The Roles of Self-esteem in the Relationship between Emotional Intelligence (EI) and Adjustment among International Students in Southern Thailand

The Roles of Self-esteem in the Relationship between Emotional Intelligence (EI) and Adjustment among International Students in Southern Thailand Yejin Kim Prince of Songkla University, Hat Yai, Songkhla, 90110 Thailand, okngood@gmail.com Wanchai Dhammasaccakarn Prince of Songkla University, Hat Yai, Songkhla, 90110 Thailand Isara Tongsamsi Songkhla Rajabhat University, Muang, Songkhla, 90000 Thailand


Introduction
The concept of adjustment among international students has evolved into a multi-faceted human behavioural phenomenon (Ward & Kennedy, 1993;Mustaffa & Ilias, 2013) since the first study by Lysgaard, who introduced the U-curve theory, in 1955. Academic, social, and psychological adjustment have been suggested to be indicators that construct the concept of adjustment (Baker & Siryk, 1999). Consequently, there has been tremendous academic interest in identifying cognitive and noncognitive predictors of these three indicators in higher education (Rosenthal et al., 2008;Mesidor & Sly, 2016).
Identifying the personal traits, predicting the three indicators in the stream of 'the problem-solving approach' (Vasilopoulos, 2016, p. 284) has been of critical importance to researchers. Educational, psychological, social, and behavioural researchers have therefore stressed the predictive roles of emotional intelligence (EI) and selfesteem on adjustment.
Nevertheless, to our knowledge, the relationship between EI and self-esteem coexisting with individuals in impacting international students' adjustment have been riddle and agreement relating each roles of both also less unanimous. Few studies have simultaneously examined the roles of self-esteem in the underlying mechanism between EI and the three indicators of adjustment. In addition, there is little academic interest on adjustment of Asian international students studying in Asian countries and on this topic. www.scholarhub.ui.ac.id/hubsasia

Adjustment among International Students
International students encountering challenges or difficulties after arriving in new environments of host countries 'have to do with the process of achieving harmony among the individual and the environment by changing individual's knowledge, attitude and emotion about his/her new environment' (Hannigan, 1990, p. 91), namely, adjustment.
According to Tinto's student integration theory (1993), integration for students in both academic and social communities is essential to reduce their rates of dropout from institutions. A recent study by Chrysikos et al. (2017) confirmed the validity of the model in the UK context, underlining the importance of both academic and social integration. In addition, Berry (2005) stressed psychological health in the acculturation model, explaining the process of adjusting by overcoming stressors, such as depression, anxiety, homesickness, and identity confusion. According to the model, international students experiencing psychological difficulties due to conflicts in social and cultural differences between two countries need to establish a coping system.
As seen in Tinto's theory (1993) and Berry's model (2005), academic, social, and psychological adjustment have emerged as three crucial aspects for understanding international students' adjustment, defined as a multifaceted phenomenon (Aycan, 1997;Berry, 2005). In addition, it is commonly accepted that multiple personal resources contribute to the three outcomes (Mesidor & Sly, 2016). The importance of EI (Roy et al., 2013;Cazan & Năstasă, 2015;Priya & Panchanatham, 2014;Thompson, 2018) and self-esteem (Mohamed, 2012;Wang & Heppner, 2012;Lopez & Bui, 2014;Liao & Wei, 2014) in international students' adjustment were stressed in numerous studies. Therefore, to seek a more complete construct of international students' adjustment, we proposed three outcome variables, indexed as academic, social, and psychological adjustment, as indicators of adjustment impacted by two personal resources, indexed as EI and self-esteem, among international students.
Academic adjustment. The prime purpose of international students leaving home to go to host countries is to obtain academic achievements. Different academic systems, relationships with professors and academic colleagues, and new academic requirements are troublesome to international students (Mesidor & Sly, 2016), causing stress, which is considered a negative emotional, cognitive, behavioural and psychological process (Lazarus & Folkman, 1987). The levels of adjustment in his/her attitudes, emotions and behaviours in new academic environments will impact the life satisfaction and academic achievements among international students.
Social adjustment. The complementary processes of desocialization and socialization are crucial among international students (Mudhovozi, 2012). International students have to keep constant social contact with others while studying in host countries that possess different social and cultural values and manners (Thompson, 2018). Culture shock due to the differences between two sociocultures remained one of the biggest challenges hindering the adjustment of international students in the United Kingdom in a recent study by Eze and Inegbedion (2015).
Thus, the greater their ability to cope with social and cultural requirements and demands, such as participating in group work, communicating well, forming friendships with other students, and joining clubs and social networks, the better their adjustment will be (Baker & Siryk, 1999). For instance, better social/socio-cultural adjustment resulting from understanding individual/social norms and behavioural patterns in communication was indicated as an important factor reducing misunderstanding in the Intercultural Adaptation Model (IAM) by Cai and Rodríguez (1997). In addition, as seen in Tinto's theory, the significant positive correlations in academic achievement and social adjustment were also indicated in the work in Malaysia by Khan et al. (2015).
Psychological adjustment. Studying in higher education increases the significance and complexity of psychological problems (Adlaf et al., 2001;Brown & Blankson, 2013). Culture shock, conceptualized by Oberg (1960), is a set of personal and emotional reactions due to unfamiliarity in new academic and socio-cultural environments. A study by Rajab et al. (2014) reported that homesickness and perceived hate were the main psychological stressors to international students in Malaysia.
In addition, there were observations of significant differences between life satisfaction of international students and domestic students in Australia (Skromanis et al., 2018); international students were found to have lower satisfaction, for instance, more smoking, gambling, and illicit drug problems were present. Consequently, maladjustment in the crisis will result in hardships in academic performance and socialization among other students. Conversely, as seen in a study on homesickness among international students in Thailand (Thomas, 2018), academic and social integration negatively and moderately impacted homesickness.

