Cytalist

100 specifically focusing on sporting apparel brands, which has become a global trend for young consumers, particularly in Thailand. This study’s objectives were to (i) investigate the relationship of global brand awareness and global brand image as they contribute to the understanding of brand knowledge; (ii) examine the impact that global brand image has on self-identification with global consumer culture for global apparel brands; and (iii) identify the mediation effect of self-identification with global consumer culture on the relationship of global brand image and brand loyalty. The findings of this study should enable marketers to understand the process by which young consumers in Thailand become loyal to global sporting apparel brands. Literature Review Brand Knowledge This study is concerned with the threat of increased competition in global markets, as well as the threat from counterfeit products. It has been suggested by (Weisheng & Keat, 2016) that if consumers are brand conscious, their intention to purchase counterfeit products will become a less likely outcome. As the study of (Weisheng & Keat, 2016) focused on the effect of brand consciousness on purchase intentions toward counterfeit products, it is theorized in this study that the affect should be inverse, suggesting that brand consciousness should lead to a higher level of intention to purchase authentic sporting good brands. In other words, more consumer knowledge about a brand should lead to more favorable outcomes. Keller (1993) introduced the concept of customer-based brand equity, explaining consumer responses to marketing strategies based on their brand knowledge. The notion of brand knowledge is “consisting of a brand node in memory to which a variety of associations are linked” (Keller, 1993, p. 3). Each node represents the information communicated through verbal, visual, abstract, and contextual nature that has been stored in the memory of consumers (Keller, 2008). As brand knowledge is an important aspect in customer-based brand equity, the concept itself can be explained through the notions of brand awareness and brand image (Keller, 1993). The concept of brand awareness can evaluate the degree to which consumers can recognise and recall brands While brand recognition is consumer ability to identify and differentiate a brand from other competitive brands, brand recall is the ability to bring the brand to mind with or without the use of any cues given (Keller, 1993). Nevertheless, both brand recognition and brand recall comprise consumer brand awareness, which can be disseminated through various marketing communication tools (Macdonald & Sharp, 2003). Another aspect of brand knowledge is brand image, which is defined as “perceptions about a brand as reflected by the brand associations held in consumer memory” (Keller, 1993, p. 3). In other words, brand associations such as product quality, brand personality, and organisational associations can be stored in the minds of consumers. In fact, brand image is an essential part in nurturing a strong brand that makes it easy for consumers to differentiate their product brands from those of competitors (Aaker, 1996; Kapferer, 2004). According to (Keller, 1993), brand awareness and brand image are related in a sequential manner; that is, brand awareness exists before consumers store the information they have learnt in their memories. A study by (Esch, Langner, Schmitt, & Geus, 2006) examined the relationship between brand awareness and brand image, and the result was found to be significantly positive. Thus, this research hypothesises that: H1: Brand awareness will influence brand image of global sport apparel brands. Self-Identification The concept of self-identification has been derived from social identity theory (SIT) defined as “an individual’s self-concept which derives from his knowledge of his membership of a social group (or groups) together with the emotional significance attached to that membership” (Tajfel, 1974, p. 69). This concept generally explains the reaction of how people view themselves as being part of an “in-group” as opposed to an “out-group.” Viewing oneself as part of an in-group tends toward favoritism to the in-group, and discrimination against the out-group (Tajfel, 1974). As a result, a person

101 will later categorise herself/himself as belonging to a particular social group, which can be explained by self-categorisation theory (SCT) (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). Importantly, before individuals can categorise themselves in a social group or category, they must understand their own self-concepts, which can be defined as “the set of cognitive representations of self available to a person” (Turner, et al., 1987, p. 44). The adaptation of this concept has been widened to explain the behaviours of people within an organization (Ashforth & Mael, 1989). For instance, the concept of organizational identification, rooted in the SIT and SCT (Riketta, 2005), explains “the perception of oneness with or belongingness to a group, involving direct or vicarious experiences of its success and failures” (Ashforth & Mael, 1989, p. 34). In other words, organizational identification can explain how an individual views the success and failure of the organization as one’s own. A similar concept was also introduced, referred to as “brand identification”, in order to explain the relationship between consumers and the brand — a consumer’s identification with the brand (Kim, Han, & Park, 2001). In general, identification is the degree of belongingness to a group of any form (i.e., organization, brand). The study of brand identification has been investigated previously in the automotive industry by (Kuenzel & Halliday, 2010), who examined how various factors such as reputation and brand personality congruence have an effect on brand identification. According to their findings, evidence suggests that both factors of reputation and brand personality congruence contribute significantly to consumer identification with the brand and result in favorable outcomes. Although previous studies sought to find relationships between the perception of how others view a brand and consumer identification with it in the automotive industry, this study suggests that if consumers identify with certain objects, favorable outcomes will follow. Thus, this study conceptualises self-identification to be related to global consumer culture. To elaborate, consumers who self-identify with global consumer culture tend to view themselves as having similar fashion tastes with consumers around the world, and they will favor international brands more than local brands (Frank & Watchravesringkan, 2016). Thus, this study conceptualises that there is a relationship between brand image and self-identification with global consumer culture, and hypothesises that: H2: Brand image will influence self-identification with global consumer culture for global sport apparel brands. Brand Loyalty In the field of marketing, extant research has investigated consequences of brand identification and found that brand loyalty is an outcome. For instance, the study of (Kim, et al., 2001) investigated the relationship of brand identification and brand loyalty for high-technology products such as cellular phones. Research by (Kuenzel & Halliday, 2010) conducted in Germany found that in order for car owners to be loyal to a brand, they needed to first identify with the brand. However, in this particular study, brand identification was incorporated as a mediator to a relationship beween brand reputation and brand loyalty (Kuenzel & Halliday, 2010). This study proposes that selfidentification with global consumer culture also plays a mediating role in the relationship between brand image and brand loyalty. H3: The relationship between brand image and brand loyalty will be mediated by the degree of self-identification with global consumer culture for global sport apparel brands. To summarise all the proposed hypotheses investigated in this study, Figure 1 illustrates the research conceptual model accordingly. The theoretical concepts underpinning these relationships are explained in the previous section.

102 Figure 1. Research Conceptual Model Research Methodology In order to examine the proposed hypotheses, questionnaires with closed-ended scale type queries were distributed to college students at a private university in Thailand. The questionnaires asked about factors related to consumer responses toward global sporting apparel brands. A convenience sampling technique was employed. The instrument used in this research was based upon established scales including brand awareness, brand image, self-identification with global consumer culture, and brand loyalty. While a four-item scale of brand awareness was adopted from (Frank & Watchravesringkan, 2016), a five-item scale of brand image was adopted from (Rageh Ismail & Spinelli, 2012). Another established six-item scale used in this study measured self-identification with global consumer culture, asking about the degree to which respondents viewed themselves to identify with global consumer tastes for sporting apparel brands (Frank & Watchravesringkan, 2016). Finally, a three-item scale of brand loyalty was employed, asking about the willingness of consumers to continue their purchases of the chosen brand (Martínez, 2015). Likert scales ranging from 1 (strongly disagree) to 5 (strongly agree) were employed for each question in the questionnaire. The measurements used in this research are illustrated in Table 1, with an acceptable Cronbach’s alpha coefficient of 0.70 or higher indicating that these scales are reliable for replication (Nunnally & Bernstein, 1994): Table 1. Cronbach’s Alpha Coefficient Scores for Variables Variable Number of Items Cronbach’s Alpha Coefficients (α) Brand Awareness (Frank & Watchravesringkan, 2016) 4 0.89 Brand Image (Rageh Ismail & Spinelli, 2012) 5 0.89 Self-Identification (Frank & Watchravesringkan, 2016) 6 0.89 Brand Loyalty (Martínez, 2015) 3 0.84 After questionnaires were distributed to college students, 276 of them were filled in and returned. The data was analyzed using a number of statistical tools, include reliability tests, Pearson’s correlation coefficient, and regression analysis. The hypothesised relationships were tested using simple and multiple regression, and a specific method for mediation effect testing was employed. Prior to hypothesis testing, a reliability analysis was performed for each of the dependent and independent variables as illustrated in Table 2.

