Writing, is a complex process and competent writing is frequently accepted as being the last language skill to be acquired. Al-Mekhlafi, 2011 Due to its complex process, writing is regarded as a challenging task not only for native speakers of a language but also for second language learners. As it is the case for any other acquired skill, writing skill also evolves continuously until the product, a piece of writing, meets the expectations of both the writer and reader. However, when it comes to learning how to apply certain conventions of writing in a formal learning environment, the role and responsibility of the learner becomes even more important.
As cited in Sheila Matoti and Almon Shumba (2011), writing problems identified by previous studies are grammar, spelling, punctuation, expression, and the ability to explain, structure and interpret facts as well as vocabulary and referencing. Although it can be claimed that the problems listed above can be fixed, to a certain extent, through formal instruction and learner’s effort, it is a fact that having some sort of knowledge about one’s capabilities and skills can indeed play an important role in motivating human behaviours. (Bandura,1986)
According to Bandura’s Social Cognitive Theory, “individuals possess a system of self-beliefs that enables them to exercise control over their thoughts, feelings, and actions.” (as cited in Garcia &Fidalgo,2008, Rahimpour & Jahan, 2010, Al-Mekhlafi, 2011, Shah et al. 2011, Galligan ,2011, Alexander & Schmidt, 2012). Broaddus (2012) explained that social cognitive theory forms “the roots of both writing self-efficacy and motivation both of which change from course to course, depending on students’ interest, efficacy for performing tasks in the course, and other social and environmental factors.” In other words, learners may create their own path of learning through engaging themselves in tasks. Therefore, students set their goals for learning a specific subject.
Self-efficacy is “a construct proposed by Bandura (1977) and has been tested in a variety of fields, focusing on a variety of topics. Broaddus (2012) and Bong (1999) define self-efficacy as people’s convictions about their own capability for successfully executing a course of action that leads to a desired outcome. Schmidt and Alexander (2012) argued that self – efficacy is not only a current but also future disposition requiring learners or participants in a research to evaluate their beliefs in their future abilities. Therefore, self-efficacy requires self-evaluation, about what one can do and what can be improved. As well as self-evaluation, there are two more factors affecting self-efficacy: modeling (coaching) such as coding the information for retention, being capable of producing the presented patterns, and maintaining a certain level of motivation; goal setting, such as acquiring skills and knowledge or making good grades. (Schunk, 2003).
When considered specifically for the writing ability, writing self-efficacy translates into a strong sense of confidence for the task of writing.( Pajares & Valiante,1997). In other words, once learners have sufficient self-belief in their ability to write in their second language, their interest in writing may increase, they may display constant efforts, and great eagerness and resiliency when they are working on a writing task. In his study conducted in 2012, Broaddus concluded that journalism course students who hold high self-efficacy for grammar have put more effort into their time management and environmental resources while studying and it was found that these high efficacy attitude towards grammar resulted in better grammar skills. This was interpreted as these students determine where the challenge or obstacle related to the task lies and they develop strategies so as to succeed in their writing assignments.
Garcia and Fidalgo stated in their 2008 study that “in difficult tasks such as writing which includes many recursively employed cognitive processes, self-efficacy plays a the key role.” While acquiring writing skills, students need to use their “cognitive, behavioral and motivational engagement and learning.”
Recently, “an enormous amount of research has been done on the writing composition proceses that student writers undertake”(Erkan & Saban, 2011). Although writing instruction has gone through numerous changes, according to an action research carried out by Yavuz and Genc ( 1998), students tend to display negative attitudes towards writing. Yavuz and Genc (1998) concluded their research findings as, for students writing is an obligation in order to pass final exam. Erkan and Saban (2004) suggest that this attitude can be interpreted as students’ apprehension and lack of self-efficacy in writing.
The reason for students’ apprehension and lack of self efficacy may be due to limited time allocated for writing practice in classes. As cited in Leki and Carson (1994), according to the results of structured interviews of a study, university students who completed language courses in intensive English program at five different university reported that they only spent 10 % of their time on writing assignments across the curriculum. However, it should be noted that in order to succeed in their academic studies, second language learners, especially the students at universities are required to gain a certain ability to write well since writing is “integral to academic success.”Although the time spared for writing practice may not always be possible, “the use of a variety of writing assignments, with clear, concise evalution criteria, could help student writers in the development of their self-efficacy while dealing with a writing task.” (Broaddus, 2012)
When examined specifically, Turkish students practising the conventions of writing in English, “even students who are proficient at other language skills share the same problems with writing: they are afraid to make writing errors; they lack self-efficacy in writing.”(Erkan & Saban, 2011)In her study, Erkan concluded , similar to previous research findings, it was confirmed that the degree of students’ negative attitude, writing apprehension, does affect Turkish Tertiary level students’ writing performance. In other words, the higher the apprehension, the weaker the writing skills / performance. This aspect of Erkan’s study, one that specifically looks at relationships between writing apprehension and writing performance, is an expansion in the research area since the number of studies conducted to investigate second language learners’ wiriting apprehension is very few.
