Content-Type: text/html This paper was presented at the Association for Education in Journalism and Mass Communication in Toronto, Canada, August 2004. If you have questions about this paper, please contact the author directly. If you have questions about the archives, email [log in to unmask] For an explanation of the subject line, send email to [log in to unmask] with just the four words, "get help info aejmc," in the body (drop the ""). (Oct 2004) Thank you. Elliott Parker ************************************************************************ I think I can write, but grammar confuses me: Pre-JMC students' high school and college language skill instruction, perceptions of self-efficacy, and variables that predict success on a required language skills exam by Peter Gade, Ph.D. University of Oklahoma 13817 Norris Circle Norman, OK 73026 [log in to unmask] Paper submitted for consideration of the Scholastic Journalism Division of the Association for Education of Journalism and Mass Communication for the annual convention, Toronto, Ontario, August 2004. Abstract This study explores the perceptions of students entering JMC programs on their experiences in English courses in high school and college, the extent of their language skill instruction, and their sense of language skills self-efficacy. This study finds that students are not getting much grammar and basic language instruction in their college composition courses. For scholastic educators, this study suggests that they do their students a service if they make a considerable effort to teach the basics of the English language. This study also finds that students enter their JMC programs saying they like to write, and their moderately high level of self-efficacy indicates they are motivated and willing to work at improving their language and writing skills. This suggests that a challenge for high school instructors is to strike a balance -- one that provides a stronger foundation in grammar and language while maintaining students' enthusiasm and affinity for writing. 2 I think I can write: I think I can write, but grammar confuses me: Pre-JMC students' high school and college language skill instruction, perceptions of self-efficacy, and variables that predict success on a required language skills exam. The importance of language skills (grammar, punctuation, usage and writing) for journalism and mass communication students has been a focus of debate for at least the past quarter century (Oukrup & Brown, 1977). More recently, there is a general concern that there is less emphasis on language skills in journalism curricula, as journalism/mass communication (JMC) schools have become more focused on new media, emerging technology, multi-cultural awareness and balancing graduate and undergraduate programs (Seamon, 2001). Media professionals have questioned what JMC schools are doing to produce graduates with sufficient language skills to succeed in their careers (Kees, 1996; Urban, 1998). Although there is widespread agreement among JMC faculty and administrators concerning the value of their students having strong language skills (Geimann, 2000; Smith, 1997), there is little agreement on how language skills fit into JMC curricula. Some administrators have argued that students should already have strong language skills when they enter a JMC school (Oregon, 1987). However, most JMC schools have required for admission into their programs some combination of: (a) English Composition courses, (b) a minimum grade-point average, and/or (c) a language skills test as evidence of language proficiency (Oukrup, Brown & Parsons, 1998). As JMC faculty struggle to find the best places for and relative emphases on language skills in their curricula, it seems relevant to explore the perceptions of students entering JMC programs on their experiences in English courses in high school and college, the extent of their language skill instruction, and their sense of language skills self-efficacy. The study reported in this paper explores these issues and variables through a 2002-03 survey of more than 200 pre-JMC students at a major southwestern university. This study can assist JMC faculty to understand better what their students perceive they know prior to entering JMC school, which can serve as a guide as educators shape the role of language skills in their curricula and individual courses. This study can also serve scholastic educators as a measure of the strength and value of language skills instruction that students perceive they are getting prior to coming to college. Literature Review Language skills instruction. The perception among JMC faculty that their students lack the necessary language skills to be successful in journalism programs is not new. Williams (1983) found that faculty said the news writing class was the most difficult journalism class to teach because students lack basic English and grammar skills. Stone (1990) lamented that the erosion of student skills considered minimal for success in higher education was a reason why teaching writing has become more difficult. Williams (1985) suggests, "The most common reason for bad writing is, I think, the simplest: Most writers have just never learned how to write clearly and directly in the first place" (p. 5). Much of "blame" for deficient language skills has been placed, correctly or not, on college English composition courses and their