R*vised Synthesis - R*search Skills

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=cher20

Higher Education Research & Development

ISSN: 0729-4360 (Print) 1469-8366 (Online) Journal homepage: https://www.tandfonline.com/loi/cher20

Graduate students as academic writers: writing anxiety, self-efficacy and emotional intelligence

Margarita Huerta, Patricia Goodson, Mina Beigi & Dominique Chlup

To cite this article: Margarita Huerta, Patricia Goodson, Mina Beigi & Dominique Chlup (2017) Graduate students as academic writers: writing anxiety, self-efficacy and emotional intelligence, Higher Education Research & Development, 36:4, 716-729, DOI: 10.1080/07294360.2016.1238881

To link to this article: https://doi.org/10.1080/07294360.2016.1238881

Published online: 04 Oct 2016.

Submit your article to this journal

Article views: 11521

View related articles

View Crossmark data

Citing articles: 26 View citing articles

Graduate students as academic writers: writing anxiety, self-efficacy and emotional intelligence Margarita Huertaa, Patricia Goodsonb, Mina Beigic and Dominique Chlupd

aDepartment of Educational and Clinical Studies, The University of Nevada, Las Vegas, NV, USA; bDepartment of Health & Kinesiology, Texas A&M University, College Station, TX, USA; cLiverpool Business School, Liverpool John Moores University, Liverpool, UK; dInspiring the Creative Within, LLC, College Station, TX, USA

ABSTRACT Researchers interested in psychological factors affecting writers in higher-education institutions, or academic writers, are concerned with internal variables affecting writing productivity; however few empirical studies explore these factors with samples of students who are in the process of earning master’s or doctoral degrees (i.e., graduate students). In this study, we examined writing anxiety, self-efficacy and emotional intelligence (EI) in a sample of graduate students at a large, research-intensive university in the United States. Using a survey, we collected measures on these variables in addition to demographic information from the participants. We then used the measures to descriptively compare groups of students with similar characteristics and to run three regression models to identify which variables best predicted writing anxiety. Our findings indicate self-efficacy is a statistically significant and large predictor of writing anxiety while EI is not, though descriptive data showed moderate effects between EI and first language (i.e., whether or not a student reported English as a first language). In the presence of self-efficacy, gender remained a significant predictor of writing anxiety, while first language did not. We discuss implications for future research and practice focused on helping graduate student academic writers succeed.

ARTICLE HISTORY Received 4 January 2016 Accepted 11 July 2016

KEYWORDS Academic writing; emotional intelligence; graduate students; self-efficacy; writing anxiety

Introduction

Based on global trends, only 1.6% of students are expected to complete an advanced research program, such as a doctoral degree (OECD, 2014). The low percentage of advanced degree recipients may be due to many different factors. However, writing is one known barrier for individuals aspiring to a master’s or a doctoral degree (from here on referred to as graduate students). For example, nearly 50% of graduate students pursuing doctoral degrees in the United States leave the university without completing their degrees, dropping out during the research proposal or dissertation-writing phases (Cassuto, 2013; Harris, 2011). Belcher (2009a) vividly describes her experience as a master’s degree student as follows: ‘When I started graduate school…my first quarter was tough… I began to suspect that everyone but me knew how to organize their time, do their research, and write successful

© 2016 HERDSA

CONTACT Margarita Huerta [email protected] Department of Educational and Clinical Studies, The University of Nevada, 4505 S. Maryland Parkway, Box 453014, Las Vegas, NV 89014-3014, USA

HIGHER EDUCATION RESEARCH & DEVELOPMENT, 2017 VOL. 36, NO. 4, 716–729 http://dx.doi.org/10.1080/07294360.2016.1238881

papers’ (p. 186). In a career where academic writing – that is, writing for academic purposes such as classroom assignments, theses or publications in academic journals – is so central to the evaluation of one’s success, understanding why graduate students struggle with writing and finding possible solutions to help students with their writing would benefit both the stu- dents and the institutions supporting them. Although studies have well documented the success of writing support groups at higher institutions in addition to providing frameworks for creating the writing groups (e.g., Boice, 1987; Murray & Newton, 2009; Murray & Thow, 2014), this study takes a different angle. Specifically, this study provides updated research on intrapersonal variables that previous scholars have noted to affect individuals’ academic writing such as writing anxiety, self-efficacy and emotional intelligence (EI; Boice & Johnson, 1984; Shao, Yu, & Ji, 2013; Zimmerman & Bandura, 1994). In doing so, we seek to inform current research and practice.

Framework

Writing anxiety

Writing anxiety can be defined as the manifestation of ‘feelings of tension, worried thoughts, and physical changes like increased blood pressure’ (American Psychological Association, n.d.) when a person is confronted with a writing task. Writing anxiety’s nega- tive impact on academic writing has been well documented in empirical studies in the United States. For example, writing anxiety was positively correlated with writer’s block amongst university teachers and negatively correlated with their writing productivity (Boice & Johnson, 1984). Writing anxiety also had a negative effect on graduate students’ writing in Bloom’s (1981) case study as well as in Onwuegbuzi’s (1997) quantitative study on graduate students taking research methodology courses. In the latter study, graduate students’ anxiety was directly related to the quality of their writing. In studies with under- graduate students taking writing-intensive classes, students with higher writing anxiety produced writing with lower quality and performed poorer on writing skills tests than stu- dents with lower writing anxiety (Daly, 1977, 1978; Daly & Miller, 1975). More recently, writing anxiety has been found to have a negative relationship on university students’ grades (Martinez, Kock, & Cass, 2011).