However, as cited in Garcia and Fidalgo (2008), Bandura maintained that misjudgement in one’s beliefs about potential performance may be detrimental, which is proved by some research that it can be prevalent in writing classes.
Although there is a great number of research examining the relationship between self-efficacy,to the best knowledge of the researchers Rahimpour and Jahan, there is no research on relationship between general self-efficacy and writing (Rahimpour and Jahan, 2010). Rahimpour and Jahan claimed that all the previous research mainly focused on the investigation of self-efficacy among groups with a very limited range of proficiency. Therefore, Rahimpour and Jahan determined their research area as ” the impact of self-efficacy and proficiency on EFL learners’ written task performance regarding load, fluency, complexity, and accuracy. The participants chosen were low-proficiency and high-proficiency learners of English aged between 18-25. In addition to completing three separate tasks, the participants were asked to fill out the General Perceived Self-Efficacy questionnaire, through which the participants’ level of confidence was aimed to rate. The results indicated that there was a significant relationship between self-efficacy and personal tasks in terms of concept load in high proficieny participants; however, there was not a significant relationship when it came to fluency, complexity, and accuracy only for low but also high proficieny participants’ performance.
Another area of research on self-efficacy is related to help seeking which is recognized as a component of self-efficacy. Help seeking is comprised of two types of behaviour: adaptive and nonadaptive behaviours. (as cited in Williams & Takaku, 2011) Various studies yielded significant correlations between self-efficacy and help seeking. Williams and Takaku cited that while high self-efficacy students tend to express high help seeking behaviour, low self-efficacy students, despite being under similar circumstances, less willing to seek help. It is commonly accepted by researchers in the field that just as help seeking, self-efficacy belief has been found to be a predictor of academic success. In addition to their teachers, most students may seek help at writing centers’ remedial courses in the institutions they attend. Writing centers have been proved to be helpful when it comes to improving students’ writing performance.
Williams and Takaku (2011) hypothesized that writing self-efficacy would influence writing performance. The participants of the study were both native and nonnative university students who attended a one semester intensive writing course for all content areas. In contrast to the researchers anticipation, the analysis showed that the frequent visitors to writing center in order to receive professional assistance from the writng center staff, managed to get higher grades than the ones who did not. What makes the results of the study more noteworthy is the fact that these results apply for both ESL learners and native-English- speakers. When the results are generalized, it can be argued that when appropriate assistance is provided, every student practising the conventions of writing can improve himself or herself.
A study carried out ,by Shah,Wan Mahmoud, Din,Yusof, Pardi in 2011 with Malaysian secondary school learners in order to determine the students’ self-efficacy in writing. The results indicated that there is ,indeed, a large, significant positive correlation between self-efficacy and writing performance. That is, it was,again, confirmed that learners with high self-efficacy tend to “pursue opportunities to write, put more effort into writing, and be more persistent in seeking writing competence.”
Another research conducted by Early and De Costa-Smith, in 2011, confirmed Bandura’s social cognitive theory. To put it differently, self-efficacy beliefs can be developed by mastery experience, social models and lived experiences, and social persuasion. The impact of multiple opportunities to practice any genre, along with peer and teacher feedback and interaction during the intervention may play a key role on the increase in students’ self-efficacy beliefs related to the genre. All in all, “while strong efficacy leads to success, overly high or low efficacy can undermine the proper development of cognitive, social, emotional, and behavioral skills.” (Broaddus, 2012)
On the other hand, Igo (2002) asserted that the students overestimated their self-efficacy because of several reasons, namely, high school teachers’ failure to provide appropiate and correctional feedback, the teachers’ negligence of promoting the importance of producing good work, and teachers’ overappraisal of mere student participation in the writing process. Conversely, when individuals see that another person, in our context a classmate, performing a desired task, self efficacy believes are positively affected. To pplace this into the context of university English preparatory school classrooms, when a student witnesses that a friend of his or her succeeds in writing a well-organized essay, he or she may be more willing to try writing an essay on his own. As Broaddus (2012) cited “observers’ self-efficacy beliefs would be reduced” if they see their peers repeatedly fail to accomplish a required task.