103 Table 2. Reliability Analysis of Variables Accordingly, the Cronbach’s alpha coefficients for all scales exceeded 0.80, which indicates excellent reliability, except for the dimension of employee concern, which obtained a Cronbach’s alpha coefficient of above 0.70. This is still considered to be acceptable according to Nunnally and Bernstein (1994), as illustrated in Table 2. Research Findings So far as the demographic characteristics of respondents was concerned, 73.6% of respondents were young Thai females between 18 and 24 years of age. This research investigated global sporting apparel brands, and found that 56.9% of young Thai consumers favor Adidas, followed by Nike (34.8%), spending on average no more than 3,000 Thai Baht per visit (73.9%). Most respondents (54.3%) purchase sporting apparel brands from department stores. Moreover, social media seemed to play an important role in communicating about the brands—73.6% of respondents received brand information through this channel—and it influenced respondents in making purchase decisions (35.5%). Table 3. Pearson’s Correlation Coefficient Prior to hypothesis testing, a Pearson correlation coefficient test was performed to examine for multicollinearity. It found that there was no substantial correlation between any of the predictors (R> 0.90) (Field, 2005), as illustrated in Table 3. Similarly as shown in Table 4, hypothesis testing on the relationship between brand awareness and brand image (H1) was performed using simple regression analysis. The results show that the predictors explained 38.3% of the variance (ß = 0.619, p < 0.001), which indicates a significant impact of brand awareness on brand image. Therefore, hypothesis H1 was accepted. Table 4. The Relationship between Brand Awareness and Brand Image (Dependent Variable) Variable R R 2 Adj. R2 Std. Error 0.619 0.383 0.381 0.463 Brand Awareness B SE B ß Sig. 0.562 0.043 0.619 0.000 Table 5 illustrates the steps in testing for mediation effect. Prior to examining mediation effect, a three-step approach is needed to detect the existence of mediation effect (Baron & Kenny, 1986). In Step 1, the relationship between brand image and brand loyalty was examined and the relationship was found to be significantly positive (ß = 0.161, p < 0.001). In Step 2, it was hypothesised that brand image has an effect on self-identification with global consumer culture, and a significant impact was found (ß = 0.279, p < 0.001). Thus, hypothesis H2 was accepted. Lastly, Step 3 tested the Variable Cronbach’s Alpha Coefficient (α) Brand Awareness 0.77 Brand Image 0.81 Self-Identification 0.88 Brand Loyalty 0.84 Brand Awareness Brand Image SelfIdentification Brand Loyalty Brand Awareness 1 Brand Image 0.619** 1 Self-Identification 0.366** 0.279** 1 Brand Loyalty 0.248** 0.161** 0.746** 1

104 relationship between self-identification with global consumer culture and brand loyalty, and once again the results showed that self-identification had a significant impact on brand loyalty (ß = 0.746, p < 0.001). Thus, the results found that there are significant relationships at the 0.05 level in steps 1- 3, and so mediation effect presumably exists. Step 4 of the model indicated that the relationship between brand image and brand loyalty is fully mediated by self-identification with global consumer culture, as brand image is no longer significant when self-identification is controlled (Baron & Kenny, 1986; Frazier, Tix, & Barron, 2004). Thus, hypothesis H3 was accepted. Table 5. The Mediator Effects of Product Attitudes Testing Steps in Mediation Model B SE B 95% CI ß Testing Step 1: Dependent: Brand Loyalty Predictor: Brand Image 0.254 0.094 0.069, 0.439 0.161** Testing Step 2: Dependent: Self-Identification Predictor: Brand Image 0.318 0.066 0.188, 0.448 0.279** Testing Step 3: Dependent: Brand Loyalty Predictor: Self-Identification 1.033 0.056 0.924, 1.143 0.746** Testing Step 4: Dependent: Brand Loyalty Mediator: Self-Identification 1.053 0.058 0.939, 1.167 0.761** Predictor: Brand Image -0.081 0.066 -0.211, 0.049 -0.051 Discussion According to the research findings, all of the proposed hypotheses have been statistically satisfied. First of all, it was found that in order for consumers to assign any image to brands, they need to be aware of brand information and able to differentiate one brand from another. Thus, this research has confirmed that the concept of brand knowledge included both brand awareness and brand image (Keller, 1993), and their relationship is a causal one (Esch, et al., 2006). In other words, brand awareness will lead to brand image; the stronger the awareness created, the more solid it is for brand image. This suggests that global sporting apparel brands must take an appropriate approach when designing their marketing communication strategies. In other words, if communicated information about the brand is indistinct, it is more likely that consumers will not have a clear image of the brand. Moreover, as mentioned previously, having little brand consciousness or awareness will damage the brand (Weisheng & Keat, 2016) because customers will not know the difference between authentic and counterfeit products. This is the point where consumers turn to conterfeit products because the price is more attractive, and no other unique selling proposition has been clearly communicated. Another important factor to consider about brand communication relates to selfidentification with global consumer culture; this research found that brand image will influence selfidentification with global consumer culture. According to a previous study by (Kuenzel & Halliday, 2010), although conducted in a different industry, it was found that self-identification is vital to favorable consumer responses. In this study, it was found that as long as consumers identified themselves with the taste in fashion of global consumers, sporting apparel brands could enjoy positive firm performance. More specifically, self-identification with global consumer culture tended to play an important role as it had a mediation effect between brand image and brand loyalty. In other words, in order for consumers to be loyal to a global brand, they must see themselves as being part of the in-group of global consumer culture— and especially favor global brands/products more than local ones. This suggests that even if an individual assigned a positive image to the global brand, but does not view one self to be a global consumer, that individual may not be a loyal customer to that particular brand.

105 Implications, Limitations, and Future Research This study contributes to the extant research which previously attempted to examine the mediation effect of self-identificaiton—referred to in other research reports as brand identification (Kim, et al., 2001; Kuenzel & Halliday, 2010). In both of these studies, different predictors were selected (i.e., brand personality, brand reputation), whilst this study focused on brand knowledge, particularly brand image. Thus, future reseach should replicate this model in different contexts globally and domestically to generalise research findings. Moreover, the findings from this research also contribute to practitioners in that marketers can incorporate the concept of brand knowledge when designing their marketing communication strategies, as it is a key factor in generating brand loyalty among consumers. Another aspect that needs to be taken into consideration is selfidentification with global consumer culture, which in this case mediates the relationship in this study. The marketer should find ways to enhance self-identification of consumers globally. The use of social network sites (SNSs) might be useful to boost self-identification with global consumer culture, especially for young consumers who are known as digital natives. The researcher acknowledges that this was drawn from student samples, although the targeted sample group is young consumers. Thus, future research projects should consider broadening data collection to include respondents from different generations. Future researchers should also consider conducting similar research in different countries to enhance cross-national comparisons. About the Author Lokweetpun Suprawan is an Assistant Professor in Marketing of the School of Business Administration at Bangkok University, Thailand. References Aaker, D. (1996). Measuring Brand Equity Across Products and Markets. California Management Review, 38(3), 102-120. Ashforth, B., & Mael, F. (1989). Social Identity Theory and the Organization. Academy of Management Review, 14(1), 20-39. Baron, R., & Kenny, D. (1986). The Moderator-Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations. Journal of Personality and Social Psychology, 51(6), 1173-1182. Chiu, W., Lee, K., & Won, D. (2014). Consumer behavior toward counterfeit sporting goods. Social Behavior and Personality: an International Journal, 42(4), 615-624. Esch, F., Langner, T., Schmitt, B., & Geus, P. (2006). Are brands forever? How brand knowledge and relationships affect current and future purchases. Journal of Product & Brand Management, 15(2), 98-105. Field, A. (2005). Discovering statistics using SPSS (Second Edition ed.). London: SAGE Publications Frank, P., & Watchravesringkan, K. (2016). Exploring antecedents and consequences of young consumers' perceived global brand equity. Journal of Product & Brand Management, 25(2), 160-170. Frazier, P., Tix, A., & Barron, K. (2004). Testing Moderator and Mediator Effects in Counseling Psychology Research. Journal of Counseling Psychology, 51(1), 115-134. Gerke, A., Chanavat, N., & Benson-Rea, M. (2014). How can Country-of-Origin image be leveraged to create global sporting goods brands? Sport Management Review, 17(2), 174-189. Kapferer, J. (2004). The new strategic brand management: Creating and sustaining brand equity long term (3rd ed.). London: Kogan Page. Keller, K. (1993). Conceptualizing, Measuring, and Managing Customer-Based Brand Equity. Journal of Marketing, 57(1), 1-22. Keller, K. (2008). Strategic brand management: Building, measuring, and managing brand equity (3rd ed.). Upper Saddle River, NJ: Pearson Education, Inc. Kim, C., Han, D., & Park, S. (2001). The effect of brand personality and brand identification on brand loyalty: Applying the theory of social identification. Japanese Psychological Research, 43(4), 195-206. Kuenzel, S., & Halliday, S. (2010). The chain of effects from reputation and brand personality congruence to brand loyalty: The role of brand identification. Journal of Targeting, Measurement and Analysis for Marketing, 18(3-4), 167-176.