instructors. Olson and Dickson (1995) contend that role of English composition courses is largely undefined, and continues to be unquestioned as faculty tend to agree that students need all the writing instruction they can get. However, this role ambiguity creates composition classes that are "disciplineless," with content that is not necessarily applicable to other fields or academic disciplines. The authors asked about 300 journalism students from programs in 10 states about the value of writing instruction they received in high school and freshman composition classes. They found that 56.2 percent of the students having quite a bit of grammar instruction in high school, but only 16.8 percent said they received quite a bit of grammar instruction in freshman composition. They also found students reported that English composition courses have significantly more focus on writing creatively and opinion writing than journalism writing classes. These findings reinforce what Olson and Dickson call the perception that writing instruction in English composition classes "rewards creativity and ignores grammatical and stylistic correctness" (p. 48). This critique of English writing courses has been connected to creating students who have "unrealistic evaluations" of their writing ability when they enter the introductory level media writing course (Collins & Bissell, 2002). Masse (1998) came to a similar conclusion in a study of his journalism students' writing journals, finding that students with high grades in English composition generally start journalism writing classes with high confidence, but their confidence plummets when their journalism grades do not match the success they had in English composition. Plopper and Rollberg (1996) studied students in the first media writing class and found that "students do not believe that what they learned in English composition courses transfers well to journalism courses" (p. 8). High school English and writing instruction has been shown to lack sufficient emphasis on grammar and basic language skills, although high school journalism experience has generally been shown to have to have positive effects on writing skills. Ryan and Findley (1982) found in a study of journalism students at the University of West Virginia that many students reported they were taught English language skills early in their high school careers and then received little or no such instruction during the remainder of high school. Further, those students who received substantial grammar instruction in 12th grade English courses scored significantly higher on a college-administered diagnostic English test than those who did not study grammar in their senior year of high school. Exploring the impact of high school English and journalism courses on language skills, Dvorak (1990) found that students who took at least one high school journalism course responded that they perceived journalism courses prepared them to fulfill language skills competencies better than English courses. Collins and Bissell (2002) studied university students' self-efficacy (their confidence in their ability) and found that students with high school journalism experience (newspaper or yearbook) had higher levels of self-efficacy. In another study, Bissell and Collins (2001) found that students with high school journalism experience wrote significantly better stories than their peers without such experience. This finding, too, held for both students with high school newspaper and yearbook experience; however, the study also found that high school journalism experience (newspaper or yearbook) was not a significant predictor of student success on a grammar exam. Writing and language skill self-efficacy. Several journalism and mass communication scholars in recent years have adapted concepts from social psychology and social learning theories to better understand predictors of student writing success. The most common concept that has been studied is student writing or language skill self-efficacy. Bandura (1978, 1986) contends that as people reflect on their capabilities, one concept that is important is their perceived self-efficacy. Self-efficacy has been tested in many studies across numerous disciplines, and the pattern of results supports that people who possess particular skills tend to have higher self-efficacy in that particular skill area. Perceived self-efficacy also affects behavior, as those people with higher levels of self-efficacy appear to have more motivation and perseverance on tasks related to the self-efficacy. Thus, self-efficacy can be understood as a type of self-fulfilling prophesy (Collins & Bissell, 2002). Students who have higher levels of self-efficacy have more confidence that they can succeed at a task, so they are more likely to work harder and perform better, which leads them to perceive a higher level of self-efficacy. In a study of predictors of success in the first media writing course, Bissell and Collins (2001) found that there was a "modest but significant" correlation (r = .16) between self-efficacy and writing performance. However, they also found no significant relationship between self-efficacy and grammar scores. College GPA was the only variable in their study that was significantly correlated with success on a writing test and a grammar test. They concluded: Students appear to be rather poor judges of their own command of