Self-efficacy

Self-efficacy is the perceived level of confidence in performing a given behavior (Bandura, 1997). Self-efficacy in writing can thus be defined as belief in one’s capability (or confi- dence) to write in a given situation. Self-efficacy is important for academic writing because the activity tends to be self-scheduled and performed alone. It also requires sus- tained creative effort and must undergo many revisions to reach publishable standards (Zimmerman & Bandura, 1994), thus requiring confidence in one’s self. Self-efficacy is correlated with writing achievement in school settings, according to several studies. For example, in studies with secondary and university students in the United States, students reporting higher self-efficacy had higher writing achievement (Pajares, 2003; Prat-Sala & Redford, 2012; Zimmerman & Bandura, 1994) and produced writing of higher quality (White & Bruning, 2005) than students who reported lower self-efficacy.

HIGHER EDUCATION RESEARCH & DEVELOPMENT 717

Emotional intelligence

EI is ‘the ability to monitor one’s own and others’ feelings and emotions, to discriminate among them, and to use this information to guide one’s thinking and actions’ (Salovey & Mayer, 1990, p. 189). According to several scholars, EI is a good predictor of academic performance (Libbrecht, Lievens, Carette, & Côté, 2014; Perera & DiGiacomo, 2013) and academic achievement (Parker, Summerfeldt, Hogan, & Majeski, 2004). EI also plays a role in students’ successful performance at the university level (Perera & DiGia- como, 2013). With respect to writing, Shao et al. (2013) recently found a positive relation- ship between Chinese university students’ EI and writing achievement in English classes offered by their university in China.

Writing anxiety, self-efficacy and EI

Empirical evidence supports the relationship between writing anxiety and self-efficacy among school-aged children, adolescents and university students (Klassen, 2002; Martinez et al., 2011; Matoti & Shumba, 2011; Pajares, 2007). For example, in a study conducted in the United States, university students who reported lower writing anxiety had higher self- efficacy than their peers who reported higher writing anxiety (Martinez et al., 2011). Not surprisingly, in a research synthesis on university business writing, Mascle (2013) argued the importance of fostering self-efficacy in order to lower students’ writing anxiety and increase their writing development.

As EI is the ability to assess, regulate and utilize emotions, and anxiety is an emotion, it is reasonable to assume that EI impacts anxiety. In the context of writing anxiety, no studies were found examining the relationship between EI and writing anxiety. Students reporting higher EI, however, report lower levels of language learning and communicative anxiety (Shao et al., 2013) and are more effective in managing their anxiety than students with lower EI (Dewaele, Petrides, & Furnham, 2008). We were therefore interested in examining the relationship between EI and writing anxiety in our study.

Research questions

The graduate students (i.e., master’s or doctoral students) in our sample participated in the POWERWriting Services (see the following description) at a large, research-intensive uni- versity in the United States. We sought to answer the following questions:

(1) What are the characteristics of graduate student writers who utilize the POWER Writing Services at a large, research-intensive university, including their writing anxiety, self-efficacy and EI?

(2) Are there significant differences between the writing anxiety, self-efficacy, and EI of the graduate students in the study in terms of gender, degree level (master’s or doc- toral), first language (speakers whose first language is English vs. speakers whose first language is not English), international status, and prior exposure to a writing service?

(3) How well do self-efficacy and EI predict writing anxiety in this sample of graduate students?

718 M. HUERTA ET AL.

POWER writing services

POWER is an acronym for ‘Promoting Outstanding Writing for Excellence in Research’. Created in 2007 at a large research-intensive university in the southeastern United States, the POWER Writing Services provide emotional and instrumental support for graduate students who wish to improve their academic writing and learn how to write more pro- ductively. As part of the services, students can sign up for a POWER writing studio and/or for the writing productivity class.

The POWER writing studio is offered to graduate students nearly every month of the academic semester, which typically runs for 15 weeks in the spring and fall in the United States. The studios are free, not for class credit, and have voluntary partici- pation. Students are able to sign up for the studios before the start of each month and the studio spots fill up quickly. Though the studios are designed for graduate stu- dents, occasionally a post-doctoral researcher or other academic member of the univer- sity enrolls. Participants and a facilitator meet once a week for two hours, during four consecutive weeks. In week 1, participants learn principles such as writing regularly, separating the generating of text from its editing, and ‘chunking’ large projects into manageable pieces. In week 2, participants learn how to write and read at the same time and how to efficiently organize a literature review. In week 3, they learn about different writing resources for writers (books, services, online tools). In week 4, partici- pants learn how to provide and solicit useful feedback, as well as why and how to estab- lish writing support groups.

Students can alternately sign up for a class (15 weeks a semester; 5 weeks in the summer; or 2 weeks in a May-mester, a short version of a semester) to learn the same prin- ciples. Students pay tuition, and receive academic credit for this class. Despite the varying lengths of the class for credit, the class is the same in its content, covering the topics in the writing studio and additional content such as writing theory, how to edit one’s own writing, and the process of writing academic articles for publication.

Method

Sampling and procedures

Participants who enrolled in the POWER studios or class were recruited for this study by receiving an email invitation, alongside a link to an online survey engine (Qualtrics). The invitation was emailed prior to the first studio or class meeting and participants were reminded of the survey during the first meeting. The university’s Institutional Review Board approved the study and data collection protocols. Student participation was volun- tary; student responses were anonymous.

