Factors Contributing to Student Attrition at English Language Institutes in Iran

Research in English Language Pedagogy

مقاله 2، دوره 11، شماره 3، آذر 2023، صفحه 317-338   

نوع مقاله: Original Article

شناسه دیجیتال (DOI): 10.30486/relp.2022.1962894.1390

نویسندگان

Shabnam Ettehad؛ Masoud Zoghi^* ؛ Hanieh Davatgari-asl

Department of ELT, Ahar Branch, Islamic Azad University, Ahar, Iran

چکیده

The study intended to fill the research gap related to academic attrition by identifying factors contributing to learners' attrition at English language institutes (ELIs). Given the classification of voluntary and compulsory attrition factors, attempts were made to determine which of these two sets of pressures could predict the attrition decision of EFL learners at ELIs. To this end, 148 enrolled students (81 males and 67 females) from seven ELIs participated in this research. Based on their self-reported proficiency accounts, the participants were considered beginner, pre-intermediate, and intermediate-level students. They completed the survey response online via Google forms, a computer-generated web-based program. Results based on hierarchical regression analysis indicated that the voluntary variables could contribute more to the predictive model for student attrition at ELIs than the compulsory factors. More specifically, the attrition decisions were more conscious and firmer for ELI students with a lower level of motivation for academic success, satisfaction with the quality of education, and satisfaction with instructors. However, the analysis did not find satisfaction with the institute environment and social integration in the institute as predictors of attrition decisions. This study concludes that voluntary pressures among ELI students are the most important predictors of attrition decisions.

کلیدواژه‌ها

Attrition؛ Compulsory Factors؛ English Language Institutes؛ English Language Learners؛ Voluntary Factors

اصل مقاله

1. Introduction

Available statistical data show that 1.5 billion English-language learners currently learn English worldwide (Beare, 2020). The number of English learners worldwide is also expected to increase in the coming years. Educational scholars often refer to this surge in English language learners as the English Effect (Beare, 2020). In much the same vein, Iranian youth have realized the importance of learning English, and, as a result, English language learning has become a considerable public concern (Moharami & Daneshfar, 2021).

Most Iranian families believe that formal EFL education cannot meet their expectations, and thus their children have to participate in English classes at English language institutes (ELIs). This has led to an upsurge in enrollments across ELIs (Gholaminejad & Raeisi-Vanani, 2021). Young EFL learners are willing to attend such institutes for different reasons. Some take classes to improve their communicative ability to continue education abroad, some wish to find better jobs with the help of English language mastery, and others need to improve their test-taking ability necessary for the national entrance examination and get scholarships (Mohammadian-Haghighi & Norton, 2016). Whatever the reason, the increasing demand for students to attend ELIs is evident, and, as a result, ELIs remain an educational necessity in the Iranian EFL context. Above all, since the profile of today's students is significantly different from the past and continues to expand, much is expected of such institutes.

Despite the fast growth of ELIs in Iran, they are not adequately monitored regarding their L2 educational outcomes (Moharami & Daneshfar, 2021). While Iran's policy on English education has played down and up the significance of English (Mohammadian-Haghighi & Norton, 2016), some students in different courses at ELIs either drop out or fail to complete the intended programs as scheduled (Gholami-Zafarani et al., 2015). Other authors apply different terminologies for this process, e.g., dropout, withdrawal, and non-completion. This study uses these technical words under the blanket term of student attrition, similar to a process in which students fail to reenroll in consecutive semesters and complete a program. As a result, this could be a permanent or temporary withdrawal from an institute.

A recent report from Young Journalists Club (2021) indicates that student attrition has become an issue of national importance since about two hundred and ten thousand elementary-level students and about seven hundred and sixty thousand secondary-level students have dropped out.. Moreover, about seven hundred and sixty thousand secondary-level students have dropped out. Informed by Gholami-Zafarani et al.'s study (2015), we postulate that attrition is taking place in academic settings like ELIs in the Iranian EFL context. This issue has been dealt with in the only existing study on attrition at ELIs by Gholami-Zafarani et al. (2015). Unfortunately, no evidence indicates the academic attrition rate among ELI students in the private sector. To add further complications to the situation is that the potential determinants of student dropout at ELIs have remained unknown, and generalizations about it can be elusive due to the uniqueness of each language institute. It is a challenging issue that, despite various lines of research, educational institutions still struggle with effective programs to decrease student attrition rates (Ashghali-Farahani et al., 2017). Hence, the search for the development of more effective ways to enhance student success could be a high priority for ELIs, as well.

As a theoretical basis, this study was informed by Bennett's (2003) postulation that a distinction had been made between voluntary and compulsory pressures for attrition factors. The choice was made because it was thought that while the work of Bennett (2003) shares similar core elements with its predecessors, it also complements those (Christo & Oyinlade, 2015). Bennett (2003) asserts that voluntary attrition is due to students' calculated decisions. These include such decisions as feelings of insufficient educational challenge, dislike of peers or teachers, loss of motivation to continue, and feelings of boredom. He argues that compulsory departure occurs because of decisions that students have not made out of their own choice. Compulsory dropout refers to push conditions where external forces make students withdraw. Common examples are serious illness, financial inability, and family obligations. That being said, the following statement provides the underlying logic for designing and conducting this investigation. If compulsory and voluntary factors can lead to departure from school, college, and university students, then the same factors will consequently predict the attrition of ELI students. Therefore, this study intended to extend this theory from university settings to ELIs to determine whether this set of factors applies to the unexplored context of ELIs.