106 Macdonald, E., & Sharp, B. (2003). Management perceptions of the importance of brand awareness as an indication of advertising effectiveness. Marketing Bulletin, 14(2), 1-11. Martínez, P. (2015). Customer loyalty: exploring its antecedents from a green marketing perspective. International Journal of Contemporary Hospitality Management, 27(5), 896-917. Nunnally, J., & Bernstein, I. (1994). Psychometric theory (3rd ed.). New York: McGraw-Hill. Rageh Ismail, A., & Spinelli, G. (2012). Effects of brand love, personality and image on word of mouth: The case of fashion brands among young consumers. Journal of Fashion Marketing and Management: An International Journal, 16(4), 386-398. Riketta, M. (2005). Organizational Identification: A Meta-Analysis. Journal of Vocational Behavior, 66(2), 358- 384. Tajfel, H. (1974). Social Identity and Intergroup Behaviour. Social Science Information, 13(2), 65-93. Turner, J., Hogg, M., Oakes, P., Reicher, S., & Wetherell, M. (1987). Rediscovering the social group: a Selfcategorization theory. Oxford: Basil Blackwell. Weisheng, C., & Keat, L. (2016). Consumers’ intention to purchase counterfeit sporting goods in Singapore and Taiwan. Asia Pacific Journal of Marketing and Logistics, 28(1), 23-36. doi: doi:10.1108/APJML-02-2015- 0031

Catalyst ISSN 2408-137X, Volume 18, 2018 107 A Pilot Study on the Relationship between Postgraduate Students’ Perception of Thesis Supervision Role and Their Research Skills in a Thai University Saowalak Roongtawanreongsri and Risper A. Awuor Abstract This quantitative pilot study aimed to test the extent of the relationship between and effect of postgraduate students’ perception of thesis supervision role and their research skills. A questionnaire, constructed based on guidelines from the handbook for supervisors in Thai universities (OHEC, 2006) with a Cronbach’s alpha of 0.955, was used as an instrument to collect data using a case study approach on one graduate faculty at a university in southern Thailand with 30 participants. Principal factor analysis was carried out, as well as correlation and regression analysis. The results showed statistically significant positive relationships between essential student research skills and their critical thinking/synthesizing skills; their critical thinking/synthesizing skills and research ethics; their critical thinking/synthesizing skills and systematic thinking skills; and their critical thinking/synthesizing skills and supervisor relationships with students. However, supervisor character was not correlated with any student research skill variables, but highly correlated with the other remaining supervision roles. The strongest predictor of student research skills is the relationship between supervisor and student. Keywords: Supervision role, research skills, graduate student, postgraduate student Introduction Postgraduate education is important for human resource development (Ghadirian, Sayarifard, Majdzadeh, Rajabi, & Yunesian, 2014), and is a key element that contributes to the growth of national income (Sidhu, Kaur, Choo, & Fook, 2015). The demand for graduate study is growing, not only in developed countries, but also in developing countries. Several studies such as those conducted by Yellowley and Farmer (2005), Mistry, White, and Berardi (2009), and Celik (2011) have reported that postgraduate studies are one of the fastest growing markets in developed countries. Similar reports are found regarding developing countries. The increased number of graduate students is evidenced in many developing countries, for example, Poland (Rozanski, 2008), Thailand (Chien & Chapman, 2014) and Malaysia (Sidhu, Kaur, Choo, & Fook, 2015). Polziehn (2011) argued that graduate educational research skills (critical and creative thinking, personal effectiveness, integrity and ethical conduct, teaching competence, societal and civic responsibilities, and leadership) are attained through doing research, or in other words, the thesis. Research projects at the postgraduate level of education are aimed at enabling students to develop sound research skills, acquiring the latest theoretical knowledge, and becoming critical thinkers in their chosen fields of study, as well as conducting research enquiries in their areas of specialization in a professional and ethical manner (Office of Postgraduate Education, 2016). Because doing research or preparing a thesis helps develop research skills in graduate students, the need to look at how such skills can be developed may thus result in valuable improvements. Several studies have highlighted that the quality of the final thesis or research produced by postgraduate students depends very largely on thesis supervision; for example, Mainhard, van der Rijst, van Tartwijk, and Wubbels (2009); Garcia, Malott, and Brethower (1988); Ghadirian et al. (2014); de Kleijn, Mainhard, Meijer, Pilot, and Brekelmans (2012); Borders, Wester, Granello, Chang, Hays, Pepperel, and Spurgeon (2012). It thus implies that thesis supervision contributes to the development of research skills via preparation of a thesis. However, there has been minimal focus on how supervisors’ roles in thesis supervision affect the quality of research and students’ research skills. A lot of research have been done on interpersonal relationships between supervisors and advisees; for example, Mainhard et al. (2009), de Kleijn et al.

108 (2012), de Kleijn, Meijer, Pilot, and Brekelmans (2014), but not on how the supervision role develops research skills. Thai universities are under the purview of the Thai Office of the Higher Education Commission (OHEC). This organisation has published a handbook, which became the principal manual for all Thai universities, entitled “Thesis Supervisor: Duties and Roles” (2006) to provide guidelines for desirable supervisors’ roles. This handbook addresses significant issues of graduate study, with details on supervisory roles, duties, ethics, morality and instructions. It also includes the quality of desirable outputs expected of graduate students. Many supervisors, however, have not been trained to deliver supervisory service to students as recommended or guided by OHEC. Many students are left to find their way by themselves, creating frustration, and sometimes leading them to drop out of the program. There is no guarantee that those who remain will receive sufficient guidance and appropriate learning to ensure their ability to become competent researchers. It would therefore be worthwhile to investigate the extent to which the roles in the handbook actually relate to the results of students’ research skills. In this present paper, the researchers aimed to present the findings from a pilot study which focused on: 1) testing the reliability of the instrument constructed; 2) testing the relationship between student perceptions of the supervision role as suggested in the OHEC handbook, and student research skills; and 3) testing the effect of student perceptions of the supervision role and student research skills. The findings of this study will be useful in preparation for future studies of student perspectives of the role of supervisors on research skills development. Definition of Key Terms Supervision is defined by Islam, Amin, Ali, Alam, Alam, and Rasul (2013) as a two-way interactional process which requires both student and supervisor to consciously engage each other with a spirit of professionalism, respect, collegiality, and open-mindedness. In this study, however, it will mainly focus on the process delivered by thesis supervisors. The supervision role in this study is summarized from the concepts given by OHEC (2006), which state that the supervision role is the responsibility of the teacher who serves as a thesis advisor (supervisor) to individual students and the institution in administering, academic advising, counselling, and assuring successful outcomes for postgraduate students. Research skills in this study also followed OHEC (2006) definitions, which mean skills that postgraduate students should possess relating to the ability to do research well – for example, critical thinking skills, analytical skills, synthesizing skills, and self-reliance (ethics). Literature Review Supervision and Thesis Success Alam, Alam, and Rasul (2013) stated that the success and quality of postgraduate education largely depend on multiple factors, where supervisor roles determine postgraduate students’ overall satisfaction, retention, and completion. Vilkinas (2008) mentions other possible indicators, such as attendance status, level of research funding available, age, completion of an honours degree, discipline areas, and gender. Furthermore, other factors identified as significant predictors of completion are suitability of research topic, intellectual environment of the department, and access to equipment and computers. However, supervisor capability is a key factor in candidates’ success. Many elements influence supervisor performance and their relationships with postgraduate students. Supervisors, for example, must have research knowledge and related skills, possess management and interpersonal skills, be able to coordinate research program activities, mentor students, where appropriate and useful, and develop supportive relationships among research students themselves (Vilkinas, 2008). On the other hand, there are various reasons for students’ failure to complete their doctoral degrees, and supervisors may affect some of these. Sheikh Abdullah,

109 Osman, Mohamad, Meerah, Osman, Zakaria, and Mahmod (2012) indicate one of the main causes for students’ failure to complete their doctoral degrees is the lack of preparation in knowledge and skills to conduct research. Students are often expected to have, for example, skills in seeking and selecting information using the library, or from on-line/web sources, in their fields of interest. They should also be well equipped in research methodology, including data collection and using data analysis tools, and of course, writing up and disseminating the research findings. Keser, Ozcinar, Kanbul, and Meerah (2010) suggested that students’ readiness in research skills will affect completion time and the quality of the research or thesis they produce. They also suggested that students need to acquire these skills early. Thus, in order to ensure the successful completion of postgraduate study, the supervisor must play a role in aiding them to gain these skills as much as possible. Another area that is vital for thesis success is supervisory mentoring skills. Mentoring as a concept becomes prominent in advising students, being an exemplary model for them, supervising them, and transferring knowledge and experience to them (Arabaci & Ersözlü, 2010). According to Clutterbuck and Megginson (1999), the mentoring skills of a supervising professor are extremely important in educating postgraduate students. Some prominent qualities of a mentor are having gained experience out of good organization, good questions, reliability, being a good listener, patience, building networks, helping mentees be themselves, mutual understanding, keeping a balance between methods and contents, helpfulness, facilitating supportiveness, helping people use information, and trustworthiness. In this study, we call these qualities as the supervisors’ ability to understand students and their works (psychological qualities). Roles of Thesis Supervisors Schulze (2011) reviewed six supervisory styles and roles. These are the functional model (gives practical advice); qualities model (identifies flexibility and high emotional intelligence of supervisors as important); mentoring process (which requires supervisors to support and challenge students); enculturalisation model (enables students to become members of a research community); criticalthinking model (develops critical thinking in students by questioning them); and feminist model (promotes same-gender supervision). A mismatch in styles may lead to poor success rates. The author also reported supervisors’ roles and responsibilities in distance education, which can also be applied to face-to-face supervision. These include maintaining regular contact with students; returning written work in a timely manner; involving students in seminars; being knowledgeable about the literature available and the key debates in the student’s research field; giving encouragement and emotional support; and being role models. These may be summed up through the supervisor’s role as hypothesized by Kayrooz and Pearson (2004), found in Bitzer and Albertyn (2011), which includes mentoring, sponsoring, progressing the candidature, and coaching. Other related roles are presented in Table 1 below. Table 1. Supervisor Roles across Different Studies Study Mizany, Khabiri, & Sajadi, 2012 Nasir & Masek, 2015 Mhunpiew,2013 Orellana, Darder, Perez, Salinas, 2016 Supervisor Role Choose a suitable research topic; encourage students in thesis writing; establish good relationships with students; have sufficient research capabilities Guide in topic selection; monitor progress of thesis writing and research; managing research timeline; provide emotional support Support in terms of technical, intellectual, administrative, management, and personal Competent researcher who communicates well, offers appropriate support and suitable skills to enhance the students’ research experience