the rules of grammar. The fact that many students clearly have deficiencies in this area yet fail to recognize this may be a byproduct of the fact that many instructors at the high school and college levels appear to pay little attention to grammar when grading assignments (p. 75). Bissell and Collins suggest the relatively low (albeit significant) correlation between language self-efficacy and writing performance is evidence that students, who have written for years before arriving to college, aren't very good judges of their own writing skills. The next year Collins and Bissell (2002) surveyed students in an introductory media writing course at the beginning and end of the semester, looking for relationships between students' self-efficacy and several performance and expectancy variables. At the beginning of the semester, nearly four of five students (79 percent) rated themselves as confident in their overall ability as students, but slightly less than half of the students (48 percent) considered themselves strong writers. By the end of the semester, the percentage considering themselves strong writers increased to 66 percent. However, the trend was reversed when students were asked to rate their language and grammar skills. The beginning of the semester found "about three out of four" students (the authors do not report a specific percentage) saying they had strong command of the English language, but this figure dropped to 64 percent at the end of the term. Similarly, 57 percent said they were confident in their grammar skills at the beginning of the semester, but that number dropped to 48 percent at the end of the course. The authors concluded "it seems telling that while students generally consider themselves better overall writers at the end of the semester, they express less confidence of their understanding of the intricacies of language" (p. 29). The authors suggest this conclusion is reinforced by an additional finding that even though students expressed confidence that their writing skills improved over the course of the semester, 35 percent reported a decrease in self-efficacy at the end of the semester. Language and grammar diagnostic exams as entrance requirements for JMC programs. Oukrop and Brown (1977), in a survey the Association for Education in Journalism President's Committee on Journalism Language Skills, found that 27 percent of the programs surveyed required students to take a language skills exam or similar test. The authors also reported that another 28 percent of the programs responded that they would soon be adding a required language skills exam. However, in a more recent study, Oukrup, Brown and Parsons (1998) reported that most JMC schools apparently did not make a language test a requirement, as 31 percent of those schools surveyed reported having required language exams, and 60 percent reported they have never had a required language exam. Interestingly, 7 percent of the schools reported having a required exam in the past, but no longer. These schools said they discontinued their exams because they were drains on the unit resources or that faculty believed the exam was not a good predictor of student success in the program. John, Ruminski and Hanks (1991), in a survey of 380 AEJMC schools found that 36.4 percent of the responding schools required an exam of English writing skills. Of these schools, about half said they use the tests to screen students, 27 percent said they use the exams to diagnose student language problems, and 12 percent said they use the scores to predict outcomes over time. Of those schools that require an exam, nearly 50 percent said the tests were written by their departments or programs; only 10.5 percent of the programs were using commercial exams. In the most recently published study on the use of language skills exams in JMC programs, Seamon (2001) found that 39 percent of the schools responding indicated they tested the grammar competencies of incoming students, and 26 percent of the schools required students to pass the test before admitting them to the JMC program. Seamon concludes that "the failure of J-schools to adopt practices such as language competency exams on a large scale would seem to suggest that grammar curricula reforms are a tough sell" (p. 62). Oukrup, Brown & Parsons (1998) report that JMC schools have apparently turned to other indicators for predicting student achievement and success, with the three most common requirements being a minimum grade-point average, completion of English composition course(s), and/or completion of an introductory JMC course. The trend that has drawn the most support is the creation of a minimum GPA; in 1977 only 4 percent of JMC schools had such a requirement, in 1997, 39 percent had minimum GPA requirements for admission in the JMC program. In sum, the literature indicates a broad-based concern among educators that language skills are not strong among students entering colleges and universities, and this concern is especially relevant to JMC programs. A substantial amount of research suggests that students are not getting sufficient language and grammar instruction in their college English composition courses, and this seems to be related to student perceptions that they are good writers when they enter JMC programs, despite clear language and grammar skill deficiencies. High school journalism experience has been shown in several