As discussed below, literature abounds with studies focusing on student attrition in different levels of education, particularly in postsecondary (tertiary) education (Andrade et al., 2020). Educational researchers have also found several factors contributing to student attrition, such as motivation resources (Litalien & Guay, 2015), social integration (Ishitani, 2016), and education quality (Daka & Changwe, 2020), to name a few. These factors of student attrition may vary by the type of educational setting in which attrition occurs (Shcheglova et al., 2020). However, one specific research gap in the literature is the lack of attention to attrition in small and private institutions like ELIs. Limited input from the related literature has caused a lack of understanding about EFL students' departure from ELIs. Against this background, this study is among the first to contribute to the much-needed evidence of student attrition in the language learning context. The current study could also contribute to the under-researched area of EFL learners' attrition in ELIs and generate further questions for future research studies in the related areas.

 

2. Literature Review

Student attrition was posited in the late 1900s. The background studies on student attrition are rich in theoretical and empirical models, together with the research that broadly investigated the learner attrition phenomenon and the involvement of learners in educational settings (Russell & Jarvis, 2019; Shcheglova et al., 2020). Different models of student attrition concentrate on the effects of academic and social systems on student behavior, such as dropout and retention (Ertem, & Gokalp, 2019, p. 4), such as Bean's Student Attrition Model (1980), Tinto's Institutional Departure Model (1975), and Bennett's Attrition Model (2003). These models account for the predictive power of either internal or external factors on student attrition.

Social integration is among the variables whose impact on attrition is best supported in the pertinent literature (Ugwu & Adamuti-Trache, 2019). Empirical data suggest that the attrition rate is associated with social integration and year of study (Ryan et al., 2021). In his research, Ishitani (2016) found that first-year students struggled with social integration. Further evidence from different studies suggests that 40% and 70% of students who cannot complete their studies within four to six years of initial enrollment are likely to drop out of the university within the first year of study (Zając & Komendant-Brodowska, 2018). Nonetheless, the impact of social integration does not hold in all educational contexts. Perhaps social integration exerts far less influence on student attrition in contexts that are characterized by an individualistic culture. Several studies indicate that a positive personal experience during the first year of study could be more influential than social integration (Qvortrup & Lykkegaard, 2022).

As is evident from the literature, there is a negative relationship between education quality and student attrition. In their review study, Daka and Changwe (2020) found that student attrition and education quality in higher education are strongly linked. In another study, employing a sequential explanatory mixed methods design, Beebe (2014) examined the school-related reasons for the withdrawal of students from private educational settings. Data from four private schools revealed that the most significant predictor of student withdrawal was disappointment with a school's quality of education.

 To further understand student attrition, it is also imperative to gain insight into the role of teacher characteristics, such as a teacher's content knowledge, pedagogical knowledge, disposition, and experience. Some studies also indicate that teachers vary widely in their influence on attrition in higher education. In a meta-analysis on the factors associated with student persistence in higher education, Hart (2012) found that meaningful interactions between students and instructors significantly predictor student ability to overcome attrition intentions. This issue has been highlighted in recent works by revealing the significant impact of teachers on their learners' (dis) engagement and motivation (Gregersen & Al Khateeb, 2022; Seebruck, 2021; Stronge, 2018).

Another line of recent research has demonstrated that student attrition is associated with motivation resources (Litalien & Guay, 2015; Strelan et al., 2020). Based on self-determination theory (Ryan & Deci, 2017), Litalien & Guay (2015) developed and tested a predictive model for doctoral attrition intentions. Their findings confirmed the significant contribution of motivation resources and psychological need support to doctoral studies completion. Specifically, they found that perceived competence played an important role in doctoral dropout intentions and was predicted by autonomous and controlled regulations and advisor support. Based on the expectancy-value theory, Schnettler et al. (2020) analyzed the intra-individual process of student attrition to understand its motivational underpinnings. Their analyses revealed that intrinsic value, attainment, and cost were linked to intra-individual changes in attrition decisions.

In a qualitative study, Russell & Jarvis (2019) show that the critical factors for student attrition can be categorized according to internal factors (structures within the university, such as institutional support mechanisms and course content) and external (structures outside of the university, such as personal and social reasons, e.g., domestic responsibility, grief and feeling of compatibility with certain courses and institutions). Previous research has also shown that life-university conflicts are pivotal in students' decision to leave before graduation. Christo and Oyinlade (2015) and Gale et al. (2015) found that compulsory variables were influential in university student attrition. They found adjusting to university difficult for some students as they had to fit personal and family commitments into their academic workload. Similarly, a retrospective case study of attrition conducted by Yates (2012) showed that the line between voluntary withdrawal and academic failure is not always clear-cut. However, personal or family health problems contributed to voluntary withdrawal motivation.

As discussed earlier, despite the potential negative influence of attrition on both students and academic institutions (Shcheglova et al., 2020), to the best of the researchers' knowledge, only one study in the Iranian EFL context to date has addressed the issue of attrition among EFL learners. To investigate the potential differences between placed and promoted students in the attrition rate at the Iran language institute (ILI) as a typical language institute in Iran, Gholami-Zafarani et al. (2015) conducted an exploratory study. Their investigation showed no significant difference in the attrition rate of placed and promoted learners, particularly in lower levels of English language learning. However, the attrition rate of placed learners in higher levels (i.e., in high, intermediate, and advanced levels) was far higher than that of promoted learners. This finding pointed to the fact that the placed learners dropped out of the institute much more than the promoted learners – an issue that highlights the inadequate assessment at the entry-level ELIs.      

