate the comments on a scale of 1-4 with 1 being definitely will use and 4 being definitely will not use.

A teenager was arrested Thursday for allegedly shooting a friend during an

argument over who would drive the car they had just stolen.

Bobby Baston, 18, allegedly shot Paul Hvastkovs, 18, in a parking lot, where they had driven a car they took from a 7-Eleven store about 5 p.m. He was shot in the arm, according to the police report.

The car belonged to Martha Katz, of 345 18th St., Mountain View. Katz admitted to cops that she left her keys in her 1994 Chevy.

"I wasn't going to be gone that long, and I shouldn't have to worry about thieves, said Katz.

Katz, who is black, said she ran into the 7-Eleven to buy a six-pack.

With Hvastkovs driving, the boys took the car back to their high school to show to friends, Police Dect. Bob Chavez said. Chavez said the boys found a loaded gun in the car's glove box when Hvastkovs refused to let Baston drive the car, Baston grabbed the gun and shot Hvastkovs.

Chavez said drugs were found in the car during a police search. He did not know the value or amount of the drugs.

Several North High School students witnessed the arrest. One student, who refused to give his name, said the suspects had stolen money from his school locker and used it to buy drugs.

INSTRUCTIONS: The teacher comments in the above story correspond with the comments below. Which teacher comments would you actually use to revise the story? Please indicate whether you would actually use the teacher comments for revising the first draft by circling one of the descriptors.

1. Where did incident happen?

   1-definitely will use 2-Might use 3- Probably won't use 4-definitely won't use

2. Did you verify name spelling?

   1-definitely will use 2-Might use 3- Probably won't use 4-definitely won't use

3. Was he hospitalized?

   1-definitely will use 2-Might use 3- Probably won't use 4-definitely won't use

4. Should you call them cops?

1-definitely will use 2-Might use 3- Probably won't use 4-definitely won't use

5. Where does quote end?

1-definitely will use 2-Might use 3- Probably won't use 4-definitely won't use

6. Relevance? Do police suspect it's hate crime?

   1-definitely will use 2-Might use 3- Probably won't use 4-definitely won't use

7. Is style correct?

   1-definitely will use 2-Might use 3- Probably won't use 4-definitely won't use

8. I don't understand sentence.

   1-definitely will use 2-Might use 3- Probably won't use 4-definitely won't use

9. Whose drugs were they?

   1-definitely will use 2-Might use 3- Probably won't use 4-definitely won't use

10. Verify anonymous information with police.

1-definitely will use 2-Might use 3- Probably won't use 4-definitely won't use

APPENDIX D

Daly-Miller Writing Apprehension Test

INSTRUCTIONS: Select the response from 1 to 5 that best suits your feelings about the following statements. Remember, there are no correct answers. Only give your honest response to each item.

1. I avoid writing.

   1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

2. I have no fear of my writing being evaluated

   1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

3. I look forward to writing down my ideas

   1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

4. I am afraid of writing essays when I know they will be evaluated

   1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

5. Taking a composition course is a very frightening experience

   1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

6. Handing in a composition makes me feel good

   1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

7. My mind seems to go blank when I start to work on a composition

   1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

8. Expressing ideas through writing seems to be a waste of time

   1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

9. I would enjoy submitting my writing to magazines for evaluation and publication

   1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

10. I like to write my ideas down

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

11. I feel confident in my ability to clearly express my ideas in writing

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

12. I like to have my friends read what I have written

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

13. I'm nervous about writing

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

14. People seem to enjoy what I write

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

15. I enjoy writing

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

16. I never seem to be able to clearly write down my ideas

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

17. Writing is a lot of fun

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

18. I expect to do poorly in composition classes even before I enter them

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

19. I like seeing my thoughts on paper

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

20. Discussing my writing with others is an enjoyable experience

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

21. I have a terrible time organizing my ideas in a composition course

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

22. When I hand in a composition I know I'm going to do poorly

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

23. It's easy for me to write good compositions

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

24. I don't think I write as well as most other people

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

25. I don't like my compositions to be evaluated

    1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

26. I'm no good at writing

1-strongly agree 2-agree 3-uncertain 4-disagree 5-strongly disagree

APPENDIX E

Writing Self-efficacy Instrument

Contact: Eric Wiltse

766-6445

No chance completely certain

Directions: On a scale from 0 (no chance) to 100 (completely certain), how confident are you that you can perform each of the following writing skills? You may use any number between 0 and 100.

Directions: People differ in their feelings about the importance of writing. On a scale from 1 (extremely unimportant) to 7 (extremely important), please rate how important you think skilled writing is for achieving each of the following.

extremely unimportant extremely important

Top