Emotional intelligence
The emotional intelligence theory  has evolved into an important predictor of human success in education (Goleman,1995;Schutte et al., 1998;Mohamed, 2012) and in personal satisfaction (Priya & Panchanatham, 2014). An understanding of one's own and others' emotions affects humans' perceptions and insight and is vital for international students (Chaffey et al., 2012). Salovey and Mayer (1990), conceptualized EI as a subset of social intelligence and reported that individuals with higher EI can more effectively solve life problems and better adjust to accomplish important tasks, when regarding individuals as emotional beings.
Based on the Tinto's theory (1993), international students are required to adjust to different social and intellectual norms in new environment. In other words, studying in new culture may trigger 'acculturative stress', requiring psychological acculturation and adjustment/adaptation to the new culture as stressed by Berry's model (2005). Adjustment is affected by personality variables. Thus, there is tremendous interest in EI as a predictor of adjustment. (Zeidner et al., 2008).
Arguably, EI, defined as 'people's self-perception of their own emotional abilities and skills, personality characteristics and behavioural dispositions influencing their ability to cope successfully with environmental demands and pressures' (Cazan & Năstasă, 2015, p. 1574, acts one of the important personal variables impacting international adjustment. Considering the language barrier (Riggio, 2010), the lack of English language skills and/or social language proficiency (Andrade & Evan, 2009) and constant academic and social contact with international students (Thompson, 2018), it is reasonable that EI has been shown to be strongly associated with academic and social adjustment (Roy et al., 2013;Cazan & Năstasă, 2015;Priya & Panchanatham, 2014). Positive correlations between EI and all aspects of communication skills, such as interpersonal/group communication and public speaking, were also indicated (Marzuki et al., 2015). Furthermore, EI acts as a predictor of psychological adjustment, which is related to one of two important categories in the acculturation model (Ramsay et al., 1999;Berry, 2005), such as life satisfaction and psychological well-being (Ward & Kennedy, 1993;Berry, 2005;Cazan & Năstasă, 2015;Jayalakshmi & Magdalin, 2015). Therefore, we propose that EI predicts academic, social, and psychological adjustment among international students in southern Thailand.

Self-esteem as a mediator
A review of the literature established self-esteem as a potential mediator between EI and the three indicators of international students' adjustment, originating from the significant positive correlations between EI and selfesteem (Mattanah et al., 2004;Mehmood & Gulzar, 2014;Bibi et al., 2016) and its mediating role between personal traits and life satisfaction (Kwan et al., 1997). Liu et al. (2014) reported self-esteem's partial mediating effects in the relationship between trait resilience and life satisfaction among Chinese students. They also indicated its full mediation effects between trait resilience and psychological distress. Rey et al. (2011) and Runcan and Iovu (2013) also found that it played a mediating role in the relationship between EI and life satisfaction.