Our final sample size included 174 graduate students enrolled in either a studio or class offered between Summer 2013 and Spring 2015. Of the 174 participants, 135 (77.6%) had enrolled in a studio and 36 (20.7%) had registered for the class. Three participants did not report whether they were enrolled in a studio or class. Detailed sample demographics are discussed under the ‘Results’ section.

HIGHER EDUCATION RESEARCH & DEVELOPMENT 719

Instrument and measures

The online survey instrument included 85 questions and took approximately 15 minutes to complete. The instrument included demographic questions (age, gender, race, first and second language when applicable, international status, country of origin), degree level (master’s or doctoral), academic department, area of study, and years in their department. To measure previous exposure to writing support services, we asked respondents whether they had ever utilized any university or POWER writing service resources. We also asked whether they were enrolled in a studio or in the class at the time of the survey, and assessed participants’ writing anxiety, self-efficacy, and EI.

Tomeasure writing anxiety, we used Daly andMiller’s (1975)Writing Anxiety Scale. The instrument includes 26 Likert-type scaled items (responses on a 5-point scale: 5 = strongly disagree) designed to quantify university students’ writing anxiety. The items contain state- ments about writing such as ‘I avoid writing,’ ‘I have no fear of my writing being evaluated,’ and ‘I enjoy writing.’ Because the instrument was developed in the context of assessing stu- dents enrolled in a composition class, we adapted the language to the context of graduate students. For example, we substituted the original term ‘essays’with ‘papers’ and the original phrase ‘composition class’ for ‘writing class’. Daly and Miller (1975) reported an internal consistency of .921 for the 26-item scale, using a split-half technique. The internal consist- ency for our graduate student sample was .928 (Cronbach’s alpha).

To measure students’ self-efficacy, we employed Zimmerman and Bandura’s (1994) Writing Self-Efficacy Scale. The original instrument includes 25 Likert-type scaled items. The items comprise statements about the students’ confidence to perform writing tasks such as eliciting suitable topics for writing in a short time, adjusting writing style to suit the audience’s needs, and finding ways to overcome being stuck on an assignment. Responses were offered on a scale of 1–7, where 1 = not able to do the task at all, and 7 = able to do the tasks very well. We adapted the language to the context of graduate studies. For example, we changed the original statement, ‘I can come up with memorable examples quickly to illustrate an important point,’ to ‘I can come up with examples from the reviewed literature, to illustrate an important point.’ In order to reflect the POWER Writing Services principles, we deleted items such as ‘I can construct a good opening sentence quickly’ and added items such as ‘I can protect my writing schedule/times,’ and, ‘I can obtain the appropriate feedback I need, during various stages of my writing project.’ Our final, adapted scale for self-efficacy included 26 items. Zimmerman and Bandura (1994) reported an internal consistency of .91 (Cron- bach’s alpha) for their 25-item scale, obtained from a sample of undergraduate students. For our sample of graduate students, we obtained an internal consistency score of .93 (Cronbach’s alpha) with a 26-item adapted scale.

To measure students’ EI, we selected items from the EI scale created by Schutte et al. (1998). The instrument was designed to quantify a person’s state of emotional development. The original instrument consisted of 33 items. From these, we discarded items focused on the ‘utilization of emotions’ dimension of EI – a dimension alluding to how one relates to others (e.g., ‘Other people find it easy to confide in me’ or ‘I present myself in a way that makes a good impression on others’). Our final scale consisted of 22 of the original 33 items. Schutte et al. (1998) reported an internal consistency of .90 for the 33-item scale. Our internal consistency was .89 for the 22-item scale (Cronbach’s alpha).

720 M. HUERTA ET AL.

Analysis

We assessed the missing data in our sample to determine patterns of missing-ness and found them to be missing at random (Buhi, Goodson, & Neilands, 2008). As the number of participants not completing the survey accounted for less than 10% of the total sample, we excluded these participants from further analyses. Because we wished to focus on graduate student data, we also eliminated the one post-doctoral student and the one faculty participant from our sample.

Using data from our final sample of 174 participants, we reverse-scored the writing anxiety and EI sores so the high values indicated high writing anxiety and high levels of EI. We used Daly and Miller’s (1975) formula for calculating writing anxiety (78 + Positive Scores –Negative Scores) and we reverse scored one negatively worded item on the EI scale before calculating the EI sum score.

Having established the data met all relevant assumptions related to multicollinearity, outliers, normality, linearity, homoscedasticity and independence of residuals (Tabach- nick & Fidell, 2013), we conducted multiple regression analyses to examine the relation- ship between writing anxiety, self-efficacy and EI, with writing anxiety as the dependent variable. We tested the fit of our proposed model, controlling for gender, degree level, first language, international status and prior exposure. In the analysis, variables were entered simultaneously in each of the models.

Results

Research question 1: What are the characteristics of graduate student writers who utilize the POWER Writing Services at a large, research-intensive university, including their writing anxiety, self-efficacy and emotional intelligence (EI)?

Between the Summer 2013 and Spring 2015 semesters, 194 participants responded to the survey. After deleting the surveys containing missing data and the one post-doctoral and one faculty respondent, the final sample comprised 174 respondents. Participants’ ages ranged from 20 to 54 (M = 30.8; SD = 6.9). Most participants were women (60.8%). Participants came from varied departments and colleges within the university.

A little over half the sample (52.9%) comprised international students, and 55.2% reported English was not their first language. Of the total sample, 83.3% reported not having participated in any other writing service prior to their studio or class enrollment; while 16.7% reported having either attended a studio and/or receiving help for their writing from other sources.