110 The OHEC Handbook, on the other hand, reviews and describes various roles of thesis supervisors. Most of the roles from the literature that the handbook reviewed are similar. The summarized roles include the supervisor being an up-to-date academic researcher, possessing the spirit of a teacher, being a role model for students, capable of critical thinking and guiding students’ work, being ready to help students psychologically and academically, being knowledgeable in both contents and administrative issues, availability, capable in managing research projects, being an evaluator, and being a colleague. Furthermore, supervisors take responsibilities in guiding students’ research, involving both academic and social aspects of advising, getting them involved in the wider research community, finding financial support and finding a position after graduation (Thai Office of the Higher Education Commission, 2006). Supervision and Student’s Research Skills While both the supervisor and supervisee assume various roles, a crucial required skill is research ability. Keser et al. (2010) suggested that research skills are important to all university graduates, regardless of whether they aim to work as researchers, in the public or the private sector. Research skills can be broadly categorised into two types: (i) skills needed to read and evaluate surveys, experiments, and other studies dealing with substantive problems in the research area; (ii) skills needed to do research including designing, conducting, analysing the data, interpreting and generalising a study’s results. Both categories of skills also rely on statistical and methodological competencies. These skills can only be learnt by someone who is actively engaged in the practice of doing research. Aside from these, there are also other meta-skills necessary for a productive research experience. In a PhD study, Blenkinsop (2003) stated that these include planning skills, communication skills, ethical awareness, self-confidence, self-awareness and self-protection, analytical skills, reflection, and self-criticism. The author also expands that learning, which takes place by undertaking a research project, is the kind of student-focused “deep learning”, which is an example of good educational practice. OHEC Handbook (2006): Thesis Supervisor: Duties and Roles To contextualize these attributes, we will consider the expectations listed by OHEC. In its handbook, guidelines for university supervisors were discussed through six chapters: 1) concepts and purpose of postgraduate students; 2) qualifications of thesis supervisors; 3) duties and roles of thesis supervisors in research steps; 4) obstacles in thesis supervising; 5) good practices for supervising; and 6) regulations and ethics for thesis supervisors. Chapter 1 of the handbook focuses on the concepts and purposes of postgraduate students, and emphasises the significance of postgraduate education as a conduit that enhances a country’s capacities in competition and development. It states clearly that research is the most significant part of postgraduate studies for students to practice how to conduct research, and how to contribute to new knowledge. This chapter also compiles the desired qualifications of postgraduate students, which can be summarised as having professional academic thinking skills, research skills, creativity, leadership, communication skills, and research ethics capacities. Chapter 2 discusses the qualifications of thesis supervisors, which should be comprised of academic competence and enthusiasm, being a role model, advisor, and counsellor. Chapter 3 describes the duties of thesis supervisors in the first part, and roles in the latter. For duties, the chapter discusses various types of supervisor behaviours. For the roles, it compiles and cites different sources for thesis supervisor roles. However, these roles are often repetitive. The last part of this chapter helps readers with the steps of thesis guiding, and what thesis supervisors should do in each step of the research: pre-proposal, proposal defense, post-proposal, and thesis defense. Chapter 4 outlines obstacles often encountered during thesis guiding, which include the ambiguous standard and quality of theses; lack of good thesis guiding mechanisms; problems of student abilities; and problems concerning supervisors and the supervising system.

111 Chapter 5 discusses the factors lead to success theses including structural factors, support factors, and exogenous variables. It also provides different strategies of thesis supervision. Chapter 6 includes the laws and regulations relating to preparing theses, which mostly cover general issues. However, several past cases of legal issues between students and institutions were given as examples. The last part of this chapter discusses research and supervision ethics. Research Methodology Participants For this pilot study, thirty graduate students from an interdisciplinary faculty in a southern Thai university were asked to respond. The participants were selected according to their willingness to participate. However, in a more complete research study, graduate students from all faculties in this same university will be included. Instrumentation A purposely-designed questionnaire was used as a tool for collecting quantitative data in this study. This was because the specific attributes of supervisor roles and student research skills outlined in the Thai university’s handbook had not been previously used in any measurement instrument. The questionnaire was divided into 5 sections: demographic data, perception toward students’ own research skills, perception toward supervisor roles, supervision type, and suggestions from students for better supervision. All questions about student perception of research skills and supervisors’ roles were extracted from the supervisors’ handbook published by the Office of the Higher Education Commission (2006). For student research skills, 30 items were extracted mainly from Chapter 1 of the handbook. A scale from one to ten was chosen for self-assessment (1 = ‘the least’ and 10 = ‘the most’). This scale was designed to allow for a wide range of scores to aid decision-making on the part of respondents as they evaluated and assessed their own skills. A sample of research skills questions is shown in Table 2. Table 2. Sample Items for Student Research Skills For research supervisory roles, items were compiled from any roles recommended in the handbook that were relevant to helping research students (mainly from Chapters 2 and 3). Redundant statements were either grouped or eliminated to keep the questionnaire within an appropriate length. All together there were 35 items on the supervisory role section. The scale for this attitudinal perception was from one to four (1 = ‘strongly disagree’ and 4 = ‘strongly agree’). No neutral scale was provided in the questionnaire in order to aid respondents in making their decisions on their supervisor’s role, which the researchers anticipated might be hard for students to decide, for in Thai culture, criticizing their teachers is not accepted nor is it a norm. A sample of items on supervision roles is shown in Table 3. Item No. Statement S7 You are knowledgeable both in theory and practice. S8 You have ability to think and analyse systematically. S9 You have good ethics and can rule yourself proudly in the wider society. S10 You have knowledge, ability, skills, attitudes, and ethics in your field of study. S13 You can think critically, can evaluate and use facts appropriately and logically. S19 You can help others or society by giving advice or suggestions and solutions to solve problems in an academic manner. S25 You have search the literature skilfully, and structure your thoughts from a review of the literature.

112 Table 3. Sample Items on Supervisor's Roles The questionnaire was reviewed for its suitability to use as a research tool by an expert panel. Data Collection The names and email addresses of all graduate students were gathered from the Registrar’s Office, and thirty samples were then selected. All questionnaires were sent via email with a letter of introduction to the selected samples. In the introduction letter, emphasis was put on describing that after the surveys were completed, the participants would return them via email to a third party (a faculty employee who was not a lecturer) so that student names and opinions would not be disclosed to lecturers. This was to ensure maximum truthfulness of their answers, as well as anonymity for the researcher. Data Analysis The data from returned questionnaires was processed using Predictive Analytic Software (PASW). The descriptive statistics, factor analysis, reliability analysis, correlation analysis, and multiple regression were run. Results Demographic Characteristics of Sample The demographic characteristics of 30 graduate students in the studied Faculty who were sampled in this study are shown in Table 4. The study investigated their perceptions toward their supervisor’s role in thesis guidance, using a questionnaire as a research tool. The results of the survey on perception of graduate students of the studied Faculty using a questionnaire to draw data from 30 participants (Table 1) show that 22 (73.3%) of the participants were females and eight (26.7%) were males. Almost all participants (29, or 99%) were younger than 40 years old (M = 31.5, SD = 5.63). Eighteen students (60%) reported that they spent 5-8 hours on their thesis daily, while only four (13.30%) worked longer hours on their thesis (M = 6.01, SD = 2.58). Most of them (27, or 90%), were studying for a Master’s degree, whereas three (10%) were working on their PhDs. The number of participants who were studying in the Social Sciences (13, or 43.30%) was almost equal to those studying in the Sciences (12, or 40%), while five (16.70%) used both approaches to integrate their study. A majority of participants (23, or 76.7%) graduated within four years of study (M = 3.66, SD = 1.08). A large number of participants (25, or 83.3%) never had to change supervisors during their course of study, whereas five (16.7%) needed to. About one third of the participants (11, (36.7%) received supervision more than six times in a month (M = 4.82, SD = 2.43). Twenty-three (76.7%) of participants’ supervisors were females, and seven (23.30%) were males. A little over half, (16, 53.3%) of the supervisor’s age was between 41-50 years old (M = 45.5, SD = 6.95). Item No. Statement T1 You had a good student-teacher relationship with your supervisor. T2 Your supervisor set a regular time for you to meet. T3 Your supervisor was academically competent and enthusiastic in academic work. T5 Your supervisor had sufficient knowledge and understanding in the problems you investigated. T9 Your supervisor was a good role model who showed passion toward research. T11 Your supervisor understood psychological principles and applied them to learn about their students, in both their academic and personal habits. T14 Your supervisor was interested in your research problem and paid attention. T17 Your supervisor respected your opinion.