studies to be linked to stronger language skills and higher self-efficacy for students entering JMC programs. Students also perceive journalism courses are more focused on language and grammar skills than English courses. Beyond this, studies show that students who have had significant amounts of language skill and grammar instruction late in their high school careers perform better on JMC-administered language and writing skills entrance exams. During the past quarter century, the number of JMC programs that require language and writing skills exams as a condition of admission has remained relatively steady at about 25 to 30 percent. JMC programs are increasingly relying on other measures (GPA, English composition courses, and introductory JMC classes) to be predictors of students' capability of succeeding in their programs. Finally, many cross-disciplinary studies have shown self-efficacy to be an important concept in predicting student performance; however, a recent study of students in an introductory media writing course found that as students became exposed to their language and grammar deficiencies in the course, 35 percent of them found their self-efficacy decreased during the semester. Based on the literature and the researcher's interest in better understanding the language skills pre-JMC students perceive they have as they enter an AEJMC-accredited program, the following research questions were posed and hypotheses tested: RQ1: How much language skill instruction do students seeking entrance to the JMC college perceive they have had? Do students think they have had an adequate amount of language skill instruction to successfully gain entrance in the JMC program? RQ2: What is the level of the students' perceived language skill self-efficacy? RQ3: Is there a significant relationship between amount of language skill instruction and perceived language skill self-efficacy? H1: Students who report greater amounts of language skill instruction will have higher perceived language skill self-efficacy. RQ4: Is there a significant relationship between amount of language skill instruction and grades on the JMC school's required Language Skills Test? RQ5: What variables are the strongest predictors of success on the JMC school's required Language Skills Test? H2: Amounts of language skill instruction will be a significant predictor of the score on the JMC school's required Language Skills Test. H3: Language skill self-efficacy will be a significant predictor of the score on the JMC school's required Language Skills Test. Methods Pre-JMC students at a major southwestern university were asked to complete a 23-statement survey following their completion of the JMC college's Language Skills Test, a requirement for admission into the college's introductory level media writing course. Prior to taking the test, students are required to have completed two English composition courses at the college or university level, or show evidence that they possess the skills taught in those classes. Most students taking the LST are second-semester freshman or sophomores. The surveys were administered by the JMC college's academic adviser between October 2002 and May 2003. The adviser asked students upon their completion of the LST if they would complete the survey. Participation was voluntary. The students were guaranteed confidentiality. Informed consent was obtained from all participants. The study's protocol was approved by the university's Institutional Review Board. Variable measures. The survey was designed to measure variables the literature suggested are important to language and writing skill ability, including: amount of language skill and grammar instruction prior to taking the LST, the grade students reported earning on their most recent college-level English composition, the extent to which they like to write, their language and writing self-efficacy, and their grade on the LST. Most of the measures are based on 5-point, Likert-like intensity scales with responses ranging from strongly disagree to strongly agree. For all measures, responses were coded so that the higher the measure, the more positive the response to the concept (e.g., higher measures reflect more language instruction). Demographic measures of gender and ethnicity were also included as control variables. The amount of language skills instruction was measured by two statements (r = .317, p < .001): "Overall, I received enough language skills instruction in high school and college to be prepared to pass the LST," and "Identify the last the last year in high school you studied English grammar" (response options were: I didn't' study grammar in high school, 9th, 10th, 11th, 12th grades). Students were asked to self-report their grades on the most recent English composition course they took at the college or university. One statement � "I like to write" -- measured their attitudes toward writing. The self-efficacy construct was measured by three statements (Cronbach's alpha = .63): "I find the rules of grammar and usage confusing" (responses to this statement were recoded so that agreement reflected higher self-efficacy), "I consider my English language skills to be:", and "I consider my writing skills to be:". On these last two statements students were provided five options ranging from poor to excellent. This measure of self-efficacy borrows one statement (I find the rules of