The theoretical midpoint is an artificial parameter to show how many of the sample scores fall above or below the mean of the instruments’ scale, allowing the reader to gauge the distribution of scores. For writing anxiety, participants’ mean score was 73.3 (SD = 16.5). The possible range of scores for the writing anxiety scale is 26 (low anxiety) to 130 (high anxiety); thus, our sample’s writing anxiety mean score fell below the scale’s theoretical midpoint (78.0). For self-efficacy, the mean score for the total sample was 97.3 (SD = 27.3). The possible range for the self-efficacy scale is 26 (low self-efficacy) to 182 (high self-efficacy); thus, our participants’ mean score fell below the scale’s theoretical midpoint (104). For EI, the sample’s average score was 83 (SD =

HIGHER EDUCATION RESEARCH & DEVELOPMENT 721

10.3). The possible range for the EI scale is 22 (low EI) to 110 (high EI), indicating the sample’s mean fell above the scale’s theoretical midpoint (66).

Research question 2: Are there significant differences between the writing anxiety, self-effi- cacy, and EI of the graduate students in the study in terms of gender, degree level (master’s or doctoral), first language (speakers whose first language is English vs. speakers whose first language is not English), international status, and prior exposure to a writing service?

Table 1 shows the results of a series of two-tailed t-tests assessing statistically significant differences between groups of students on their writing anxiety, self-efficacy, and EI.

As Table 1 illustrates, there were no statistically significant differences for gender or degree level among the dependent variables. Still, gender and degree level exhibited small to moderate effect sizes with respect to writing anxiety (For gender: Cohen’s d = .269; for degree level: Cohen’s d = .307). Females reported higher writing anxiety (M = 75.13; SD = 17.45) than males (M = 70.63; SD = 14.49), and master’s students reported higher writing anxiety (M = 77.25; SD = 15.36) than doctoral students (M = 71.76; SD = 16.77).

Writing anxiety and self-efficacy scores did differ significantly and by moderate-to- large amounts, for speakers of various languages. Participants who reported English was not their first language had higher writing anxiety (M = 76.06; SD = 14.89; Cohen’s d = .368) and lower self-efficacy (M = 89.2; SD = 25.60; Cohen’s d = .635) than participants for whom English was their first language (writing anxiety:M = 70.06; SD = 17.87; Cohen’s d = .368; self-efficacy:M = 106.50; SD = 26.64; Cohen’s d = .635). In addition, EI scores dif- fered by a moderate amount (albeit not statistically significant) between speakers and non- speakers of English as a native language (Cohen’s d = .305). Participants who reported English as their first language had higher EI scores (M = 84.74; SD = 10.70) than partici- pants who did not report English as their first language (M = 81.61; SD = 9.88).

Self-efficacy scores also exhibited a statistically significant difference and a moderate effect size, when non-international and international students were compared (Cohen’s d = .357). Specifically, non-international students had higher self-efficacy (M = 103.77; SD = 25.77) than international students (M = 91.63; SD = 27.44).

Last, though non-significant, a moderate effect size was observed in self-efficacy for stu- dents who had been exposed to writing services, prior to the survey, compared to those who had no exposure (p < .020; Cohen’s d = .357). Participants reporting prior exposure to a writing service had higher self-efficacy (M = 108.03; SD = 28.34) than participants who did not have prior exposure to a writing service (M = 95.2; SD = 26.65).

Research question 3: How well do self-efficacy and EI predict writing anxiety in this sample of graduate students?

To answer this question, we ran three multiple regression models as shown in Table 2 (with variables entered simultaneously in each model).

Model 1 explained 6.2% of the variance in writing anxiety (adjusted R2 = .062, F (5, 172) = 3.268, p < .008, Cohen’s f2 = .066). Language made the largest significant contribution to the model (β = .297; p < .031) followed by gender (β = .162; p < .037). Degree level also made a significant contribution to the model (β =−.153; p < .040). Participants who indi- cated either that English was not their first language, or that they were female, or that they were master’s students had higher writing anxiety scores. The other variables did not con- tribute significantly to Model 1.

722 M. HUERTA ET AL.

Table 1. Differences in writing anxiety, self-efficacy and emotional intelligence on selected independent variables. Writing anxiety Self-efficacy Emotional intelligence

Variable M (SD) p Cohen’s d M (SD) p Cohen’s d M (SD) p Cohen’s d

Gender .084 .269 .626 .074 .978 .004 Male 70.63 (14.49) 96.09 (26.80) 83.04 (9.95) Female 75.13 (17.45) 98.16 (27.67) 83.00 (10.64) Degree level .046 .307 .059 .290 .520 .098 Master’s 77.25 (15.36) 91.29 (20.10) 83.80 (10.45) Doctoral 71.76 (16.77) 99.86 (26.64) 82.69 (10.39) Language .017** .368 .001** .635 .047 .305 English as first language 70.06 (17.87) 106.50 (26.64) 84.74 (10.70) English not first language 76.06 (14.89) 89.92 (25.60) 81.61 (9.88) International status .167 .212 .003** .457 .335 .148 Non-international status 71.53 (17.29) 103.77 (25.77) 83.82 (11.22) International status 75.01 (15.71) 91.63 (27.44) 82.83 (9.50) Prior exposure .557 .089 .020 .357 .822 .034 Some prior exposure to writing services 71.72 (19.39) 108.03 (28.34) 83.41 (8.84) No prior exposure to writing services 73.70 (15.94) 95.21 (26.65) 82.94 (10.65)

Note: df = 172. Values in bold are effect sizes. **p < .017 (Bonferroni correction value for testing three hypothesis for each demographic variable).