Notwithstanding many studies on student attrition, scholars have scarcely zoomed in on learner attrition at small and private institutions. A higher attrition rate is observed at small and private institutions compared to other academic settings (Ishitani, 2016). Given the absence of attrition studies at ELIs and the nature of ELI learners (typically students aged 18 to 22), the researchers in this study assumed that voluntary factors would probably be stronger for student attrition than compulsory factors. It was largely based on the researchers' assumptions that compulsory factors could negligibly provide any explanations for ELI student attrition. Non-traditional students are more prone to be influenced by external, compulsory pressures. ELI learners in Iran comprise those who take classes mainly to fulfill their personal needs. That is why we assumed that personal internal factors such as low motivation, low satisfaction with the quality of education, and low academic interests would force EFL learners to drop out. Overall, we postulated that voluntary pressures would be important predictors and outweigh any compulsory pressures that cause learners to drop out of ELIs.

With this in mind, this study examined EFL learners' attrition at ELIs by obtaining statistical results about the role of compulsory and voluntary pressures. More specifically, it intended to determine the extent to which compulsory variables add to the incremental change (in the criterion variable) after voluntary variables with higher priority could contribute their share to the prediction of the attrition decision. Accordingly, the following research questions were addressed:

1. Which of the two sets of voluntary and compulsory variables explained attrition more significantly than the other?
2. How can each variable in each set of variables individually predict the decision of attrition?

3. Methodology

3.1. Design and Context of the Study

Informed by the quantitative approach, the current research used the correlational research design to determine potential predictors (the voluntary and compulsory variables) of EFL learners' attrition decisions. The data for the present study were collected in 2019 over three months in Karaj, Iran. Due to the difficulty of recruiting former ELI students (i.e., non-completers) in a relatively short period, we contacted eleven accessible ELIs in Karaj. Only seven private ELIs out of the contacted institutes agreed to cooperate.  

 

3.2. Participants

The intended population in this study was the inactive students who had recently dropped out of ELIs located in Karaj, Iran. As such, the population included the students who had withdrawn from the ELI programs before the winter of 2019. Recruiting of participants occurred by employing a student database maintained by the seven ELIs' management offices. Based on the purposive sampling method, the criteria for selecting the participants included: (1) Having completed at least one academic semester at an ELI in Karaj; it was necessary to ensure that they could provide dependable responses to some of the questions on the survey, (2) Being a recent dropout at the time of conduct of this study; they must have dropped out during the period of winter 2018 to 2019, (3) Being an adult-age student; they had reached the age of 18 at the time of dropping out, and (4) Being either inactive or withdrawn from the English language program at the time of conduct of this study.

It is worth noting that a priori power analysis using G*Power was run to determine the minimum number of participants for hierarchical multiple regression analysis. The minimum sample size required to detect an effect size = 0.15, probability level = 0.05, and power level = 0.95 for the additional set of variables to the model was 107. In light of the sample size calculation, a sample of email addresses and phone numbers of 256 dis-enrolled students was obtained from the management offices of seven ELIs. Some individuals were contacted (via email and phone call) more than once. Of these, 216 dis-enrolled students agreed to participate in this study by signing and returning a consent form. However, 28 of them failed to complete the survey. A sample of 40 disenrolled students was selected to participate in a pilot study.

Moreover, for the final study, 148 individuals (81 males and 67 females) completed the survey responses. This sample size was regarded enough based on the a priori power analysis. The participants were of different ages (M= 20.14, SD= 2.49); had different self-reported proficiency levels (beginner, pre-intermediate, intermediate); and had dropped out of ELIs before the winter  of 2019 (Table 1).

 

 

 

 

Table 1.

Demographic Characteristics of Participants (n = 148)

Demographic Variable	Numerical Data
Gender
Male	81
Female	67
Proficiency Level
Beginner	74
Pre-intermediate	49
Intermediate	25
Age
Mean	20.14
Std. Deviation	2.49

 

3.3. Instrument

Since the data collection was done during lockdown - the coronavirus (COVID-19) pandemic - the data for the present study were gathered remotely. The Revised Persian Version of the Attrition Questionnaire (RPVAQ) was developed and employed as a survey instrument to measure different aspects of attrition – compulsory and voluntary attrition pressures and EFL learners' attrition decisions at Iranian ELIs. The survey was adapted from Christo & Oyinlade's (2015) self-report attrition questionnaire. The 63-item RPVAQ consists of demographic details (age, gender, and self-reported proficiency level) and is organized into three sections or subscales: subscales for voluntary attrition factors [motivation for academic success (items 1 to 6), satisfaction with quality of education (items 7 to 14), satisfaction with instructors (items 15 to 24), satisfaction with institute environment (items 25 to 28), and social integration (items 29 to 36)]; subscales for compulsory attrition factors [institute self-estrangement (items 37 to 44), institute-life conflict (items 45 to 49), life-institute conflict (items  50 to 53),] and a subscale for the attrition decision (items 54 to 64). The items on the instrument are rated on a 4-point Likert-type scale ranging from Strongly Agree = 4 to Strongly Disagree =1. The total score is obtained by summing all items for each component, and higher scores indicate greater levels of each measured component.

Both translation and validation involved expert judgments on the content of the RPVAQ. To achieve both semantic and functional equivalence in translation, a panel of English language and psychology experts commented on its relevancy and linguistic congruence by suggesting modification or removal of certain non-relevant items. Three well-qualified independent translators applied forward (or one-way) and backward translations in the translation process—forward translation allowed for translating the English questionnaire into Persian. The translated Persian questionnaire was compared with the original English version to discover any possible ambiguities and discrepancies in words, sentences, and meanings. Given the pragmatic value of back translation as a translation quality assessment tool, the translated questionnaire (Persian) was later translated back into its source language (English). Neither original nor back-translated versions of the questionnaires yielded any major discrepancies or ambiguities.