grammar and usage confusing) from Bissell and Collins (2001) and Collins and Bissell (2002). These authors used different measures of self-efficacy in their two studies; the first study used seven statements, the second used five. In both cases, the self-efficacy measure used by these researchers was created to reflect skills beyond language and writing (for example, in one study they asked students whether they perceived themselves to be good students, and in the other study asked students about their abilities to remember details and thrive under pressure). This researcher thought these types of statements measured issues broader than language and writing self-efficacy, so they were discarded for the current study. The students' grades on the LST were recorded on the survey by the JMC college adviser after the tests results were calculated. The LST is an 80-question exam with sections on word usage, spelling, identifying parts of speech and parts of a sentence, punctuation and grammar. A score of 70 percent (56 correct answers) is required to pass the exam. Students receive three opportunities to take the exam. If they don't pass on their second effort, students are advised to take the JMC college's remedial course in language skills prior to taking the exam the third time. Results During the period the survey was administered, a total of 313 students took the LST. Of these, 225 completed the survey, a response rate of 71.9 percent. More than a third (34.6 percent) reported having "clepped" out of one English composition course (placed out based on high school advanced placement courses or a diagnostic exam given through the English department), and 18.6 percent said they clepped out of both English composition courses required by the university. Of those who took English composition in college, 64.4 percent said they made "A" on their most recent composition course, and 26.8 percent reported making "B"; together, this means that more than 9 of 10 students (91.2 percent) taking the LST reported earning at least a "B" in their most recent English composition course. The mean score on the exam for all respondents was 57.59 (56 is the minimum passing score); there were no significant differences based on attempts at the exam (first attempt mean = 57.14; second attempt mean = 56.48; third attempt mean = 56.58). Of all students taking the exam (regardless of attempt), 60.3 percent passed the exam, and 39.7 percent failed. By gender, 70.3 percent of respondents are women, and 29. 7 percent are men. Ethnicity data show that 2.8 percent of the respondents are African American, 1.4 percent are Asian American, 5.5 percent are Hispanic, 8.3 percent are Native American, 79.4 percent are white, and 2.8 percent indicated international student or other. Because the number of responding students of many ethnicities is small, the ethnicity variable was recoded for statistical analysis into minority and white. Differences by gender and ethnicity are discussed below. The first research question asked the amount of language skill instruction students had prior to seeking entrance to the JMC college, and whether they perceived this amount of instruction was adequate to prepare them to pass the LST. As a partial answer to this question, students were asked to indicate the last year in high school that they studied English grammar. Approximately two of five students reported they studied grammar as seniors; however approximately one of eight said they didn't study any grammar in high school (See Table 1). _______________________________________________________________________ Table 1: Last year in high school that students studied English grammar N Percentage Never studied grammar 28 12.4 Ninth grade 27 12.0 Tenth grade 33 14.7 Eleventh grade 41 18.2 Twelfth grade 96 42.7 Totals 225 100.0 _______________________________________________________________________ Students were also asked if they thought they had received sufficient language instruction in high school and college to pass the LST. Their mean response (on a five-point scale where 1 = strongly agree, 2 = somewhat agree, 3 = neutral, 4 = somewhat agree, and 5 = strongly agree) was 3.44 (SD = 1.19). This response falls almost squarely between neutral and somewhat agree. In other words, this response suggests that students show tentative agreement that the totality of their language skill instruction (in both high school and college) was sufficient for them to pass the test. Those students who said they took college English composition courses were asked one additional question about how much grammar instruction was in these courses. Their response options ranged from 1 = none, 2 = very little, 3 = some, 4 = quite a lot, and 5 = a great deal. The mean response to this statement was 2.27 (SD = .90), which means that students who took composition in college remember receiving slightly more than "very little" grammar instruction. These findings support much of the previous research that suggests many high school students do not receive grammar instruction in their later high school years, and at the college level there is very little emphasis on grammar in English composition courses. Perhaps most interesting, though, is that despite this relative dearth of instruction, students show slight agreement that they've received enough language skill and grammar instruction to prepare them to pass the college's