Whether high or low, self-efficacy in writing has a crucial role in students’ writing performance. Broaddus (2012) cites from Bandura (1997) and draws our attention to environmental context and says that it often affects efficacy. In a formal language learning environment where there is only limited time can be allocated to writing, it is an undeniable fact that determining students with low self-efficacy and assisting these students while they are developing their writing skills is highly important. Instructors should not only provide the standard techniques of writing and provide corrective feedback on vocabulary and grammar but also provide one-to-one assistance whenever their students seek help. The amount of assistance provided by teachers should also have limits. As cited in Broaddus (2012), Bandura (1997) says “Successes achieved with external assistance carry little efficacy value because they are likely to be credited to external aids rather than to personal capabilities”
Participants were 20 Turkish learners of English at a preparatory school of an English medium university, Turkey. The participants initially took a standard language proficiency test prepared by the preparatory school testing office. The proficiency test consisted of 80 four-option multiple choice questions before the semester started. The multiple choice items aimed to test the students’ level of grammar, vocabulary, and reading comprehension abilities, The passing grade for the proficiency test was determined as 60.The students who got lower than 60 took the placement test consisting of 80 questions. The students were also required to write an essay on a topic provided by the testing office after completing the multiple choice section. According to the placement test scores, the students were placed in three different levels-A, B, and C levels. The intermediate level students were placed in A level classes. In other words, the participants can understand and speak English with some confidence and they have the grammar and the vocabulary to talk and read about a wide number of subjects. Moreover, they can assess and consciously improve their own pronunciation. They have studied all the main tenses and can confidently make sentences, question forms and clauses in all of them. However, intermediate students are only in the beginning phase of improving their use of phrasal verbs and modal verbs. On the other hand, intermediate level is equivalent to IELTS score 4.5 – 5.5, TOEFL IBT score 55, TOEIC score 450 – 650, Cambridge PET.
Out of 20 participants, there were only 3 students who were private high school graduates. In Turkey, private high schools’ English curriculum is more intense when compared to state schools English curriculum. Both males and females were involved in the study. They ranged in age from 18 to 20.
The questionnaire administered in the study was developed for the purpose of assessing international students’ in Marquette University, U.S.A. Marquette University is a private university in Milwaukee, U.S.A. The development of the questionnaire was one of the four assessment areas that Marquette university aimed to assess the four advanced ESL academic bridge courses, one of which focused on international students’ efficacy in their writing performance. The original questionnaire is a 36-item self-assessment scale to grade the strength of participants’ belief, or self-efficacy, in their writing performance. However, the questionnaire used to investigate the relationship between students’ self-report on their writing efficacy and their writing performance is a 20-item questionnaire. The reason why the original questionnaire was not used in the current study was the fact that some of the items did not serve the purpose of the study. The adopted writing efficacy questionnaire features 20 items with positive polarity (I canaˆ¦).The instrument of 20 – item questionnaire was used to assess students’ self- efficacy beliefs related to the skills needed in different essay organizations and other factors that determine students’ writing performance as successful or unsuccessful. Some of the investigated factors through the questionnaire are brainstorming, outlining, as well as their self- efficacy in terms of grammar, vocabulary, and spelling. Level of efficacy in writing ability was measured on a five – point- likert scale included the responses (usually not true of me, somewhat true of me, usually true of me, always or almost always true of me). The questionnaire will be used as the independent variable when the relationship between self-efficacy and writing scores is investigated.
Questionnaire is distributed to participants at the beginning of the semester as a self-completing questioonnaire. Since they were not received any writing instructions before the semester, it is assumed that this questionnaire served to measure their self-efficacy without any intervention of teacher. During the semester, they learned basics of essay writing such as paragraph organization, thesis statement and some grammar backup.
Quiz scores which will be used as the dependent variable in my essay are based on the Second Quiz, took after XXXXXXX which constitutes 20% percent of the midterm grade.