Although the validity and reliability of the original attrition questionnaire have been established (Christo & Oyinlade, 2015), a psychometric evaluation of the RPVAQ was deemed necessary to ensure its appropriateness for assessing EFL learners' attrition at Iranian ELIs. To this end, in a pilot study, the responses of forty students were used to test the psychometric properties (construct validity and internal consistency) of the RPVAQ through factor analysis (FA) and reliability coefficient estimation.

To this end, the suitability of data, the sampling adequacy, and the sphericity assumption were checked through Pearson correlation coefficients, Kaiser–Meyer–Olkin (KMO) measure, and Bartlett's test, respectively. Results showed that none of the assumptions were violated. Then, the 64 items of the RPVAQ were subjected to principal component analysis without rotation. The commonalities for each variable (the proportion of variance each item has in common with other items) were estimated. Communalities for these data ranged between .51 and .91. As a further validation check, we compared inter-scale correlations with internal consistency reliability coefficients (Cronbach's alpha). Results supported that each scale measured a unique concept.

In addition, Cronbach's alpha, the most common measure of scale reliability, was utilized to check the reliability of each scale. The scales of the RPVAQ all had acceptable reliabilities, namely Satisfaction with Quality of Education, Cronbach's α = .884; Social Integration in Institute, Cronbach's α = .874; Satisfaction with Instructors, Cronbach's α = .935; Motivation for Academic Success, Cronbach's α = .749; Satisfaction with Institute Environment, Cronbach's α = .857; Self-estrangement from Institute, Cronbach's α = .939; Institute-Life Conflict, Cronbach's α = .897; and Life-Institute Conflict, Cronbach's α = .911. Therefore, the reliability coefficient for the attrition decision was noted to be very high, Cronbach's α = .902. In sum, the results of this analysis supported the use of the RPVAQ as a valid and reliable measure for assessing voluntary and compulsory factors along with ELI students' attrition decisions.

 

3.4. Data Collection Procedure

The survey data for three months was collected. First of all, eleven ELIs in Karaj were invited to participate in the study. Seven ELIs showed interest, consequently obtaining consent to participate in the study. As compensation, the first author agreed to teach two classes free of charge for one semester. The participating ELIs were also informed that the outcomes of this research would be shared with them.

Permission to adapt and use the attrition questionnaire was also obtained. The electronic survey was then developed using Google forms, a computer-generated web-based program. Afterward, a participation consent letter was emailed to the intended sample (former ELI students), inviting them to participate in this research study. The letter identified the main characteristics of the study while discussing compensation and confidentiality and providing contact information for the researcher. Moreover, included in this email was a copy of the RPVAQ. Upon completion of the study, the researcher (the first author) emailed the study's findings to each participant. Reminder follow-ups were used to increase the response rate for the study. A follow-up email was sent ten days later to those who did not respond by the due date. Once again, after ten days, a second reminder was emailed; after two more weeks, a final telephone follow-up was done. The researcher personally contacted those who did not respond on the due date and invited them to participate. Individuals choosing to participate were requested to answer the RPVAQ online via Google forms before the survey window closed.

 

3.5. Data Analysis Procedure

All the analyses were carried out based on the significance level of 0.05 and through the statistical software of SPSS version 22. The hierarchical or sequential multiple regression was employed as it was intended to determine the relative significant contributions of each set of independent variables (voluntary and compulsory factors) to predict the attrition decision among the former ELI students. This statistical test analyzes the amount of variance explained in a dependent variable by independent variables entered in blocks or groups. Using hierarchical regression analysis to predict ELI students' attrition has several advantages. It allows us to important input variables in light of theory and logic. Furthermore, this analysis can indicate whether the predictor variables, entered in a given order, could predict the incremental change (R2) in the criterion variable by evaluating variances after adding each set of predictors.

 

4. Results

To answer the research questions, two models based on theoretical considerations of this research were developed:

Model 1: Students’ attrition decision = Intercept + Voluntary Pressures

Model 2: Students’ attrition decision = Intercept + Voluntary Pressures + Compulsory Pressures

Table 2 shows the descriptive statistics for the voluntary and compulsory variables and the predicted variable (student attrition).

Table 2.

Descriptive Statistics Related to the Variables (n = 148)

Variables	Mean	Std. Deviation
Attrition	3.64	1.15
Satisfaction with Quality of Education	2.40	1.25
Social Integration in Institute	2.06	1.01
Satisfaction with Instructors	2.28	1.08
Satisfaction with Institute Environment	2.19	1.07
Motivation for Academic Success	2.45	1.31
Life-Institute Conflict	3.30	.77
Self-estrangement from Institute	2.94	1.01
Institute-Life Conflict	3.20	1.20

 

                                           

 

Before data analysis, the following assumptions of hierarchical multiple regression were also checked and found tenable: (a) normality, (b) independence of observations (autocorrelation), (c) linearity, (d) multi-collinearity, and (e) homoscedasticity. In the Voluntary Model (Model 1), five variables (i.e., motivation for academic success, satisfaction with the quality of education, satisfaction with instructors, satisfaction with the institute environment, and social integration) were block-entered into the hierarchical regression equation. It enabled us to assess voluntary pressures' individual and collective abilities to explain attrition decisions. As shown in Table 3, the overall voluntary model was significant and accounted for a significant proportion (19%) of the variance in ELI students' attrition decision, R² = .190, F (5, 142) = 6.683, p = .000.   

 

Table 3.