113 Table 4. Demographic Data of Research Participants Factor Analysis The main applications of factor analytic techniques are: (1) to reduce the number of variables; and (2) to detect structure in the relationships between variables – that is, to classify variables. Therefore, factor analysis is applied as a data reduction or structure detection method (StatSoft, 2017). Since there were 30 items on students’ perception, and 35 items for supervision roles, which were quite a number of variables, factor analysis was necessary to group, reduce, and classify the variables under the similar themes. For the 30 items on perceptions in Section 2 of the questionnaire, two items investigated students’ perception of success, and another three items about supervisor’s role in their research success. These items were grouped into students’ perception of success (items 2 and 3), and perception of supervision role for their success (items 4, 5, and 6), and were not included Variables Frequencies Percentage Mean Standard Deviation N=30 Gender - - Male 8 26.7 Female 22 73.3 Age Range 31.5 5.63 20-30 13 43.3 31-40 16 53.3 41-50 1 3.3 Hours Working on Thesis 6.01 2.58 1-2 hours/day 2 6.7 3-4 hours/day 6 20.0 5-6 hours/day 9 30.0 7-8 hours/day 9 30.0 >8 hours/day 4 13.3 Degree Studied - - Master 27 90.0 PhD 3 10.0 Field of Study - - Science 12 40.0 Social Science 13 43.3 Integrated of both 5 16.7 Total Years of Study 3.66 1.08 < 2 years 2 6.7 2-3 years 11 36.7 3-4 years 10 33.3 > 4 years 7 23.3 Change of Supervisor - - Yes 5 16.7 No 25 83.3 Supervision Frequencies 4.82 2.43 1-2 times/month 7 23.3 3-4 times/month 8 26.7 5-6 times/month 4 13.3 > 6 times/month 11 36.7 Supervisor Gender - - Male 7 23.3 Female 23 76.7 Supervisor Age 45.5 6.95 31-40 7 23.3 41-50 16 53.3 51-60 7 23.3

114 in factor analysis. Item number 1 asked about students’ role in writing the thesis, so it was also not included in factor analysis. A principal factor analysis (PFA) of the remaining 24 items was then conducted, using varimax rotation. The analysis extracted four factors that successfully explained 73.54% of the variance. The labels for each factor were given according to the theme of the items, which were: Students’ Essential Research Skills (10 items), Student Critical Thinking and Synthesizing Skills (7 items), Student Research Ethics (5 items), and Student Systematic Thinking Skill (2 items). Samples of each items of these themes are given in Table 5. Table 5. Sample of Students’ Research Skills Items under Each Factor Analysis Theme A PFA was also conducted for 35 items on supervision roles in Section 3 of the questionnaire. Eight factors were extracted which explained 84.5% of the variance. However, due to limited space of this article, only four variables were used as independent variables in multiple regression analysis, which included Supervisor Characteristics (7 items), Supervisor’s Dedication to Thesis Supervision (6 items), Relationships with Students (7 items), and Supervisor’s Understanding of the Thesis (6 items). Table 6. Sample of Supervisor’s Roles Items under Each Theme of Factor Analysis Reliability Analysis Internal consistency or reliability for each factor was examined using Cronbach’s alpha, which resulted in a Cronbach’s alpha of 0.955. The alphas for factors shown in Table 7 were as follows: 0.930 for Students’ Essential Research Skills (10 items); 0.913 for Student Critical Thinking and Synthesizing Skills (7 items); 0.846 for Student Research Ethics (5 items); 0.856 for Student Systematic Thinking Skills (2 items); 0.949 for Supervisor Characteristics (7 items); 0.954 for Supervisor’s Dedication to Thesis Supervision (6 items); 0.933 for Relationships with Students (7 items); and 0.894 for Supervisor’s Understanding of the Thesis (6 items). Skills Items Students’ Essential Research Skills S17 - You can creatively produce academic work and can conduct your own research correctly according to the right methodology. S19 - You can help others or society by giving advice or suggestions and solutions to solve problems in an academic manner. Student Critical Thinking and Synthesizing Skills S13 - You can think critically, evaluate, use facts appropriately and logically. S25 - You can search the literature skilfully, and structure your thoughts from a review of the literature. Student Research Ethics S9 - You have good ethics and can rule yourself proudly in the wider society. S10 - You have knowledge, ability, skills, attitudes, ethics in field of study. Student Systematic Thinking Skills S7 - You are knowledgeable in both theory and practice. S8 - You have ability to think and analyse systematically. Roles Items Supervisor Characteristics T3 - Your supervisor was academically competent and enthusiastic in academic work. T9 - Your supervisor was a good role model who showed passion toward research. Supervisor’s Dedication to Thesis Supervision T2 - Your supervisor set a regular time for you to meet. T14 - Your supervisor was interested in your research problem and paid attention. Relationships with Students T1 - You had a good student-teacher relationship with your supervisor. T17 - Your supervisor respected your opinion. Supervisor’s Understanding of the Thesis T5 - Your supervisor had sufficient knowledge and understanding in the problems you investigated. T11 - Your supervisor understood psychological principles and applied them to learn about their students, both in academic and personal habits.

115 Table 7. Descriptive Statistics for Four Student Skill Factors & Four Supervision Role Factors (N = 30) Factors No. of Items Mean (S.D.) Skewness Kurtosis Cronbach’s Alpha Student Skill Factors Essential Research Skills 10 7.32 (1.04) -0.478 0.864 0.930 Synthesizing Skills 7 7.79 (0.92) -0.294 -0.686 0.913 Research Ethics 5 8.43 (0.79) -0.214 -0.361 0.846 Systematic Thinking Skills 2 7.37 (0.80) -0.621 -0.229 0.856 Supervision Role Factors Supervisor Characteristics 7 3.67 (0.48) -1.400 0.738 0.949 Supervisor’s Dedication to Thesis Supervision 6 3.59 (0.52) -1.098 0.180 0.954 Relationships with Students 7 3.61 (0.43) -1.080 -0.182 0.933 Supervisor’s Understanding of the Thesis 6 3.12 (0.35) -0.223 -1.544 0.894 Correlation Analysis Correlations were computed among the four supervisor’s roles, and student perceptions of their research skills as shown in Table 8. The results suggest that 16 out of 56 correlations were statistically significant, and 14 of them were highly significant at the level of p < 0.01, two tailed. The remaining two correlations were statistically significant at the level of p < 0.05, two tailed. The correlation analysis indicates that there are strong positive relationships with high statistical significance between Students’ Essential Skills and Students’ Critical Thinking and Synthesizing Skills (r = 0.69, n = 30, p < 0.001), and between Students’ Essential Skills and Students’ Research Ethics (r = 0.57, n = 30, p < 0.001). Student Critical Thinking and Synthesizing Skills and Student Research Ethics were positively correlated with a high level of statistical significance (r = 0.69, n = 30, p < 0.001), as well as Student Systematic Thinking Skills (r = 0.51, n = 30, p < 0.01). These skills were also moderately correlated to Supervisor’s Relationships with Students (r = 0.37, n = 30, p < 0.05) and Supervisor’s Understanding of the Thesis (r = 0.39, n = 30, p < 0.05). We also found that Student Research Ethics were positively correlated with a high level of statisticalsignificance to Students’ Systematic Skills(r = 0.48, n = 30, p < 0.01), Supervisor’s Dedication to Thesis Supervision (r = 0.48, n = 30, p < 0.001), Relationship with Students (r = 0.57, n = 30, p = 0.001), and Supervisor’s Understanding of the Thesis (r = 0.55, n = 30, p < 0.005). This variable is the most interesting, as it is correlated with all remaining variables except for Supervisor Characteristics. Supervisor Characteristics was not correlated with any student research skill variables, but was highly correlated with the other remaining supervision roles: Supervisor’s Dedication to Thesis Supervision (r = 0.75, n = 30, p < 0.001), Relationships with Students (r = 0.76, n = 30, p < 0.001), and Supervisor’s Understanding of the Thesis (r = 0.62, n = 30, p < 0.001).