The hypothesized framework
This investigation is grounded within the framework of Tinto's Student Integration Theory (1993), the Emotional Intelligence Theory , and the Acculturation Model (Berry, 2005). The aim of the study is to examine the underlying mechanism through which EI affects academic, social, and psychological adjustment among international students in southern Thailand. Based on the above review of the existing literature, we propose that: (1) EI is positively correlated with self-esteem; (2) self-esteem is positively correlated with academic, social, and psychological adjustment; and (3) self-esteem mediates the association between (a) EI and academic adjustment, (b) EI and social adjustment, and (c) EI and psychological adjustment.

Participants
Though one hundred fifty-six international students responded, the data from five participants were dropped from the data pool due to unfaithful responses (1 case) and being out of targeted group (4 cases). The data of the remaining one hundred fifty-one international graduate students enrolled in programmes at Hat Yai campus of Prince of Songkla University in southern Thailand was used in this research. The research focused on first-and second-year MA programme students and first to third year Ph.D. programme students who attended classes on campus regularly at one campus of university in the same academic semester. Eighty-six MA programme students (57%) and sixty-five Ph.D. programme students (43%) majoring in various curriculums participated in the research: first year (31.8%) and second year (25%) of MA programme; first year (13.9%), second year (19.2%), and third year (9.9 %) of Ph. D. programme. After describing the purpose and confidentiality of the research, their voluntary consent for participation was obtained. Fortythree international students responded by online and one hundred eight by paper-based questionnaires.

Measures
Schutte Self Report Emotional Intelligence Test (SSEIT; Schutte et al., 1998). SSEIT, one of the most widely used scales for measuring EI, was used to assess EI. The English version of SSEIT was adopted for the international students. The scale consists of 33 items (e.g., "I am aware of my emotions as I experience them", "I seek out activities that make me happy", "when I am in a positive mood, solving problems is easy for me") containing all portions of the EI as conceptualized by Salovey and Mayer (1990). For each statement, participants were asked to rate how much they agree on a five-point Likert scale ranging from 1 = "strongly disagree" to 5 = "strongly agree". The internal consistency reliability () in this research was 0.87.
Rosenberg Self-Esteem Scale (RSES: Rosenberg, 1965). RSES ( = 0.81), which is the most widely used measure of the construct (Blacscovich & Tomaka, 1991), was adopted to evaluate self-esteem. It consists of 10 items (e.g., "I take a positive attitude towards myself"), consisting of two factors, self-liking and self-competence (Tafarodi & Milne, 2002). The participants rate each item on a fourpoint Likert scale ranging from 1 = "strongly disagree" through 4 = "strongly agree". One half of all items are stated positively, and the other half are stated negatively.

Student Adaptation to College Questionnaire (SACQ:
Baker & Siryk, 1999). The three outcome variables were evaluated using the SACQ ( = 0.93), devised as a measurement tool for conceptualization with 67 items (Baker & Sirky, 1999) consisting of four subscales: 1) academic adjustment (24 items, e.g., "I am finding academic work at college difficult"), 2) socio-cultural adjustment (20 items, e.g., "I feel that I fit in well as part of the college environment"), 3) personal-emotional adjustment (15 items, e.g., "I have felt tired much of the time lately"), and 4) institutional attachment (15 items, e.g., "I am pleased now about my decision to go to college"). The participants rated each question on a nine-point Likert scale ranging from "applies very closely to me" to "doesn't apply to me at all", with values from 1 to 9 in 34 negatively keyed items or from 9 to 1 in 33 positively keyed items. The present study adopted three subscales in the questionnaire: academic adjustment ( = 0.87) for academic adjustment variables, socio-cultural adjustment ( = 0.81) for social adjustment variables, and personal-emotional adjustment ( = 0.85) for psychological adjustment variables.