required language skills exam. The second research question asked what is the level of the students' perceived language skill self-efficacy? To answer this question, the responses to the three statements used to create the self-efficacy construct were averaged to create a self-efficacy quotient (responses were coded so that the higher the measure, the higher the self-efficacy). This self-efficacy quotient is 3.50 (See Table 2). This result suggests that students see their overall language and writing skills as between "average" and "above average." However, a look at the individual statements in the construct provides additional insights. Students are much less confident in their grammar and usage than in their language and writing skills. This suggests a couple things. First, students tend to see language skills as something other than grammar and usage, as they clearly rate themselves as more confident on "language skills" than grammar and usage. Second, despite their uncertainty about grammar and usage, students rate their writing skills as close to "above average," the highest of the three statements measures. This, too, suggests that students distinguish their perceptions of their writing skills from their grammar and usage skills. The data suggest that students perceive they are above average writers, even when they aren't so sure of their grammar and usage skills. _______________________________________________________________________ Table 2: Student perceptions of their language and writing skill self-efficacy N Mean SD Self-efficacy quotient 224 3.50 .66 Statements I find the rules of grammar 225 3.04 1.10 and usage confusing *I consider my English language 224 3.69 .74 skills to be: *I consider my writing skills 225 3.85 .70 to be: *(these statements used the following scale: 1 = poor, 2 = below average, 3 = average, 4 = above average, 5 = excellent) _______________________________________________________________________ Research question three asked whether there is a significant relationship between the amount of language skill instruction and perceived language and writing skill self-efficacy. Because there is a substantial body of research suggesting that the more instruction one receives on a subject, the more confident one should be in that subject area, a hypothesis was posed that predicted students who report greater amounts of language and writing skill instruction will have higher perceived self-efficacy. This hypothesis was tested by a correlation between amount of instruction and self-efficacy. The correlation between the constructs is positive and significant (r = .339, p < .001). The hypothesis is supported. The data reveal that as language and writing instruction increases, levels of self-efficacy increase as well, and those students who report lower amounts of instruction also have lower language and writing skill self-efficacy. The fourth research question asked whether there is a significant relationship between amount of language skill instruction and grades on the JMC school's required Language Skills Test. The correlation of the two constructs is .132 (p = .048), which is significant, but not strong, as it explains less than 2 percent of the variance between the variables (r2 = 1.742). The data indicate that as the amount of language skill instruction increases, LST grades increase, and less instruction is related to lower LST grades. However, given the relatively low correlation, there are other variables that need to be considered. Toward this end, research question five asked what variables are the strongest predictors of success on the JMC school's required Language Skills Test. To answer this question, a regression model was built, with the independent variables ethnicity, gender, self-efficacy, liking to write, amount of instruction, and grade on most recent college-level English composition course. The dependent variable is LST score. The research suggests these hypotheses: The amount of language skill instruction will be a significant predictor of the score on the JMC school's required Language Skills Test, and language and writing skill self-efficacy will be a significant predictor of the score on the JMC school's required Language Skills Test. In constructing the regression model, the researcher used the enter method, controlling first for ethnicity and gender. As the data indicate, when all the independent variables are considered, the significant predictors of passing the LST are self-efficacy and liking to write (See Table 3). Hypothesis two predicted amount of instruction would be a significant predictor of the LST score. This hypothesis was not supported (standardized beta = .056, sig. = .417). Hypothesis three predicted that self-efficacy would significantly predict the LST score. This hypothesis was supported (standardized beta = .290, sig. = .000), as self-efficacy was the strongest predictor of score on the LST of all the independent variables, explaining 13.2 percent of the variance. _______________________________________________________________________ Table 3: Regression Model , standarized beta weights and adjusted R2 on LST score Block 1 Block 2 Block 3 Block 4 Block 5 Stand. Stand. Stand. Stand. Stand. Beta Beta Beta Beta Beta Variable Ethnicity -.135* -.078 -.087 -.090 -.092 Gender -.086 -.084 -.085 -.092 -.095 Self-efficacy .371** .319** .303** .290** Liking to write .140* .143* .153* Amount of .043 .056 Instruction Eng Comp grade .095 Cumulative .015 .147 .157 .160 .162 adjusted r square _______________________________________________________________________ Discussion This study finds, much like previous studies, that students perceive their writing skills to be relatively strong, but they aren't so sure about their grammar and usage skills. Despite these reservations about grammar and usage skills, students indicate mild agreement that they had sufficient preparation (in high and college) to pass the language skills test that is required for entrance into the first media writing course. In terms of high school instruction, approximately 60 percent of the pre-JMC students did not study grammar in their senior year of high school, and one in eight students say they did not study any grammar in high school. This study also supports previous studies that found students do not think they are getting much grammar and usage (basic language skills) in college English composition. This means for a clear majority of pre-JMC students that by the time they have fulfilled the JMC college's general education requirements and are ready to take the LST (usually their sophomore year), it has been at least two years, and in most cases three, since students have had substantial instruction in grammar. Perhaps these findings begin to explain why students don't feel as confident in their grammar and usage skills as their writing skills. The data also indicate that if pre-JMC students don't get sufficient language skills instruction in high school, they aren't going to get it in college before they begin JMC classes. This finding suggests that high school English and journalism teachers serve their students well when they make language skills and grammar an important part of the high school curricula. Several studies have found that students entering JMC programs have unrealistic evaluations of their language and writing ability. This study produces data that both support and refute this notion. In perceptions of their ability, students clearly distinguish their grammar and usage skills from their language skills and writing ability. In general, respondents in this study say they like to write (mean = 4.25 on a scale of 1 to 5, where 5 = strongly agree), have above average language skills and above average writing ability. However, they aren't so sure of their grammar and usage skills. In other words, the pre-JMC students in this study don't think their lack of grammar and usage skills holds them back as writers or is likely to prevent them from being successful on the LST. This seems to support those studies that have found students might not be very good judges of their writing. Considered another way, 64.4 percent of the students in this study reported making "A" on their most recent English composition course (91.2 percent said they made at least "B"), and these same students reported receiving "very little" grammar instruction in English composition. The lesson students seem to take from their composition experience is that they are good to excellent writers, even if they find the rules of grammar and language confusing. However, self-efficacy was the strongest predictor of scores on the LST, which seems to suggest that students' perceptions of their language and writing skills might be more accurate than many JMC faculty and researchers think. This finding seems to illustrate a core assumption of Bandura's work in self-efficacy. Bandura (1986) asserts that self-efficacy in a subject or skill area mediates the effects of actual skills because self-efficacy increases effort, persistence and perseverance. Students with higher self-efficacy have more confidence they can succeed at a task, so they work harder at it, which generally leads to stronger performance, which in turn increases levels of self-efficacy. This type of self-fulfilling prophecy is supported by the data in this study, which finds self-efficacy a stronger predictor of LST performance than the amount of language instruction students report having or their grade on their most recent English composition course. This finding might also begin to explain why Hypothesis 2 -- The amount of language skill instruction would be a significant predictor of LST score -- was not supported. The data show a significant correlation between amount of instruction and self-efficacy (r = .339, p < .001), and the amount of instruction and score on the LST (r = .132, p = .048). Yet, when the more robust regression test looked at the relative impact of all the independent variables on LST score, it was self-efficacy -- not the amount of instruction -- that predicts LST score. These are interesting results. Students are getting their self-efficacy from somewhere, and although one can not claim causation based on tests of correlation, there is a moderately strong relationship between the amount of instruction they receive and their self-efficacy. Attribution theory would suggest that students attribute their success and skill to their internal capabilities (their skill) more than external causes (the instruction they have received). Although this study did not attempt to test attribution theory per se, the findings suggest that self-efficacy is a complex process, and its sources or causes might not be easily -- or uniformly -- identified. There are several shortcomings that should be