11 Table 8. Correlation Analysis Students’ Essential Research Skills Students’ Critical Thinking and Synthesizing Skills Re Students’ Essential Research Skills Pearson Correlation 1 0.693** Sig. (2-tailed) 0.000 Student Critical Thinking and Synthesizing Skills Pearson Correlation 0.693** 1 Sig. (2-tailed) 0.000 Student Research Ethics Pearson Correlation 0.571** 0.690** Sig. (2-tailed) 0.001 0.000 Student Systematic Thinking Skills Pearson Correlation 0.299 0.510** Sig. (2-tailed) 0.109 0.004 Supervisor’s Characteristics Pearson Correlation -0.034 0.084 Sig. (2-tailed) 0.860 0.661 Supervisor’s Dedication to Thesis Supervision Pearson Correlation 0.091 0.290 Sig. (2-tailed) 0.634 0.119 Relationships with Students Pearson Correlation 0.282 0.370* Sig. (2-tailed) 0.131 0.044 Supervisor’s Understanding of the Thesis Pearson Correlation 0.317 0.399* Sig. (2-tailed) 0.088 0.029 * Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (2-tailed).

16 Students’ esearch Ethics Students’ Systematic Thinking Skills Supervisor’s Characteristics Supervisor’s Dedication to Thesis Supervision Relationships with Students Supervisor’s Understanding of the Thesis 0.571** 0.299 -0.034 0.091 0.282 0.317 0.001 0.109 0.860 0.634 0.131 0.088 0.690** 0.510** 0.084 0.290 0.370* 0.399* 0.000 0.004 0.661 0.119 0.044 0.029 1 0.481** 0.310 0.484** 0.567** 0.552** 0.007 0.095 0.007 0.001 0.002 0.481** 1 -0.001 0.182 0.231 0.104 0.007 0.996 0.335 0.219 0.583 0.310 -0.001 1 0.752** 0.764** 0.620** 0.095 0.996 0.000 0.000 0.000 0.484** 0.182 0.752** 1 0.795** 0.732** 0.007 0.335 0.000 0.000 0.000 0.567** 0.231 0.764** 0.795** 1 0.674** 0.001 0.219 0.000 0.000 0.000 0.552** 0.104 0.620** 0.732** 0.674** 1 0.002 0.583 0.000 0.000 0.000

117 Multiple Regression Analysis The results of multiple regression analysis are shown in Table 9. Table 9. Multiple Regression with Reverse Scale The multiple regression model used in this study was comprised of the following independent variables: Supervisor Characteristics, Supervisor’s Dedication to Thesis Supervision, Relationships with Students, and Supervisor’s Understanding of the Thesis. The multiple regression model with all four predictors produced R² = 0.404, F(4, 25) = 4.23, p < 0.005. Supervisor Characteristics had a negative statistical significance effect on student research skills (β = -0.635, t = 2.459, p = 0.021). Supervisor’s Dedication to Thesis Supervision had a positive effect on student research skills (β = -0.177, t = -0.595, p = 0.557), but with no statistical significance. Relationships between supervisors and students was the strongest predictor of student research skills, with a statistical significance (β = 0.728, t = 2.560, p = 0.017). Supervisor’s Understanding of the Thesis had no significant effect on student research skills (β = 0.461, t = 1.978, p = 0.059). Discussion The results from this pilot study showed the potential for conducting a larger study on this topic with some improvements. For instrument testing, factor analysis grouped similar items under common themes, totalling four themes for student research skills and eight themes for supervisor roles, with four considered in this paper. For student research skills, the items under the same theme were related, and names for the new variable could be easily given: Students’ Essential Research Skills, Student Research Ethics, Student Critical Thinking and Synthesizing Skills, and Student Systematic Thinking Skills. On the contrary, for the supervisor’s role, only Supervisor Characteristics seemed to group all similar items together. For the remaining three themes, which were Supervisor’s Dedication to Thesis Supervision, Relationships with Students, and Understanding of the Thesis, some items that were not similar were grouped within the same theme. For example, three items were placed under Relationships with Students – item No. 4 “Your supervisor had ability to evaluate the quality of research in the field”, item No. 18 “Your supervisor possessed the ability to think critically and analytically”, and item No. 27 “Your supervisor had potential to evaluate the thesis and the student’s ability to prepare the thesis.” However, it seems that not all of these belong under this theme. Likewise, there were two items each under Supervisor’s Dedication to Thesis Supervision and Understanding of the Thesis variables that were not so related to the rest of the items. This problem was probably caused by the complexity and ambiguity of the statements, which needs to be improved to make the statements shorter and clearer if the questionnaire is to be used in the future. Despite the issue mentioned above, however, the items when grouped under the reported themes attained a satisfactory level of reliability, meaning that the questionnaire could be used as an instrument in a larger study. The findings of this pilot study show the correlations between perception of supervision role and postgraduate research skills. In summary, the results showed positive statistically significant R R 2 F Standardized Coefficients Beta t P 0.636 0.404 4.23 Supervisor Characteristics -0.635 -2.459 0.021 Supervisor’s Dedication to Thesis Supervision -0.177 -0.595 0.557 Relationships with Students 0.728 2.560 0.017 Supervisor’s Understanding of the Thesis 0.461 1.978 0.059

118 relationships between Student Essential Research Skills and Student Critical Thinking and Synthesizing Skills; Student Critical Thinking and Synthesizing Skills and Student Research Ethics; Student Critical Thinking and Synthesizing Skills, and Student Systematic Thinking Skills; and Student Critical Thinking and Synthesizing Skills and Supervisor Relationships with Students. These results supported the work of de Kleijn et al. (2012) who also reported a relationship between perceptions of the supervisor and student learning. Thus, these findings support a further and more detailed study with larger samples. Some correlations show interesting relationships that should be investigated in the future. For example, Student Critical Thinking and Synthesizing Skills and Student Research Ethics were correlated with all of the remaining skills, implying that students who possess these two skills may also possess other skills as well. In other words, it can be argued that these skills are vital for postgraduate students as an advanced researcher. Thus, supervisors should nurture these skills in young researchers, which is in part consistent with Egan et al. (2009), who argued that effective supervision is a crucial factor to help students succeed in their thesis development. Furthermore, only one variable, Student Research Ethics, was associated with most supervisor roles (Supervisor’s Dedication to Thesis Supervision, Relationships with Students, and Supervisor’s Understanding of the Thesis). Interestingly, this one variable also shows association with all Student Skills. This can generate an assumption that supervisor roles can probably create student research ethics, which are strongly associated with all the remaining student skills. Several studies have been carried out to examine different models of supervision such as Nordentoft et al. (2013) and Agu and Odimegwu (2014). However, not many studies have been carried out to focus on creating particular skills. This finding points to a research gap that need to be further explored. Relationships with Students was shown to have a statistically significant effect on student research skills. Thus, the study confirms what has been previously found regarding this issue by other researchers such as Mainhard et al. (2009) and de Kleijn et al. (2012). It would be worthwhile to study these factors in detail in future research. However, the Supervisor Characteristics variable had a negative statistically significant effect on student research skills, which is contrary to most of the literature. No sufficient logical explanation is offered for this phenomenon through this pilot study. A more comprehensive study may need to be carried out to verify this finding. It is also important to point out that despite several studies showing that supervisor roles are vital for thesis completion (for example, Keser et al., 2010; Sheikh Abdullah et al., 2012), with Keser et al. (2010) also mentioning the importance of building research skills for graduate students, a study that associates the two factors is not yet available. Most of these research projects studied the two factors separately; therefore, at this juncture, it is not possible to discuss the findings based on previous works. The authors hope that this paper will bring more light to this area. Conclusions and Recommendations This pilot study was conducted in preparation for a deeper study to examine the relationship between students’perception of supervision role and research skills, following the OHEC guidelines for supervision. A questionnaire was constructed and used to collect demographic data and perceptions from 30 graduate students in a Faculty at a southern Thai university. A ten-point ranking scale was used to evaluate students’ perception of supervisors’ roles and their own research skills. Student research skills were assessed through 24 items in the questionnaire. Factor analysis grouped the research skills into four components; all of them had a high Cronbach’s reliability score ranging between 0.846 and 0.930, whereas the supervision role was grouped into eight components. Only four of these were included in this study, and all of them also had a high Cronbach’s reliability score of between 0.894 – 0.954. The correlation analysis showed many variables that had statistically significant relationships with student research skills, between supervision roles, and between student research skills and supervision roles. However, the regression analysis predicted that their effect on research skills depended significantly on relationships between supervisors and students, which agrees with earlier research studies. Even though not much can be summarised from this pilot study, it has shown