Self-Efficacy of Participants
Coefficient of Variation
clear thesis statement
using my independent thinking
support /develop main point
spell, capital, punctuation
a good conclusion
wording, gram., pun,spelling
outline to logically organize
a good introduction
vocab/ word forms
effectively write under time constraints
revise to improve dev.& org.
write quickly in English
different types of organization
academic style &tone
Above table shows descriptive statistics of items used in the questionnaire. Since the questionnaire is composed of Likert scale questions, 1 stands for the lowest score and 5 is the highest possible score. According to this table, participants’ self-efficacy is relatively high in items like “topic sentence”, “clear thesis statement”, “using my independent thinking”, “effectively brainstorm” and “support/develop main point” where average scores are above 3,50 over 5. Their weakest points are stated as “academic style&tone”, “different types of organization”, “paragraph” and “write quickly in English”, where averages are below 3 over 5. Generally speaking, participants seem to be confident in idea development and they feel themselves relatively weaker in the structural aspects such as academic style and tone.
Coefficient of variations, calculated by dividing standard deviation to mean for each items show how responses are dispersed in each item. Highest dispersion is observed on item19, “write quickly in English”” and on item12, which measures self-perception about the ability to use “different types of organization”, both coefficient of variations are above 0,34. Lowest scores are observed on these items “topic sentence”, “support/develop main point”, “summary”, “vocab/word forms” and “paragraph”, all of them are 0,17.
Descriptive statistics give some hints about overall level of self-efficacy of participants, and it is almost sure that this is not something uni-dimensional. However a multivariate analysis is beyond the limits of this essay and number of cases is very small to conduct further analyses. Consequently remaining parts of this essay will be based on the assumption that self-efficacy is uni-dimensional, something possible to measure in a single dimension.
Efficacy scores of respondents are calculated on their raw scores obtained by using above described self-efficacy questionnaire. Total efficacy score is calculated by using a simple summation of responses. Theoretical maximum score is 100 showing the highest self-efficacy score, and minimum theoretical score is 20, indicating that respondent stated him/herself on the lowest end in each question.
Table . Descriptives of Self Efficacy Scores
Descriptive statistics are presented above. Mean self-efficacy score is 66,15 with a standard deviation of 7,34. Coefficient of variation (V) is 0,11, indicating a low level of variation. Minimum score is 53, which is above theoretical minimum and observed maximum 86 is below theoretical maximum. These findings show that scores are located between 60 and 90.
In order to facilitate interpretation of analyses, all raw scores are transformed to standardized (Z) scores. Newly created self-efficacy scores have a mean of 0 and standard deviation of 1. Distribution of this new scores are presented in the below histogram. Although number of participants is limited (we have only 20 cases), this distribution is very close to be normally distributed.
Figure . Histogram of Standardized Self Efficacy Scores
Measurement of Academic Performance: Quiz Scores
It is so far argued that quiz scores are used to measure academic performance of participants. Quiz scores are graded over 100.
Table . Descriptives of Quiz Scores
Cause and Effect Quiz Scores
Above table shows that average quiz score is 60, and standard deviation is 7,78. If we consider that range between minimum and maximum scores is only 31, it is possible to conclude that quiz scores are relatively less dispersed.
Our expectation is that academic performance of students is correlated with their self efficacy. As one students is more confident about his/her efficacy, it is expected that he/she performs better in courses; and as he/she performas better academically; his/her self efficacy increases. Since direction of causality is not clear; correlation analysis is the best tool to present the nature of the relationship between self-efficacy and academic performance.
In our cases, self-efficacy of participants are measured by using a self completed questionnaire and their self efficacy scores are calculated based on their answers; meanwhile their quiz scores are accepted as indicators of their academic performance.
Table . Correlation Analysis
Compare and Contrast Quiz
Above table shows that correlation coefficient between self efficacy and compare and contrast quiz scores is 0.60. Correlation coefficient is a score located between -1 and 1, where -1 stands for perfect negative correlation and 1 stands for perfect positive correlation. As this coefficient comes closer to 0, it shows that power of relationship is relatively week, if it is 0, it means that there is no relationship between examined variables.
A correlation coefficient of 0.60 means that there is a positive correlation between self efficacy of participants and their academic performance, a finding not falsifying our expectations. If we emphasize on the strength of this relationship, 0.60 shows a more than moderate relationship between our two variables.
This finding shows that there is a positive relationship between self-efficacy of participants and their academic performance. As one’s self efficacy increases, his/her academic performance also increases; and since correlation coefficient is a symmetrical measure; this finding also shows that as one’s academic performance increases; his/her self-efficacy also increases.