Model Summary

Model	R	R Square	Adjusted R Square	Std. Error of the Estimate	Change Statistics					Durbin-Watson
					R Square Change	F Change	df1	df2	Sig. F Change
1	.436	.190	.162	1.058	.190	6.683	5	142	.000
2	.468	.219	.180	1.047	.028	2.540	2	140	.082	2.032

 

Only three variables individually contributed to the model for attrition decision (Table 4). Attrition decision was found to be more conscious and firmer for ELI students with a lower level of motivation for academic success (β = -.23, p = .03), satisfaction with quality of education (β = -.40, p = .00), and satisfaction with instructors (β = -.197, p = .02). However, the analysis did not find satisfaction with institute environment (β = -.02, p = .77), and social integration in institute (β = -.11, p = .17) as predictors of attrition decision.

 

 

Table 4.

Standardized Regression Coefficients

Standardized Regression Coefficients Model		Unstandardized Coefficients		Standardized Coefficients	t	Sig.
		B	Std. Error	Beta
1	(Constant)	4.826	.262		18.412	.000
	Satisfaction with Quality of Education	-.370	.101	-.402	-3.653	.000
	Social Integration in Institute	-.132	.096	-.116	-1.372	.172
	Satisfaction with Instructors	-.210	.092	-.197	-2.273	.025
	Satisfaction with Institute Environment	-.028	.098	-.027	-.291	.771
	Motivation for Academic Success	-.209	.097	-.238	-2.149	.033
2	(Constant)	3.471	.668		5.199	.000
	Satisfaction with Quality of Education	-.324	.104	-.352	-3.111	.002
	Social Integration in Institute	-.126	.096	-.111	-1.318	.190
	Satisfaction with Instructors	-.174	.093	-.163	-1.870	.064
	Satisfaction with Institute Environment	-.027	.097	-.025	-.280	.780
	Motivation for Academic Success	-.243	.097	-.277	-2.498	.014
	Self-estrangement from Institute	-.115	.087	-.108	-1.322	.188
	Life-Institute Conflict	.249	.124	.166	2.008	.047
	Institute-Life Conflict	.074	.083	.078	.900	.370

 

 

As for the Compulsory Model (Model 2), three variables of compulsory pressures (self-estrangement from an institute, institute-life conflict, and life-institute conflict) were added to the hierarchical equation (Table 3). It was the full model since it contained the voluntary factors from the previous model and the compulsory variables. Results of this analysis indicated that the full model accounted for 21% of the variance in ELI students' attrition decision, R2 = .219, F (7, 140) = 5.603, p = .000. However, the compulsory factors failed to explain any significant addition of variance in ELI students' attrition decision, ΔR2 = .028, p = .082. Because the difference of Δ R2 between Model 1 and Model 2 was found to be statistically non-significant, we deduced that the added set of variables in Model 2 could not account for ELI students' attrition decision beyond the variables in Model 1.

 

5. Discussion

To begin, it is worth mentioning that the literature link to student attrition at ELIs is weak and underexplored. Given the lack of empirical research on student attrition in this context, our discussion of findings draws on research findings in other educational settings, such as colleges and universities.

To answer the first research question, the results suggested that the voluntary variables could contribute more to the predictive model for student attrition at ELIs than the critical factors. Overall, regression findings partially supported our postulate that voluntary factors account for ELI students' attrition. Given the second research question, three internal to the program factors (satisfaction with quality of education, motivation for academic success, and satisfaction with instructors) were found to be predictors of student attrition in the program. The other two, i.e., satisfaction with the institute environment and social integration in the institute, did not reach statistical significance. Overall, ELI students' attrition decision was implicated in lower motivation levels for academic success, satisfaction with the quality of education, and satisfaction with instructors. Voluntary and compulsory variables of the model in light of the current research were discussed, and possible explanations were offered.

 

5.1. Findings Pertinent to Voluntary Factors

The hierarchical analysis shows that satisfaction with the quality of education explained attrition decisions among the ELI students studied. Specifically, the results suggest that satisfaction with quality education was the strongest factor associated with attrition decisions among ELI students in the first and final models. Numerous studies reported that such internal and voluntary factors positively impacted students' attrition decisions (Beebe, 2014; Daka & Changwe, 2020; Shcheglova et al., 2020). In particular, the finding related to education quality is compatible with the findings of Daka & Changwe (2020), whose review of the attrition literature revealed that the main factors affecting attrition potentials in educational institutions are good teaching, students' satisfaction with courses, and the careful matching of courses with students.

Many investigations emphasizing the predictive power of course quality have also detailed why directors must be aware of the quality of their educational programs (Russell & Jarvis, 2019; Strelan et al., 2020). Specifically, Toffler (1990) strongly contended that today's curriculum is academic content that has survived from an earlier time and has no meaning. After such a long time, there have been insignificant changes in the situation, though (Litalien & Guay, 2015). We are beginning to face a newer generation of students (known as Generation Alpha) who demonstrate a preference for a different type of education with greater flexibility. Their notable and defining characteristics are (1) a tendency to spend less time on tasks and achieve success with little effort, (2) in want of immediate feedback, and (3) a need to have a quick guide to be successful in and out of class. Undoubtedly, with these students, we need to accept the new normal and majorly revise the whole concept of the EFL curriculum.