119 interesting variables that should be given attention. These variables are supervisor characteristics (which are associated with all other supervisory roles), supervisor’s dedication to the thesis, supervisor’s comprehensive understanding of the student’s work, and supervisor’s good relationship with student. Therefore, supervisors should consider investing in developing these skills to provide better help to students. In other words, postgraduate research supervision should also receive formal supervision training (Orellana et al., 2016) in order to ensure better quality and completion of research theses. Since there is minimal research work done on the association between supervisor’s role in thesis supervision and student research skills, a larger study should focus on these factors in particular, in order to determine with certainty the association and effect between these factors. About the Authors Dr. Saowalak Roongtawanreongsri is an Associate Professor in the Environment Economic Research Unit, Faculty of Environmental Management, Prince of Songkla University, Hat Yai, Songkhla, Thailand. Her research interests include environmental economics, social science of environmental management, learning and education, environmental education, and community participation. Email: [email protected] Professor Dr. Risper Awuor is Deputy Vice Chancellor at Adventist University of Africa, Kenya. She has published articles on educational measurements, multivariate approaches to quantitative research studies, and factors affecting academic performance. References Agu, N., & Odimegwu, C. (2014). Doctoral dissertation supervision: Identification and evaluation of models. Education Research International, 2014, 1-9. Alam, F., Alam, Q., & Rasul, M. (2013). A pilot study on postgraduate supervision. Procedia Engineering, 56, 875-881. doi: http://dx.doi.org/10.1016/j.proeng.2013.03.210 Arabaci, I., & Ersözlü, A. (2010). Postgraduate students’perceptions of their supervisors’ mentoring skills (Gaziosmanpa University example). Procedia Social and Behavioral Sciences 2 (2010), 4234-4238. Bitzer, E., & Albertyn, R. (2011). Alternative approaches to postgraduate supervision: A Planning Tool to Facilitate Supervisory Processes. South African Journal of Higher Education, 25, 875-888 Blenkinsop, C. (2003). Research: an essential skill of a graduate nurse? Nurse Education Today, 23, 83-88. Borders, L., Wester, K., Granello, D., Chang, C., Hays, D., Pepperel, J., & Spurgeon, S. (2012). Association for counselor education and supervision guidelines for research mentorship: Development and implementation. Counselor Education & Supervision, 51, 162-175. Celik, S. (2011). What are the factors comprising postgraduate education demands, and what are levels of agreement with these factors? Education Research International, 132(1), 185-202. Chien, C., & Chapman, D. (2014). Expanding graduate education in Malaysia and Thailand. Retrieved from http://www.universityworldnews.com/article.php?story=20140610125523334 Clutterbuck, D., & Megginson, D. (1999). The lessons for executive and director mentoring. In Mentoring executives and directors (pp. 131-163). Oxford: Butterworth-Heinemann. de Kleijn, R., Mainhard, T., Meijer, P., Pilot, A., & Brekelmans, M. (2012). Master’s thesis supervision: Relations between perceptions of the supervisor–student relationship, final grade, perceived supervisor contribution to learning and student satisfaction. Studies in Higher Education, 37(8), 925-939. de Kleijn, R., Meijer, P., Pilot, A., & Brekelmans, M. (2014). The relation between feedback perceptions and the supervisor–student relationship in master’s thesis projects. Teaching in Higher Education, 19(4), 336-349. Doğan, N., & Bıkmaz, Ö. (2015). Expectation of Students from their Thesis Supervisor. Procedia - Social and Behavioral Sciences, 174 (Supplement C), 3730-3737. Retrieved from https://doi.org/10.1016/ j.sbspro.2015.01.1106 Egan, R., Stockley, D., Brouwer, B., Tripp, D., & Stechyson, N. (2009). Relationships between area of academic concentration, supervisory style, student needs and best practices. Studies in Higher Education, 34(3), 337-345. doi: 10.1080/03075070802597143 Garcia, M., Malott, R., & Brethower, D. (1988). A system of thesis and dissertation supervision: Helping graduate students succeed. Teaching of Psychology, 15(4), 186-191.

120 Ghadirian, L., Sayarifard, A., Majdzadeh, R., Rajabi, F., & Yunesian, M. (2014). Challenges for better thesis supervision. Medical Journal of the Islamic Republic of Iran, 28, 1-9. Keser, H., Ozcinar, Z., Kanbul, S., & Meerah, T. (2010). World Conference on Learning, Teaching and Administration PapersReadiness of preparing postgraduate students in pursuit of their doctoral programme. Procedia - Social and Behavioral Sciences, 9, 184-188. doi: http://dx.doi.org/10.1016/ j.sbspro.2010.12.133 Mainhard, T., van der Rijst, R., van Tartwijk, J., & Wubbels, T. (2009). A model for the supervisor–doctoral student relationship. Higher Education, 58, 359-373. doi: 10.1007/s10734-009-9199-8 Meerah, T., Osman, K., Zakaria, E., Ikhsan, Z., Krish, P., Lian, D., & Mahmod, D. (2012). Developing an instrument to measure research skills. Procedia - Social and Behavioral Sciences, 60, 630-636. Retrieved from https://www.sciencedirect.com/science/article/pii/S187704281203889X Mhunpiew, N. (2013). A Supervisor's Roles for Successful Thesis and Dissertation. Online Submission. Mistry, J., White, F., & Berardi, A. (2009). Skills at Master’s Level in Geography Higher Education: Teaching, Learning and Applying. Journal of Geography in Higher Education, 33(1), 123-148. Mizany, M., Khabiri, M., & Sajadi, S. (2012). A study of the capabilities of graduate students in writing thesis and the advising quality of faculty members to pursue the thesis. Procedia - Social and Behavioral Sciences, 31, 5-9. doi: http://dx.doi.org/10.1016/j.sbspro.2011.12.006 Nasir, S., & Masek, A. (2015). A model of supervision in communicating expectation using supervisory styles and students learning styles. Procedia - Social and Behavioral Sciences, 204, 265-271. Retrieved from https://www.sciencedirect.com/science/article/pii/S1877042815047989 Nordentoft, H., Thomsen, R., & Wichmann-Hansen, G. (2013). Collective academic supervision: A model for participation and learning in higher education. Higher Education, 65, 581-593. doi: 10.1007/s10734- 012-9564-x Office of the Higher Education Commission. (2006). Supervisor: Duties and roles. Bangkok: Thailand Office of the Higher Education Commission. Office of Postgraduate Education. (2016). Aims and Objectives of Research Postgraduate Education. Retrieved 10 April 2016 from http://pg.ust.hk/cgi-bin/pg/eng/secondary.php?catid=1&sid=1&tid=164 Orellana, M., Darder, A., Perez, A., & Salinas, J. (2016). Improving doctoral success by matching PhD students with supervisors. International Journal of Doctoral Studies, 11, 87-103. Polziehn, R. (2011). Skills expected from graduate students in search of employment in academic and nonacademic settings: Faculty of Graduate Studies and Research. Rozanski, A. (2008). The educational orientation of participants in postgraduate managerial study programmes in Poland. Human Resource Development International, 11(1), 91-99. Schulze, S. (2011). A survey of students' views of supervision at Unisa. South African Journal of Higher Education, 25(4), 784-802. Sidhu, G., Kaur, S., Choo, L., & Fook, C. (2015). Postgraduate students’ level of dependence on supervisors in coping with academic matters and using digital tools. Paper presented at the 12th International Conference on Cognition and Exploratory Learning in Digital Age. StatSoft, Inc. (2017). Principal components and factor analysis. https://store.fmi.uni-sofia.bg/fmi/statist/ education/textbook/eng/stfacan.html Vilkinas, T. (2008). An exploratory study of the supervision of Ph.D./research students’ theses. Innovative Higher Education, 32(5), 297–311. doi: 10.1007/s10755-007-9057-5 Yellowley, W., & Farmer, M. (2005). Team teaching in higher education reflections on the added value of team teaching on student and staff learning experiences. International Journal of Learning, 12(6), 85-89.

Catalyst ISSN 2408-137X, Volume 18, 2018 121 Variables Affecting the Servant Leadership of Educational Institute Administrators Under the Primary Educational Service Area Office Phitzesa Phongpanphaengpanga Abstract The research reported here involved primary school administrators (n = 42) under the Primary Education Service Area Offices for the provinces of Buriram, Chaiyaphum, Khon Kaen, Nakhon Ratchasima, and Surin in the lower northeastern region of Thailand during the 2016 academic year. It was found that: 1) The educational administrators displayed a “very high” level of servant leadership, which differed slightly among the administrators. 2) The variable that directly influenced administrators’ servant leadership was achievement motivation, which accounted for 65.30 percent of the variance (p < .01 level); and 3) The administrators’ servant leadership was indirectly influenced by their practices according to the four sublime states of mind: Loving-kindness, Compassion, Sympathetic Joy, and Equanimity and Learning Motivation, acting through their achievement motivation. Their work responsibilities directly influenced their practices, according to the four sublime states of mind and learning motivation. Keywords: Servant leadership, primary educational administrators Introduction Economic and social changes, technological advancement, and globalization mean that countries face higher levels of competition. In order to cope with these changes, there is an urgent need to recognize and develop human resources. In order to meet these needs, it is necessary to improve the quality of students’ educational experiences. This is true in many countries, including Thailand. Good quality education enables the country to become visible on the world stage in the era of globalization. However, the low quality of Thailand’s education has been a problem at all educational levels. This is reflected in low achievement scores in science, mathematics and English language subjects, as well as unsatisfactory levels and skills in the area of critical thinking (Martin, Gregory, & Stemler, 2000). Although knowledge and educational opportunities are improving, the quality of Thai education remains in need of substantial reform. Failure to address the shortcomings will affect all Thai people adversely. This critical situation is an important reason for seeking to improve the quality of educational experiences (Sinlarat, Teerapijit, & Chaodamrong, 2010). Administrators play a significant role in guiding institutions to implement educational changes successfully according to the Government’s educational reform guidelines. They should possess good competency, knowledge, ability, ethics, and morals. In order to manage and administer an educational institute, administrators need to possess good knowledge and understanding about the guidelines for educational reform and educational quality assurance. In addition, they need to be competent educational managers, possess knowledge and understanding in the art of developing learnercentered curricula that facilitate classroom research, and be able to make the school a true learning organization. Administrators also need to have leadership skills, especially servant and academic leadership skills (servant leadership means service for teachers, students, and the parents), to enhance educational administration and management (Roonjoraen, 2010). This is consistent with the concept provided by PhraBrahmagunabhorn (2010) and Boonruangrut (2013). They claimed that to succeed in administering Thai education, according to the guidelines for educational reform, administrators must be competent, knowledgeable, and capable of carrying out their work based on the practice of the four sublime states of mind. These consist of loving-kindness, compassion, sympathetic joy, and