Data analyses
Initially, according to the recommendations of Kenney (2018), the four steps of mediation analyses were conducted to test the mediation model for all students as a preliminary analysis for confirming the possibility of the mediation role of self-esteem. Second, with the positive results from the previous test, evaluation of the fit for the hypothesized model was conducted using maximum likelihood estimation. In the current research, the criteria for evaluating the fit of the models consisted of the chisquare statistic (χ 2 ), the χ 2 /df ratio, the Comparative Fit Index (CFI), and the Root Mean Square Error of Approximation (RMSEA); it was considered a satisfactory model if χ 2 /df ≤ 2, CFI > 0.90 and if the RMSEA was below 0.08 (Kelloway, 1998: Byrne, 2010. Finally, testing the mediation estimation in the path model was performed with standardized estimation of path coefficients and the significance of the direct and indirect effects' p value by bootstrapping (Preacher & Hayes, 2008).
The data were analysed using the SPSS version 25 and Amos 25.0 programmes. Missing data were replaced by the mean of the full scales based on the manual of SACQ (Baker & Sirky, 1999) and data imputation approach (Arbuckle, 2017). The research ethics was approved by the Centre for Social and Behavioural Sciences Institutional Review Board, Prince of Songkla University.

Preliminary analyses
The means, standard deviations, and intercorrelations between the research variables for all students are presented in Table 1. EI, the predictor, was positively correlated with the three outcomes: academic adjustment (r = 0.45, p < 0.01), social adjustment (r = 0.28, p < 0.01), and psychological adjustment (r = 0.31, p < 0.01), as well as the mediator, self-esteem (r = 0.47, p < 0.01), confirming the positive results of the first two steps suggested by Kenney (2018). In addition, the other correlations were statistically significant; self-esteem was also positively correlated with academic adjustment (r = 0.67, p < 0.01), social adjustment (r = 0.35, p < 0.01), and psychological adjustment (r = 0.57, p < 0.01); the three outcomes had positive correlations, r = 0.52, p < 0.01, r = 0.65, p < 0.01, and r = 0.30, p < 0.01, respectively, academic and social adjustment, academic and psychological adjustment, and social and psychological adjustment, indicating statistical significance of all correlations among the variables at the level below 0.70. Thus, the multicollinearity was not violated.
The results from the multivariate regression based on the last two steps by Kenney (2018) are presented in Table 2 for all participants. A mediation effect of self-esteem in the association between EI and academic adjustment was found; self-esteem (semipartial r = 0.58, p < 0.001) had a significant semipartial correlation with academic adjustment, and the semipartial correlation (semipartial r = 0.21, p < 0.01) between EI and academic adjustment was smaller than the zero-order correlation (r = 0.45). In the relationship between EI and social adjustment, the semipartial effect of self-esteem was found to be significant (semipartial r = 0.25, p < 0.01), and the semipartial correlation (r = 0.14, p = 0.08) between the predictor and the dependent variable was smaller than zero-order (r = 0.28). Last, the mediation effect of selfesteem in the association between EI and psychological adjustment was also indicated by the significant semipartial correlation (semipartial r = 0.51, p < 0.001) between self-esteem and psychological adjustment and the smaller semipartial correlation (r = 0.05, p < 0.05) than zero-order correlation (r = 0.31). Thus, the mediation test in the path analysis was followed.

Mediation analyses Model estimation.
Due to the unsatisfactory model fit for all students, χ 2 (3, N = 151) = 57.54, (p < 0.001), χ 2 /df = 19.18, CFI = 0.79, RMSEA = 0.35, the fully mediated hypothesized model required modification: eliminating of the two direct pathways from EI to social adjustment and psychological adjustment based on the modification indices and the result indicating nonsignificant predictions in the pathways, respectively, β = 0.15, p = .07 and β = 0.05, p = 0.50, and adding two non-direct paths between error variances of academic adjustment and social adjustment, and psychological adjustment. Consequently, the re-specified model that eliminated the direct pathways between EI and social adjustment and psychological adjustment was re-run, and was indicated to be a satisfactory model, χ 2 (3, N = 151) = 6.15, (p = 0.10), χ 2 /df = 2.05, CFI = 0.99, RMSEA = 0.08.

Mediational analyses.
The Boot estimation procedure was used to test the significance of mediation effects for self-esteem. Direct, indirect and total effects were tested based on standardized beta coefficients and bootstrapped standard errors: 1000 bootstrap samples and 95% biascorrected confidence intervals (CI) (Shrout & Bolger, 2002). Figure 1 demonstrates the results of direct effects among the variables with p values based on bootstrapping (1000 samples) and Squared Multiple Correlations (SMC). Table 3 shows the results of indirect effects and total effects in the direct model and the final mediation.
Thus, the associations between EI and social adjustment and between EI and psychological adjustment can be considered fully mediated by self-esteem. However, self-esteem may act as a partial meditator in the relationship between EI and academic adjustment. In addition, self-esteem predicted the three variables much more strongly than EI. The indirect effects of self-esteem on the total effect in the association between EI and academic adjustment was 70.73%.