H IG H ER

ED U C A TIO

N RESEA

RC H & D EV

ELO PM

EN T

723

Model 2 included the variables examined in Model 1 plus self-efficacy as the indepen- dent variables, and explained 55.1% of the variance in writing anxiety (adjusted R2 = .551, F (6, 172) = 36.125, p < .0001; Cohen’s f2 = 1.227 – very large). Of the independent vari- ables, self-efficacy made the largest unique contribution to the model (β =−.747; p < .0001). Notably, in the presence of self-efficacy, whether or not the student had prior exposure to a writing service became statistically significant (β = .107; p < .041) – a demo- graphic variable that did not contribute to Model 1. Participants who reported lower self- efficacy were female, had had no prior exposure to writing support services, and reported higher writing anxiety. In addition, language and degree level, both contributors in Model 1, became non-significant in the presence of self-efficacy as a predictor variable. Analysis of the part correlation coefficients for self-efficacy (−.691), gender (.151), and prior exposure (.105) indicated self-efficacy explained 47.75% of the variance, gender explained 2.3%, and prior exposure explained 1.10% of the variance in writing anxiety.

Model 3 included all variables in the previous models alongside EI as the independent variables, and explained 55.2% of the variance in writing anxiety (adjusted R2 = .552, F (7, 172) = 31.234, p < .0001; Cohen’s f2 = 1.232 – very large). Self-efficacy maintained its strong relationship with writing anxiety, again making the largest unique contribution to the model (β =−.722; p < .0001) while EI did not contribute to the model. In this model gender, again, made a small but significant contribution to writing anxiety (β = .157; p = .004), and prior exposure made a small but marginally significant contri- bution (β = .104, p = .047). Participants who reported lower self-efficacy were female, had had no prior exposure to writing support services, and reported higher writing anxiety. Last, analysis of the part correlation coefficients for self-efficacy (−.625), EI (−.060), and gender (.150), indicated self-efficacy explained 39.06%, EI explained .36%, and gender explained 2.25% of the variance in writing anxiety.

Discussion

This study’s purpose was to examine the writing anxiety, self-efficacy and EI of a sample of graduate students at a large, research-intensive university. In doing so, we wished to characterize our sample, compare the sample’s measures on the variables of interest, and analyze how well self-efficacy and EI predicted writing anxiety. Our final sample included 174 graduate students enrolled in either a POWER writing studio or class.

Table 2. Standardized beta coefficients for predictors of writing anxiety, among a sample of graduate students at a research-intensive university, according to different regression models.

Model 1 Adj. R2 = .062

Model 2 Adj. R2 = .551

Model 3 Adj. R2 = .552

Predictors β p β p β p

Gender .162 .037* .158 .004** .157 .004* Degree level −.153 .040* −.049 .348 −.055 .290 Language .297 .031* .007 .939 −.002 .981 International status .093 .504 .025 .793 .017 .864 Prior exposure −.013 .864 .107 .041* .104 .047** Writing self-efficacy (self-efficacy) −.747 .0001*** −.722 .0001*** Emotional intelligence −.065 *p < .05. **p < .01. ***p < .001.

724 M. HUERTA ET AL.

Student characteristics and notable measures

The average student in our sample was female, pursuing a doctoral degree, and not parti- cipating in any other writing service prior to the POWER writing studio or class. In addition, the average student was not overly anxious about academic writing. The average student reported less than average self-efficacy but high EI.

Notably, students who reported English was not their first language had statistically sig- nificant higher writing anxiety and lower self-efficacy compared to native English speak- ers. International students also showed statistically significant lower self-efficacy than students who reported not to be international. Our findings align with theorists and researchers noting the challenge K-12 students face in learning when simultaneously being on the academic English language learning trajectory while learning new content (Fang, 2006; Janzen, 2008). Graduate students are learning the content of their specific areas of research alongside academic writing. These findings are important as institutions think about how to best serve non-native English-speaking students who are studying in higher education settings where English is the primary language and/or who are not native to the country where they are studying.

Predictors of writing anxiety

Self-efficacy In our regression models, self-efficacy exhibited a significant and large association with writing anxiety. When introduced in the second model, self-efficacy increased the variance explained to 48.9% (recall: the first model included only demographic variables). Self-effi- cacy also made a large, statistically significant and negatively directed (i.e., higher self-effi- cacy equaled lower writing anxiety) contribution in both the second and third regression models.

Though, to our knowledge, no other studies have empirically explored the relationship between self-efficacy and writing anxiety with a sample of graduate students, the findings are in line with empirical findings noting negative relationships between writing anxiety and self-efficacy among students at the undergraduate level and below (Klassen, 2002; Martinez et al., 2011; Matoti & Shumba, 2011; Pajares, 2007). In addition, the study sup- ports researchers arguing the importance of fostering university students’ self-efficacy in order to lower writing anxiety and increase writing development (e.g., Mascle, 2013). This finding is useful to guide the development of writing services: providing students strategies that build their confidence in writing academically could increase students’ self-efficacy (Bandura, 1982, 1997; Mascle, 2013).

Model variations Self-efficacy also contributed to whether or not groups of students with similar character- istics experienced writing anxiety. Students who reported English was not their first language or who reported to be master’s students had higher writing anxiety in Model 1; however, the variables disappeared in significance in the presence of self-efficacy in Model 2. As researchers have noted, increasing students’ skills for self-efficacious learning in an academic setting is beneficial to their writing outcomes (Pajares, 2003; Prat-Sala & Redford, 2012; Zimmerman & Bandura, 1994). Systematically providing graduate students

HIGHER EDUCATION RESEARCH & DEVELOPMENT 725

with tools, models and support to build their academic writing skills could help graduate students with their self-efficacy in academic writing.