Motivation for academic success was associated with attrition but not as strongly as satisfaction with the quality of education. Motivation for academic success accounted for the second-largest portion of the variance in ELI students' decision to drop out. Our present finding supports the view that student attrition can be motivational. Other studies have also yielded similar results and have found that motivation had a direct and significant influence on dropout decisions (Schnettler et al., 2020). One possible reason could be that ELI students with poor motivation for academic success feel drained of the necessary skills to learn English; this might make them more prone to attrition intentions. Low academic motivation is, in general, regarded as one of the strongest predictors of attrition decisions (Litalien & Guay, 2015). Past research has shown an overall decline in motivation over time (Strelan et al., 2020). If such a decrease is combined with other factors, a firmer decision to drop out from ELI students could be made. Students, therefore, need to retain motivation for a long while. The multidimensionality of motivation suggests that various motivational sources could impact students' motivation for academic success (Litalien & Guay, 2015), such as teachers, peers, and parents, to name but a few. However, the intraindividual aspect of motivation seems to play a more central role in attrition decisions (Schnettler et al., 2020).

In addition, this finding gives more credence to the work of Eccles et al. (1983) – Expectancy-Value Theory (EVT). EVT consists of two components, i.e., expectancy and value. These two important constructs of EVT seem to have the potential to explain the motivationally-based aspect of attrition decisions. EVT postulates that students' decision to pursue a task is impacted by expectations of success and subjective task value. Expectancy refers to the probability of success on the task (Schnettler et al., 2020). Expectancy could suffer when students do not perceive that their effort pays off. The ELI students' attrition decision may be likely due to their low expectancy levels, as they felt that something other than effort (e.g., lack of resources) would predict performance.

Therefore, the value component can be regarded as an incentive to complete the task (Schnettler et al., 2020). The four values in EVT are attainment value (the importance of successful task completion), utility value (the immediate or future importance of a task), intrinsic value (the pleasure of doing a task), and cost (the amount of sacrifice to engage in a task). In line with EVT, as evidence suggests here, when ELI students realized that learning English at a given ELI was less enjoyable (intrinsic value), less important (attainment value), less consequential (utility value), and less beneficial (cost), they decided more firmly to drop out. In other words, when the pain side of the task (learning English) was stronger than the gain side, students had no hesitation in discounting attendance at ELIs.

Additionally, satisfaction with Instructors accounted for the third-largest portion of the variance in ELI student attrition decisions. This result supports the research literature that the instructor factor is a predominant predictor of student retention or attrition (Hart, 2012; Stronge, 2018). It may be counterintuitive that several studies have identified instructor quality as the most critical factor for student success and demotivation (Gregersen & Al Khateeb, 2022). In this study, it became evident that such factors as instructors' (1) individual characteristics and qualifications, (2) level of commitment to teaching, (3) teaching experience, and (4) teaching quality may, to a great extent, contribute to instructor quality. Any deficit in the quality of ELI instructors seemed to provide the motivation necessary for students to withdraw from the programs.

Another way to appreciate this finding is to note the research evidence highlighting the importance of the student-teacher relationship, though mainly at the primary or secondary education level (Seebruck, 2021). While confirming an old idea, this finding may appear new as the importance of such a relationship becomes known in ELI settings. From the perspective of self-determination theory or SDT (Ryan & Deci, 2017), we can posit that the role of instructors in predicting student attrition is considerable. According to this theory, students' sustainable behavior is weakened without emotional support from external factors like teachers. A construct that has been underscored in SDT is a sense of relatedness. Although it seems necessary to pay attention to intrapersonal factors, it is also essential to consider the social context of the classroom where the teacher-student relationship is shaped; a weak bond with instructors may make students less resilient during their study, as evidenced by this study. Positive classroom relationships in all classroom types are vital for successful learning (Gregersen & Al Khateeb, 2022). However, such relationships are significant for language classes at ELIs as they may energize students to remain engaged and work harder despite difficulties. ELI instructors have the power to enhance or dampen motivation for learning in their students; students may feel demotivated to learn if they realize that their instructors do not care about them. Therefore, these results are evidence of the importance of the instructor factor in line with previous literature (Seebruck, 2021; Stronge, 2018).

Notwithstanding the significant contributions of the voluntary factors of satisfaction with the quality of education, motivation for academic success, and satisfaction with instructors, the remaining two voluntary factors, i.e., satisfaction with the institute environment and social integration in the institute, were not significant contributors to the models. In other words, neither of these variables could bring any influence to bear upon student attrition. Our findings concerning satisfaction with institute environment and social integration in institute are inconsistent with previous research (Ryan et al., 2021; Ugwu & Adamuti-Trache, 2019; Zając & Komendant-Brodowska, 2018). One possible explanation may be that there are differences in the institutions studied. ELI students in private and small institutes probably experience language education differently. Unlike college or university students, ELI students seemed to adjust more easily to their social environment. More specifically, students' positive experiences with the social systems existing in ELIs, or positive relationships with peers and other individuals (e.g., staff) could strengthen their social integration. It is possibly true that students may have certain attributes that integrate well into the social systems of ELIs. Alternatively, these institutes have adapted successful policies to look after social and environmental factors such as peer culture, identification with a group, social involvement and regulations, ease of registration, and staff attitudes. Overall, this research's evidence indicates a personal dimension to attrition decisions for ELI students (Ishitanil, 2016; Qvortrup & Lykkegaard, 2022).

 

5.2. Findings Pertinent to Compulsory Factors

Including compulsory factors, the full model accounted for 21 percent of attrition decisions among the participating ELI students. However, none of the compulsory factors were individually significant contributors to the model. Neither the first nor final model, self-estrangement from the institute, institute-life conflict, and life-institute conflict, reached statistical significance and could add any variance to the predictive models. These findings counter the empirical findings of scholars who have reported compulsory pressures as the most important predictors of attrition (Christo & Oyinlade, 2015; Yates, 2012). Compulsory factors, such as serious illness, family obligations, financial inability, and unexpected social conditions, have already been identified (Russell & Jarvis, 2019). However, given the lack of evidence, it is hard to discern the extent to which these compulsory factors are essential in dropping ELIs.