Discussion and Conclusion
Having international students is of benefit to the institutions, because of their contributions to diversity and internationalization, with academic, cultural, and financial influences in host countries (Wu et al., 2015). The present study investigated the important role of self-esteem as a mediator between EI and academic, social, and psychological adjustment among international students in southern Thailand.
As predicted, the results of the mediation path analysis for all the participants showed that self-esteem fully mediates the relationships between EI and social and psychological adjustment, accounting for approximately 12% and 33% of the variance, respectively. In addition, the partial mediator roles of self-esteem in the association between EI and academic adjustment were indicated. The results of the present work are consistent with a previous study by Rey, Extremera, and Pena (2011), who reported that selfesteem acts as a mediator between emotional ability and life satisfaction. Similarly, Ruvlacaba-Romero et al. (2017) indicated self-esteem as a mediating variable between EI and emotional clarity, emotional repair and life satisfaction.
Of note, self-esteem acts as a momentous mediator in the underling mechanism between EI and the three indicators of adjustment among international students. Self-esteem is estimated to affect 71% in the path analyses, indicating its huge role in increasing academic adjustment among international students. Additionally, EI did not exert direct and decisive roles in social and psychological adjustment.
Noteworthily, the concept of constructs on EI is still unclear. Zeidner et al. (2008) documented the on-going debates, categorizing into three issues, namely, conceptualization, assessment and applications; Any concept of EI constructed based on self-reported questionnaires has not reached to 'a satisfactory justification or definitional framework' (p. 74; Mayer et al., 2004); The effectiveness of programmes originated for enhancing EI is also questionable due to 'the lack of the theoretical and methodological rationales' (p. 74). The critical points on the concept of EI may provide clues to understand the findings of the research, indicating less exertion of EI on the three outcomes in the final mediation model than self-esteem.
Thus, enhancing international students' self-esteem in the association between EI and adjustment may impact the levels of international students' academic, social, and psychological adjustment, leading to the reduction of dropout rates, based on the Tinto's student integration theory (Tinto, 1993), and to the better establishment of their coping system in the acculturation model (Berry, 2005), while studying in Thailand. In addition, the results are also consistent with the assertion that self-esteem is one aspect of the four dimensions predicting job satisfaction and performance in the Core Self-Evaluations (CSE: Judge et al., 1997;Judge et al., 1998). Self-esteem may assume a critically important role in the relationship between EI and adjustment.
We believe that the results obtained in this study will allow stakeholders to develop projects to enhance international students' adjustment in Thailand, using English language in primarily academic settings and the Thai language in most social circumstances. Although previous studies strongly stressed the impacts of EI on adjustment among international students in higher education, this study indicated that it does not act as an independent personal trait influencing adjustment. In other words, personal cognitive and/or noncognitive traits constitute a certain mechanism which impacts an individual's adjustment.
Based on these results, we suggest two practical projects or programmes to improve students' self-esteem and EI for better adjustment (Telbis et al., 2014): (1) A community acceptance programme providing useful information and practical help related to new academic and social opportunities; this programme can be modelled on social support, which has been proven to contribute to global self-esteem among Chinese students (Kong et al., 2013), it can improve cultural adjustment and life satisfaction among international students, as seen in Thailand (Rujiprak, 2016), and it can assist with psychological adjustment, as seen in Malaysia (Lashari et al., 2018); (2) Academic and/or social language programmes aiming to improve communication skills in academic and social environments. The important roles of both academic and social language proficiency in enhancing the adjustment of international students were stressed in one ASEAN country, as Malaysia has two different language environments (Mustaffa & Ilias, 2013).
Nevertheless, several limitations of the present study should be taken into consideration. First, this study relied on a small sample size at one university in southern Thailand; thus, the results need to be extended to larger populations and to other universities. Second, the study is cross-sectional in nature, providing evidence rather than causation; prospective and longitudinal research should be conducted to determine the causation between the relationships among the variables. Third, only one measure of EI was adopted, and it did not distinguish between ability EI and trait EI. Therefore, further research is necessary to adopt more sensitive measures, elucidating each of the two aspects. Finally, only one personal trait, EI, was tested as a predictor. Future studies, therefore, should employ various personal traits as predictors.