Gender, on the other hand, was statistically significant in predicting writing anxiety in all three regression models. Females exhibited higher writing anxiety. Our findings align with Martinez et al. (2011) who found females to experience higher writing anxiety than males among their sample of undergraduate students. However, our findings also bring up the question as to why self-efficacy did not mitigate the gender differences in the present study’s sample and regression analysis. Pajares’ (2003) review of self-efficacy noted gender differences in self-efficacy were non-significant when the analysis controlled for gender orientation beliefs (i.e., whether the person believes their gender is associated with success in an academic domain). Other factors not measured in the present study may have contributed to female students exhibiting higher writing anxiety regardless of also reporting high self-efficacy.

Prior exposure (i.e., whether students had previously sought out specific help on aca- demic writing or not) made significant contributions to the model when self-efficacy was introduced (Models 2 and 3). Specifically, students with no prior writing exposure had higher writing anxiety. Although tempting to conclude that writing services enhance self-efficacy, because the study is cross-sectional, we cannot override the possi- bility that it is the students with stronger self-efficacy who seek out the assistance of writing services.

Although we did not find empirical studies showing that increases in self-efficacy among academic writers decreases their writing anxiety, theory (Bandura, 1997), research (Zimmerman & Bandura, 1994), and books concerned with academic writers (Belcher, 2009b; Boice, 1990; Goodson, 2017) all point to the importance of building self-efficacy as a way to lower writing anxiety.

Emotional intelligence Results from the third regression model (including demographic variables, self-efficacy, and EI) indicated EI accounted for very little of writing anxiety and that the contribution was not significant. While the most immediate conclusion is that inclusion of EI within a framework to predict writing anxiety among graduate students is inaccurate, limitations related to our measures of EI can also account for the absent relationship. Because EI is multidimensional and culturally specific, the measures may not have been appropriate for a sample with such a large group of international participants. Future studies with a larger sample size would do well to examine this issue in particular, by analyzing EI scores of non-international students as compared to those of international students. EI and anxiety (outside the academic writing context) are often associated with higher work productivity and performance (Lam & Kirby, 2002; Zeidner, Matthews, & Roberts, 2004) and general academic achievement (Perera & DiGiacomo, 2013).

However, our descriptive data found a moderate effect (Cohen’s d = .305) with respect to EI when analyzing the language variable (i.e., whether or not the student reported English to be their native language). Students who reported English as their first language had higher EI scores than students who did not report English to be their first language. Past studies have noted students with higher EI tend to do better in language learning than students with lower EI (Aki, 2006). Future research on EI would be beneficial, especially given the interest in EI in terms of language and culture differences (Ekermans, 2009).

726 M. HUERTA ET AL.

Limitations, directions for future research, and practical implications

Our study has several limitations. As previously noted, we may have overlooked cultural subtleties in the EI instrument that could have skewed our results. Our sample did include overall highly emotional intelligent students, and the lack of variability in the scores may have affected the relationship between the EI variable and writing anxiety. Research should also look into the EI instrument’s fidelity with non-native English- speakers as it is possible the instrument was not culture-sensitive enough and made more sense to native English speakers, therefore accounting for the discrepancy in our findings.

Second, our sample is not generalizable to all graduate students in higher education – it was limited to a large, research university in the United States. In addition, students who took a studio or class did so voluntarily; therefore, the sample is not necessarily represen- tative of all graduate students, many of whom may not have chosen to take a class to help them in their academic writing. Nonetheless, the findings contribute to the limited research on graduate students and academic writing and can inform present practitioners and future researches in varied settings.

Third, our survey did not account for writing productivity measures. Because we wished to maintain participant anonymity and given the difficulties in accurately assessing productivity, we opted not to include this variable in our analyses. Still, writing pro- ductivity measures, including writing outcomes or writing achievement, could be invalu- able in future studies attempting to connect these variables with writing anxiety, self- efficacy, and EI measures. Future studies could also explore if these measures change over time as students participate in academic writing support structures. The studies could also continue to explore variables of interest such as gender and native language. Last, subsequent studies could certainly use larger samples than the one in this study, with the ability to apply more sophisticated statistical techniques.

Our findings, nonetheless, provide a foundational understanding of the relationship between writing anxiety, self-efficacy, and EI amongst a group of academic writers. More importantly, the study provides the first in-depth analyses of these variables in the context of graduate student academic writers, who arguably are the future of the academy.

It is to the universities’ best interest to ensure graduate students are well equipped with the tools allowing them to successfully communicate ideas and innovation in writing. Ideas for reducing graduate student writing anxiety include providing workshops and ser- vices similar to the ones mentioned in this study, in which participants are taught specific productivity strategies and afforded writing support. Tactics such as self-regulating one’s writing, writing regularly and having a writing support group have been well-documented by book authors and researchers as helping academic writers be more self-efficacious and less anxious (Belcher, 2009b; Boice, 1990; Goodson, 2017; Murray & Thow, 2014). In addition, helping non-native English speaking students in higher education settings increase their EI could benefit their academic writing.

In conclusion, it is our hope that the study lends to a first step in forging new studies and future directions in practice to help reduce writing anxiety and increase self-efficacy and EI of graduate student writers.

HIGHER EDUCATION RESEARCH & DEVELOPMENT 727

Disclosure statement

No potential conflict of interest was reported by the authors.