Nevertheless, it stands to reason to state that ELI students are mainly traditional and do not typically play multiple and conflicting roles; as a result, their attendance at ELIs may not interfere with their daily life activities. This situation is reversed when we consider college or university students as non-traditional students. They are prone to experience institute-life conflict, which may, in turn, affect their intention to stay in an educational program (Christo & Oyinlade, 2015; Gale et al., 2015).

In this study, self-estrangement from the institute was not evident; this finding runs counter to previous research showing that self-estrangement was one of the most powerful predictors of attrition likelihood (Christo & Oyinlade, 2015). It may be that the circumstances at ELIs were not different from students' expectations, and thereby they did not have to conform to any condition. However, such a postulation might hold for the study participants, and in some other settings, this factor could influence students' decision to withdraw from the course.

In general, the results add support to the contributory role of voluntary (internal) factors (e.g., educational experiences of learners) in student attrition and confirm the view that voluntary factors influence students' attrition decisions to a greater degree than compulsory (external) factors.

 

6. Conclusion

Student attrition does not occur in one defining moment; instead, students decide to leave for various interrelated reasons (Ertem & Gokalp, 2019; Russell & Jarvis, 2019). Some of these are voluntary and related to the quality of teaching, teacher characteristics, and academic motivation; others are compulsory and include life-institute or institute-life conflicts. The variety of views on attrition speaks to the need for caution in interpreting and judgments about attrition factors. This study provided corroborative evidence only for personal, voluntary variables relevant to ELI student attrition. The current research added to this understanding by highlighting that these personal variables are better understood as closely linked factors important to the attrition decision of ELI students.

This study's findings have several implications for policy and practice at ELIs. As became evident, voluntary factors explained the most significant reasons for attrition decisions among ELI students. The strong implications of the strongest association with attrition were voluntary factors. This finding will behoove ELI directors to investigate and find solutions to the poor education quality that causes attrition among ELI students. It may be related to structural conditions. Courses and instructional methods may need retrofitting and redesigning to increase learning outcomes.

Importantly, some issues related to the admission policy of students that may affect their performance outcomes need reforming (Gholami-Zafarani et al., 2015). ELIs must rethink their admission policies related to the initial proficiency check procedures. Over-reliance on placement tests in paper-and-pencil formats may no longer work. A different way of assessment, like oral interviewing, can reveal a lot about an entering student's language proficiency, needs, and expectation. ELIs can place each entering student in classes at the right level.

Despite their contributions to existing knowledge, the study's results should be considered in light of several limitations that may, to some extent, affect external and internal validity. The findings of this study are limited in application to the selected ELIs with a limited number of participants. This study was also limited to a certain research area. It might affect the generalizability of the findings outside the boundaries of the present research.

From a quantitative perspective, the Revised Persian Version of the Attrition Questionnaire (RPVAQ) used in this study is a self-report measure and is subject to bias. Self-report data are considered biased as respondents or interviewees are consciously or unconsciously influenced by social desirability. This research did not exhaust all other possibilities of correlative factors on an individual's attrition decision. For example, factors such as learning styles, emotional intelligence, and demographic and characteristics information were not included in the current research. Therefore, caution is encouraged while generalizing the study's findings.

Our assumption regarding the voluntary factors seems most pressing for attention by all ELI officials. Earlier work by Ertem and Gokalp (2019), Ishitani (2016), and Russell and Jarvis (2019), for instance, focused heavily on school, college, and university students. We encourage future studies to address theoretical aspects of ELI student attrition. There are few well-established theoretical bases to explain ELI student attrition decisions. In general, there is a lack of qualitative studies on student attrition in Iran (Ashghali-Farahani et al., 2017). It speaks to the necessity of more qualitative research in this substantive area of inquiry. Research like Grounded Theory (GT) study sounds appropriate as little is known about this phenomenon in ELI settings. GT can help generate an explanatory theory that unveils a range of issues about attrition decisions in ELIs. As the literature shows, this potential area of research remains unexplored in the local EFL context. In this study, we adopted and tested some predetermined constructs of the student attrition model of Bennett (2003), and we acknowledge that this may limit the investigation and its findings. Although such theoretical models might be utilized as theoretical lenses to guide initial investigations, we recommend employing more exploratory methodologies to gain a deeper understanding of attrition. This kind of investigation can help close the research gap.

مراجع

Ashghali-Farahani, M., Ghaffari, F., Oskouie, F., & Zagheri Tafreshi, M. (2017). Attrition among Iranian nursing students: A qualitative study. Nurse Education in Practice, 22(1), 98-104. https://doi.org/10.1016/j.nepr.2017.01.002

Bean, J. P. (1980). Dropouts and turnover: The synthesis and test of a causal model of student attrition. Research in Higher Education, 12(2), 155-187. https://doi.org/10.1007/BF00976194

Beare, K. (2020). How many people learn English? ThoughtCo. https://www.thoughtco.com/how-many-people-learn-english-globally-1210367

Beebe, R. (2014). Factors contributing to student attrition in private schools. [Doctoral dissertation, Walden University]. ProQuest Dissertations & Theses. https://www.proquest.com/

Bennett, R. (2003). Determinants of undergraduate student dropout rates in a university business studies department. Journal of Further and Higher Education, 27(2), 123-141. https://doi.org/ 10.1080/030987703200065154

Christo, Z., & Oyinlade, A. O. (2015). Factors of student attrition at an urban university. International Journal of Humanities and Social Science, 9(1), 9-22.