Although the measurement of self-efficacy precedes the measurement of academic performance; it is not sufficient to talk about a causal relationship between self-efficacy and academic performance; because self-efficacy of participants most probably affected by their previous academic performance, such as their performance in the high school or other exams they attended. Consequently accepting self-efficacy scores as an indicator of academic performance is a less risky choice compared to using them to forecast future academic success. In order to make such a forecasting, we need much more detailed measurement of academic performance and self-efficacy and higher number of respondents.
A further analysis is conducted to see whether there is a significant difference between students, based on their self-efficacy scores. Analysis of Variance (ANOVA) is preferred because it allows us to test statistical significance of an intervening factor on a dependent variable. We expect that students having high level of self-efficacy are more successful in academic terms. Above discussed measurement of self-efficacy is accepted as the independent variable and Quiz Scores are dependent.
Since self-efficacy is an interval level variable, recoding these scores to a nominal level variable is needed. It is preferred to use standardized self-efficacy scores to facilitate interpretation.
Standardized self-efficacy scores are recoded as follows: Those having a score less than 0 are recoded as “low efficacy” and those having a score less than or equal to 0 are accepted as “high efficacy”. 12 students are categorized as “low-efficacy” students and 8 are grouped as “high efficacy” students.
Table . Descriptives of Quiz Scores according to the Level of Self Efficacy
Low Self Efficacy
High Self Efficacy
Descriptive statistics of both groups are presented above. It is known that overall average of quiz scores is 60.05 with a standard deviation of 7.78. Students with low self-efficacy scores have an average of 57.33 and standard deviation of this group is 3.45. Meanwhile, students having high self-efficacy scores are also more successful in quizzes, their average is 64.13. At this point, we need to consider that standard deviation of these students is very high, 10.7; compared to other students.
Table . ANOVA Results
Low Self Efficacy
High Self Efficacy
Above presented results of ANOVA shows that difference between average quiz scores of low self-efficacy and high self-efficacy students is statistically significant. F score is 4.3 and this score is statistically significant at 0.05 level. Thus, we can falsify our hypothesis that group means are equal, meaning that students with high self-efficacy scores have higher scores in the Quiz, than students with low self-efficacy scores.
Similar to correlation analysis, ANOVA analysis does not allow us to make generalizations about the causal relationship between self-efficacy of students and their academic performance. However, it is clear that there is a statistically significant difference between these two groups of students. I believe that causal relationship has to elaborated by using more sophisticated data analysis tools and using these instruments with higher number of participants.
This essay targeted to discover whether there is a relationship between students’ perceptions about their efficacy in the English language and their academic performance. Literature review led us to think that there is a positive relationship between self-efficacy scores of students and their academic performance. There are a lot of reasons of this relationship discussed in details in literature review section of the essay.
Our hypothetical expectation is that as self-efficacy scores of students increase, their academic performance in writing also increases. Since this relationship is recursive, meaning that self- efficacy leads to better performance in exams and better performance may also lead to higher self-efficacy, causal relationship won’t be tested in this essay.
In order to measure self-efficacy of students, an adopted version of the questionnaire developed by the Marquette University has been used with 20 students at the beginning of semester. Original questionnaire was shortened to limit it with writing related items. By using these questionnaires, an index of self-efficacy is calculated and standardized. Meanwhile academic performance of students are measured with their grades in a Quiz held on XXXth week of the semester.
Correlation analysis showed that there is a positive correlation between self-efficacy of students and their academic performance. With a coefficient of 0.60, this relationship seems to be moderately powerful. As a result of restrictions of correlation analysis, it is not possible to understand direction of causality between these two variables.
My second analysis also showed that difference between mean quiz scores of students with high and low level of self-efficacy is statistically significant. Standardized self-efficacy scores are transformed to a binary variable. ANOVA showed that these two groups have different mean quiz scores and this difference is statistically significant.
While correlation analysis showed that there is a positive relationship between self-efficacy scores and academic performance; ANOVA results presented that students having high and low level self-efficacy scores have different quiz scores. Both findings indicate that self-efficacy scores matter.
It is so far argued that these findings cannot give any idea about direction of causality, since relationship between self-efficacy and academic performance is recursive. However, my analyses presented the existence of a relationship and it deserves further discussion.
It is possible to repeat this exercise with a larger sample and to prevent a possible falsification of hypotheses as a result of small sample size. Questionnaire revised to measure self-efficacy may be refined and developed to make a better measurement of this concept. Fu