References

Aki, O. (2006). Is emotional intelligence or mental intelligence more important in language learn- ing? Journal of Applied Sciences, 6, 66–70.

American Psychological Association. (n.d.). Anxiety. Retrieved from http://www.apa.org/topics/ anxiety/

Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, 37, 122–147. doi:10.1037/0003-066X.37.2.122

Bandura, A. (1997). Self-efficacy: The exercise of control. New York, NY: Freeman. Belcher, W. L. (2009a). Reflections on ten years of teaching writing for publication to graduate stu-

dents and junior faculty. Journal of Scholarly Publishing, 40, 184–200. doi:10.3138/jsp.40.2.184 Belcher, W. L. (2009b). Writing your journal article in 12 weeks: A guide to academic publishing

success. Thousand Oaks, CA: Sage. Bloom, L. Z. (1981). Why graduate students can’t write: Implications of research on writing anxiety

for graduate education. Journal of Advanced Composition, 2, 103–117. Boice, R. (1987). A program for facilitating scholarly writing. Higher Education Research &

Development, 6, 9–20. Boice, R. (1990). Professors as writers: A self-help guide to productive writing. Stillwater, OK: New

Forums Press. Boice, R., & Johnson, K. (1984). Perception and practice of writing for publication by faculty at a

doctoral-granting university. Research in Higher Education, 21, 33–43. Buhi, E. R., Goodson, P., & Neilands, T. B. (2008). Out of sight, not out of mind: Strategies for hand-

ling missing data. American Journal of Health Behavior, 32, 83–92. doi:http://dx.doi.org/10.5993/ AJHB.32.1.8

Cassuto, L. (2013). Ph.D. attrition: How much is too much? The Chronicle of Higher Education. Retrieved from http://chronicle.com/article/PhD-Attrition-How-Much-Is/140045/

Daly, J. (1977). The effects of writing apprehension on message encoding. Journalism & Mass Communication Quarterly, 54, 566–572.

Daly, J. (1978). Writing apprehension and writing competency. The Journal of Educational Research, 72, 10–14.

Daly, J., & Miller, M. D. (1975). The empirical development of an instrument to measure writing apprehension. Research in the Teaching of English, 9, 242–249.

Dewaele, J., Petrides, K. V., & Furnham, A. (2008). The effects of trait emotional intelligence and sociobiographical variables on communicative anxiety and foreign language anxiety among adult multilinguals: A review and empirical investigation. Language Learning, 58, 911–960. doi:10.1111/j.1467-9922.2008.00482.x

Ekermans, G. (2009). Emotional intelligence across cultures: Theoretical and methodological con- siderations. In C. Stough, D. H. Saklofske, & J. D. A. Parker (Eds.), Assessing emotional intelli- gence (pp. 259–290). New York, NY: Springer.

Fang, Z. (2006). The language demands of science reading in middle school. International Journal of Science Education, 28, 491–520. doi:10.1080/09500690500339092

Goodson, P. (2017). Becoming an academic writer: 50 exercises for paced, productive, and powerful writing (2nd ed.). Thousand Oaks, CA: Sage.

Harris, C. S. (2011). The case for partnering doctoral students with librarians: A synthesis of the literatures. Library Review, 60, 599–620. doi:10.1108/00242531111153614

Janzen, J. (2008). Teaching English language learners in the content areas. Review of Educational Research, 78, 1010–1038. doi:10.3102/0034654308325580

Klassen, R. (2002). Writing in early adolescence: A review of the role of self-efficacy beliefs. Educational Psychology Review, 14, 173–203.

728 M. HUERTA ET AL.

Lam, L. T., & Kirby, S. L. (2002). Is emotional intelligence an advantage? An exploration of the impact of emotional and general intelligence on individual performance. The Journal of Social Psychology, 142, 133–143. doi:10.1080/00224540209603891

Libbrecht, N., Lievens, F., Carette, B., & Côté, S. (2014). Emotional intelligence predicts success in medical school. Emotion, 14, 64–73. doi:10.1037/a0034392

Martinez, C. T., Kock, N., & Cass, J. (2011). Pain and pleasure in short essay writing: Factors pre- dicting university students’ writing anxiety and self-efficacy. Journal of Adolescent and Adult Literacy, 54, 351–360. doi:10.1598/JA AL.54.5.5

Mascle, D. D. (2013). Writing self-efficacy and written communication skills. Business and Professional Communication Quarterly, 76, 216–225. doi:10.1177/1080569913480234

Matoti, S., & Shumba, A. (2011). Assessing the writing efficacy of post-graduate students at a uni- versity of technology in South Africa. Journal of Social Sciences: Interdisciplinary Reflection of Contemporary Society, 29, 109–118.

Murray, R., & Newton, M. (2009). Writing retreat as structured intervention: Margin or main- stream? Higher Education Research & Development, 28, 541–553. doi:10.1080/ 07294360903154126

Murray, R., & Thow, M. (2014). Peer-formativity: A framework for academic writing. Higher Education Research & Development, 33, 1166–1179. doi:10.1080/07294360.2014.911252

OECD. (2014). Education at a glance 2014: OECD indicators. Author. Retrieved from http://dx.doi. org/10.1787/eag-2014-en; doi:10.1787/eag-2014-en

Onwuegbuzi, A. J. (1997). Writing a research proposal: The role of library anxiety, statistics anxiety, and composition anxiety. Library & Information Science Research, 19, 5–33. doi:10.1016/S0740- 8188(97)90003-7

Pajares, F. (2003). Self-efficacy beliefs, motivation, and achievement in writing: A review of the lit- erature. Reading & Writing Quarterly: Overcoming Learning Difficulties, 19, 139–158. doi:10. 1080/10573560308222

Pajares, F. (2007). Empirical properties of a scale to assess self-efficacy in school contexts. Measurement and Evaluation in Counseling and Development, 39, 239–249.