Daka, H., & Changwe, R. (2020). An exploration of education quality in the light of the grade point average and examination attrition rate. International Journal of Humanities Social Sciences and Education, 7(6), 196–207. http://doi.org/10.20431/2349-0381.0706021

Eccles, J. S., Adler, T. F., Futterman, R., Goff, S. B., Kaczala, C. M., Meece, J. L., & Midgley, C. (1983). Expectancies, values, and academic behaviors. In J. T. Spence (Ed.), Achievement and achievement motivation (pp. 75–146). W. H. Freeman.

Ertem, H. Y., & Gokalp, G. (2019). Role of personal and organizational factors on student attrition from graduate education: A mixed-model research. Journal of College Student Retention: Research, Theory & Practice, 23(4), 903-928. https://doi.org/10.1177/1521025119881391

Gale, J., Ooms, A., Newcombe, P., & Marks-Maran,  D. (2015). Students' first-year experience of a BSc (Hons) in nursing: A pilot study. Nurse Education Today, 35(1), 256–64. https://doi.org/10.1016/j.nedt.2014.08.016

Gento, S. & González, R. (2014). The axiological basis for a curriculum design in educational institutions of quality. In D. B. Napier (Ed.), Qualities of Education in a Globalised World, (pp. 131–148). Sense Publishers.

Gholaminejad, R. & Raeisi-Vanani, A. (2021). English language teaching in Iranian mainstream schools: Pedagogical, societal and government policy environments. Issues in Educational Research, 31(1), 111-129.

Gholami-Zafarani, S. I., Kasaian, S. A., & Abdolmanafi-Roknis, S. J. (2015). Language learners' attrition rate: A case study in ILI. Research Journal of English Language and Literature, 3(3), 9-19.

Gregersen, T., & Khateeb, A. A. (2022). Motivation contagion: The reciprocal influence of language teachers and learners. In A. H. Al-Hoorie & F. Szabó (Eds.). Researching language learning motivation: A concise guide (pp. 153–164). Bloomsbury Academic.

Hart, C. (2012). Factors associated with student persistence in an online program of study: A review of the literature. Journal of Interactive Online Learning, 11(1), 19-42.

Ishitani, T. T. (2016). Time-varying effects of academic and social integration on student persistence for first and second years in college. Journal of College Student Retention: Research, Theory & Practice, 18(3), 263-286. https://doi.org/10.1177/1521025115622781

Mohammadian-Haghighi, F. & Norton, B. (2016). The role of English language institutes in Iran. TESOL Quarterly 51(2), 1-10. https://doi.org/10.1002/tesq.338

Moharami, M. & Daneshfar, S. (2021). The impacts of learning English on Iranians' everyday life: An ethnographic example from Piranshahr. Issues in Educational Research, 31(4), 1156-1174.

Qvortrup, A., & Lykkegaard, E. (2022). Study environment factors associated with retention in higher education. Higher Education Pedagogies, 7(1), 37-64. https://doi.org/10.1080/23752696.2022.2072361

Russell, L., & Jarvis, C. (2019). Student withdrawal, retention, and their sense of belonging; their experience in their words. Research in Educational Administration & Leadership, 4(3), 494- 525. https://doi.org/10.30828/real/2019

Ryan, R. M., & Deci, E. L. (2017). Self-determination theory: Basic psychological needs in motivation, development, and wellness. Guilford Publications.

Ryan, J. M., Potier, T., Sherwin, A., & Cassidy, E. (2021). Identifying factors that predict attrition among first-year physiotherapy students: A retrospective analysis. Physiotherapy, 110, 26-33. https://doi.org/10.1016/j.physio.2017.04.001

Shcheglova, I., Gorbunova, E., & Chirikov, I. (2020). The role of the first-year experience in student attrition. Quality in Higher Education, 26(3), 307-322. https://doi.org/10.1080/13538322.2020.1815285

Schnettler, T., Bobe, J., Scheunemann, A., Fries, S., & Grunschel, C. (2020). Is it still worth it? Applying expectancy-value theory to investigate the intraindividual motivational process of forming intentions to drop out from university. Motivation and Emotion, 44(4), 491-507. https://doi.org/10.1007/s11031-020-09822-w

Seebruck, R. (2021). How teacher rotation in Japanese high schools affects the clustering of teacher quality: Comparing the distribution of teachers across public and private education sectors. Education Policy Analysis Archives, 29(91). https://doi.org/10.14507/epaa.29.5362

Strelan, P., Osborn, A., & Palmer, E. (2020). Student satisfaction with courses and instructors in a flipped classroom: A meta‐analysis. Journal of Computer Assisted Learning, 36(3), 295-314. https://doi.org/10.1111/jcal.12421

Stronge, J. H. (2018). Qualities of effective teachers (3rd ed.). ASCD.

Tinto, V. (1975). Dropouts from higher education: A theoretical synthesis of recent research. Review of Educational Research, 45(1), 89-125. http://doi.org/10.3102/00346543045001089

Ugwu, D. N., & Adamuti-Trache, M. (2019). Post-graduation plans of international science and engineering doctoral students attending US universities. Journal of International Students, 7(1), 1-21.https://doi.org/10.32674/jis.v7i1.242

Yates J. (2012). When did they leave, and why? A retrospective case study of attrition on the Nottingham undergraduate medical course. BMC Medical Education, 12, 1-7. https://doi.org/ 10.1186/1472-6920-12-43

Young Journalists Club. (2021). YJC.News. Statistics of students dropping out of school. https://www.yjc.news/fa/news/7956639

Zając,T., & Komendant-Brodowska, A. (2018). Premeditated, dismissed and disenchanted: Higher education dropouts in Poland. Tertiary Education and Management, 25(1). 1-16. https://doi.org/10.1007/s11233-018-09010-z