discussed. First, this study -- based on a sample of students from one university -- can not be generalized. Although, anecdotally, it appears the demographic variables of ethnicity (approximately 80 percent white, 20 percent minority) and gender (approximately 70 percent female, 30 percent male) are quite close to those of students entering JMC programs. Second, the timing of administering the survey might have impacted some responses. Students were asked to complete the survey immediately after completing the LST. Their perceptions on how they fared on the exam might have impacted their responses to statements on the survey that asked them to assess their skill levels. However, it appears the most important issue is a lack of tested measures. The self-efficacy measure in this study was different from that used in earlier studies, but the two most recent JMC-based studies that considered self-efficacy used two different measures of self-efficacy in them. Self-efficacy appears to be an important variable in language skill competency and fertile ground for future studies, but there needs to be more agreement on how to measure it. In this study, the focus was on language skills (grammar, usage and writing), but in the Collins and Bissell studies, they are studying broader concepts of self-efficacy (including ability to thrive under pressure and make deadlines). Future research should consider refining a self-efficacy measure as it relates to language and writing skills. Conclusion This study adds a modest amount to the work on self-efficacy as a predictor of language and writing skills. In this way, it contributes to a small, but growing, body of research on the impact of self-perceptions on writing skills and outcomes on standardized language tests. More important to scholastic educators is that the findings in this study suggest that they do their students a service if they make a considerable effort to teach the basics of the English language. This study finds that students are not getting much grammar and basic language instruction in their college composition courses. This study also finds that students enter their JMC programs saying they like to write, and their moderately high level of self-efficacy suggests they are motivated and willing to work at improving their language and writing skills. This suggests that a challenge for high school instructors is to strike a balance -- one that provides a stronger foundation in grammar and language while maintaining students' enthusiasm and affinity for writing. References Bandura, A. (1978, April). The self system of reciprocal determinism. American Psychologist, 33, 344-358. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall. Bissell, K., & Collins, S. (2001, Summer). Early predictors of ability in a news writing course. Journalism & Mass Communication Educator, 56 (2), 69-80. Collins, S., & Bissell, K. (2002, Winter). Student self-efficacy in a media writing course. Journalism & Mass Communication Educator, 57 (4), 19-35. Dvorak, J. (1990, Spring). College students evaluate their scholastic journalism courses. Journalism Educator, 45 (1), 36-46. Geimann, S. (2000, August). A grade for the graduate. Quill, 38-39. John, J.A., Ruminski, H., & Hanks, W. (1991, Autumn). Trends in writing skills tests for admissions to programs. Journalism Educator, 46 (3), 44-49. Kees, B. (1996, December). Do we need journalists or generalists? The Education Digest, 62, 65-69. Masse, M. (1999, Autumn). Evaluating students' progress by reading their journal. Journalism & Mass Communication Educator, 54 (3), 43-56. Olson, L., & Dickson, T. (1995, Summer). English composition courses as preparation for news writing. Journalism & Mass Communication Educator, 50 (2), 47-54. Oregon Report: Project on the future of journalism and mass communication education. Planning for curricular change in journalism education (revised edition). Eugene: University of Oregon. Oukrup, C., & Brown, L. (1977, August). A brief report on what J schools are doing about language skills. Paper presented to the Association for Education in Journalism, Madison, Wisc. Oukrup, C., Brown, L., & Parsons, P. (1998, Summer). Language exams and entrance requirements: A 20-year update. Journalism & Mass communication Educator, 53 (2), 40-44. Plopper, B., & Rollberg, J. (1996). Assessment and student outcomes in entry-level news writing courses: Another look. Southwestern Mass Communication Journal, 12 (1), 1-9. Ryan, M., & Findley, P. (1982, Spring). J-schools can help by monitoring H.S. language programs. Journalism Educator, 36 (1), 17-19. Seamon, M. (2001, Winter). Assessing the need for change in J-school grammar curricula. Journalism & Mass Communication Educator, 55 (4), 60-69. Smith, M. A. (1997, January). Perceptions of quality in journalism and communications education: A delphi study. Journal of the Association of Communication Administration, 32-50. Stone, G. (1990). Measurement of excellence in newspaper writing courses. Journalism Educator, 44 (4), 4-19. Urban, C.D. (1998). Examining our credibility: A report for the American Society of Newspaper Editors Credibility Project. Reston, Va.: ASNE Foundation. Williams, H.L. (1983, Winter). Which are the toughest j-courses to teach and why? Journalism Educator, 37 (4), 12-14. Williams, J. (1985). Style: Ten lessons in clarity and grace. (2nd ed.) Glenview, Ill.: Scott, Foresman, and Co.