Parker, J. D., Summerfeldt, L. J., Hogan, M. J., & Majeski, S. A. (2004). Emotional intelligence and academic success: Examining the transition from high school to university. Personality and Individual Differences, 36, 163–172. doi:10.1016/S0191-8869(03)00076-X

Perera, H. N., & DiGiacomo, M. (2013). The relationship of trait emotional intelligence with aca- demic performance: A meta-analytic review. Learning and Individual Differences, 28, 20–33. doi:10.1016/j.lindif.2013.08.002

Prat-Sala, M., & Redford, P. (2012). Writing essays: Does self-efficacy matter? The relationship between self-efficacy in reading and in writing and undergraduate students’ performance in essay writing. Educational Psychology: An International Journal of Experimental Educational Psychology, 31, 9–20. doi:10.1080/01443410.2011.621411

Salovey, P., & Mayer, J. D. (1990). Emotional intelligence. Imagination, Cognition and Personality, 9, 185–211. doi:10.2190/DUGG-P24E-52WK-6CDG

Schutte, N. S., Malouff, J. M., Hall, L. E., Haggerty, D. J., Cooper, J. T., Golden, C. J., & Dornheim, L. (1998). Development and validation of a measure of emotional intelligence. Personality and Individual Differences, 25, 167–177.

Shao, K., Yu, W., & Ji, Z. (2013). The relationship between EFL students’ emotional intelligence and writing achievement. Innovation in Language Learning and Teaching, 7, 107–124. doi:10.1080/ 17501229.2012.725730

Tabachnick, B. G., & Fidell, L. S. (2013). Using multivariate statistics (6th ed.). Boston, MA: Pearson. White, M. J., & Bruning, R. (2005). Implicit writing beliefs and their relation to writing quality.

Contemporary Educational Psychology, 30, 166–189. doi:10.1016/j.cedpsych.2004.07.002 Zeidner, M., Matthews, G., & Roberts, R. D. (2004). Emotional intelligence in the workplace: A criti-

cal review. Applied Psychology, 53, 371–399. doi:10.1111/j.1464-0597.2004.00176.x Zimmerman, B. J., & Bandura, A. (1994). Impact of self-regulatory influences on writing class

attainment. American Educational Research Journal, 31, 845–862.

HIGHER EDUCATION RESEARCH & DEVELOPMENT 729

<< /ASCII85EncodePages false /AllowTransparency false /AutoPositionEPSFiles false /AutoRotatePages /PageByPage /Binding /Left /CalGrayProfile () /CalRGBProfile (Adobe RGB \0501998\051) /CalCMYKProfile (U.S. Web Coated \050SWOP\051 v2) /sRGBProfile (sRGB IEC61966-2.1) /CannotEmbedFontPolicy /Error /CompatibilityLevel 1.3 /CompressObjects /Off /CompressPages true /ConvertImagesToIndexed true /PassThroughJPEGImages false /CreateJobTicket false /DefaultRenderingIntent /Default /DetectBlends true /DetectCurves 0.1000 /ColorConversionStrategy /sRGB /DoThumbnails true /EmbedAllFonts true /EmbedOpenType false /ParseICCProfilesInComments true /EmbedJobOptions true /DSCReportingLevel 0 /EmitDSCWarnings false /EndPage -1 /ImageMemory 524288 /LockDistillerParams true /MaxSubsetPct 100 /Optimize true /OPM 1 /ParseDSCComments false /ParseDSCCommentsForDocInfo true /PreserveCopyPage true /PreserveDICMYKValues true /PreserveEPSInfo false /PreserveFlatness true /PreserveHalftoneInfo false /PreserveOPIComments false /PreserveOverprintSettings false /StartPage 1 /SubsetFonts true /TransferFunctionInfo /Remove /UCRandBGInfo /Remove /UsePrologue false /ColorSettingsFile () /AlwaysEmbed [ true ] /NeverEmbed [ true ] /AntiAliasColorImages false /CropColorImages true /ColorImageMinResolution 150 /ColorImageMinResolutionPolicy /OK /DownsampleColorImages true /ColorImageDownsampleType /Bicubic /ColorImageResolution 300 /ColorImageDepth -1 /ColorImageMinDownsampleDepth 1 /ColorImageDownsampleThreshold 1.50000 /EncodeColorImages true /ColorImageFilter /DCTEncode /AutoFilterColorImages false /ColorImageAutoFilterStrategy /JPEG /ColorACSImageDict << /QFactor 0.90 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> /ColorImageDict << /QFactor 0.40 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000ColorACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 15 >> /JPEG2000ColorImageDict << /TileWidth 256 /TileHeight 256 /Quality 15 >> /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 150 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages false /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict << /QFactor 0.90 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> /GrayImageDict << /QFactor 0.40 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000GrayACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 15 >> /JPEG2000GrayImageDict << /TileWidth 256 /TileHeight 256 /Quality 15 >> /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Average /MonoImageResolution 300 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict << /K -1 >> /AllowPSXObjects true /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False /Description << /ENU () >> >> setdistillerparams << /HWResolution [600 600] /PageSize [595.245 841.846] >> setpagedevice