122 equanimity as the model for good self-practice, as well as knowledge and understanding of educational reform guidelines and a Buddhism-based leadership method. Greenleaf (2002) has suggested that the success of an organization is based on the result of the servant leadership of its administrators. However, it is hypothesized here that four variables affect the characteristics displayed by leaders. These are administrators’ achievement motivation (AAM: the wish to do something successfully), administrators’ learning motivation (ALM: the need to improve knowledge for career development), administrators’ work responsibility (AWR: competency to work more efficiently), and administrators’ practice of the four sublimes state of mind (APM: lovingkindness, compassion, sympathetic joy, equanimity). Servant leadership is a philosophy and set of practices that enriches the lives of individuals, builds better organizations, and ultimately creates a more just and caring world. The administrators should not focus on competition, but make decisions on a group basis, putting more emphasis on the perceived values of their colleagues. They will be more interested in developing their colleagues’ potential as the important basis of practice. Literature Review A number of relevant papers on servant leadership in the Thai context were reviewed. Mekkayai (2007) sought to develop a servant leadership training curriculum for student leaders at Kasetsart University. This study demonstrated that individuals can be trained successfully to adopt a servant leadership approach to challenging administrative issues. Servant leadership is composed of 10 factors; conceptualization, stewardship, persuasion, foresight, listening, healing, commitment to the growth of people, building community, empathy and awareness (Chaiyapheuk, 2007). Lengpanich’s (2011) research found four factors, which consisted of credibility, creativity, vision, and integrity of primary school administrators. Servant leadership, team work, and organizational citizenship behavior together provide a reasonable explanation of the results of managers’ job performances (Ruenthai, 2008). Servant leadership and the spirituality shown in organizations has a positive effect on managers’ job performances (Meemheuy, 2010). Nithananon (2014) studied the relationship between servant leadership and job performance of teacher coordinators connected with the World Morality Revival Project, a project seeking to make the world a better place. The result was commitment to personnel development, followed by that of community establishment, and then listening. Phongpanphaengpanga (2014) considered servant leadership for educational institution administrators in the lower Northeastern region of Thailand. The research findings indicated that exercising moral authority, having a service personality, giving priority to serving others, being compassionate collaborators, using foresight, being systems thinkers, and being skilled communicators were essential factors. Learning the secrets of servant leadership is essential for administrators who manage and train educational personal. There is no “silver bullet” for being a good leader. But if there was, maybe it would be first serving the people around you. As servant leadership influences organizational management and educational management development, this researcher was interested in investigating factors which influence the levels of servant leadership of educational institute administrators who work under the Primary Educational Service Area Office. Research Objectives The objectives of this research were: 1) To investigate the level of servant leadership of educational institute administrators under the Primary Educational Service Area Office. 2) To investigate variables affecting the level of servant leadership of educational institute administrators under the Primary Educational Service Area Office.

123 Methodology The population chosen came from among the educational institutes’ administrators under the Primary Educational Service Area Offices. Sixty administrators were chosen from primary schools in Buriram, Chaiyaphum, Khon Kaen, Nakhon Ratchasima, and Surin provinces in northeastern Thailand. This was done on the proportional probability to size (PPS) sampling method according to the number of students in each school (Sudman, 1976). This gave a good distribution of samples from the target area. Returns came from 42 administrators, representing 70% of educational institutes’ administrators under the Primary Educational Service Area Offices in the 2016 academic year. In order to increase the validity and reliability of data, four people were selected to measure the variables for each school administrator. These four people consisted of a school administrator, one head of department, and two teachers selected by random numbers taken from a table. The data from the four people were averaged to give a mean score for each variable and for each school administrator. Research Variables The dependent variable was the administrators’ servant leadership (ASL). The independent variables were the administrators’ achievement motivation (AAM), practice of the four sublime states of mind (loving-kindness, compassion, sympathetic joy, equanimity – APM), learning motivation (ALM), work responsibility (AWR), and their self-development (ASD). Research Tools The tool used for measuring the dependent variable was the administrators’ servant leadership questionnaire, which consisted of 10 items with five rating levels (very low, low, moderate, high, and very high). The questionnaire had discrimination power of between 0.600 and 0.800, and a reliability score of 0.822. Tools for measuring independent variables consisted of the following: 1) An administrators’ self-development questionnaire, consisting of 17 items with a five-point rating scale, having discrimination power from 0.500-0.790 and reliability of 0.930; 2) An administrators’ learning motivation questionnaire, containing 24 items with a five-point rating scale, having discrimination power from 0.911-0.919 and reliability of 0.919; 3) An administrators’ achievement motivation questionnaire, containing 24 items with five-point rating scale, having discrimination power from 0.800- 1.000 and reliability of 0.920; 4) An administrators’ work responsibility questionnaire, containing 22 items with a five-point rating scale, having discrimination power from 0.200-0.659 and reliability of 0.939; and 5) A questionnaire on administrators’ practice of the four sublime states of mind (loving-kindness, compassion, sympathetic joy, equanimity) containing four response items, having discrimination power from 0.950- 0.960 and reliability of 0.950. Data Analysis The data were analyzed by means, standard deviations, coefficients of variation (CV) for each dependent and independent variable, simple correlation, beta weight of independent variable on the dependent variable, squared multiple correlation, and path analysis (PAL), using the appropriate software. Results Table 1 shows levels of the relationships found between administrators’ servant leadership, and five independent variables.

124 Table 1. Levels of the Administrators’ Servant Leadership (Dependent Variable) as Influenced by Five Independent Variables. Variables N X S.D. C.V. Administrators’ Servant Leadership (ASL) 42 4.410 0.317 7.188 Administrators’ Self-development (ASD) 42 3.895 0.388 9.961 Administrators’ Learning Motivation (ALM) 42 4.364 0.341 7.814 Administrators’ Achievement Motivation (AAM) 42 4.348 0.334 7.682 Administrators’ Work Responsibility (AWR) 42 4.419 0.300 6.789 Administrators’ Practice of the Four Sublime States of Mind (Lovingkindness, Compassion, Sympathetic Joy, Equanimity – APM) 42 4.537 0.368 8.111 Educational institute administrators reported high levels of servant leadership ( X = 4.410), with slight differences among individual respondents (C.V. = 7.188). Similar results were also obtained for their self-development( X = 3.895; C.V. = 9.961), learning motivation ( X = 4.364; C.V. = 7.814), and achievement motivation ( X = 4.348; C.V. = 7.682). The administrators rated their level of work responsibility as high ( X = 4.419; C.V. = 6.789), along with their practice of the Four sublime states of mind (loving-kindness, compassion, sympathetic joy, equanimity) ( X = 4.537; C.V. = 8.111). Figure 1 shows a model constructed after path analysis detailing the influences, in term of β weight, contributing to administrators’ servant leadership (ASL). 0.457 0.737** 0.264** 0.147 0.793** 0.857** 0.733** 0.266 Chi-Square = 8.54, df = 5, P-value = 0.12867, RMSEA = 0.090; **p< 0.01 Figure 1. Model of the Variables Affecting the Administrators’ Servant Leadership Details of the analytical data are given in Table 2. This Table shows the goodness of fit (R 2 ) of the data to the model. The β values given in Figure 1 are again shown for completeness. It should also be noted that the RMSEA score was 0.09, which is higher than the usual cutoff level, and so is a limitation of this model. AWR ALM APM AAM ASL 0.371

125 Table 2. Variables Affecting Administrators’ Servant Leadership Variables R 2 β Administrators’ Servant Leadership (ASL) Administrators’ Achievement Motivation (AAM) Level of Administrators’ Achievement Motivation (AAM) Administrators’ Practice of the Four Sublime states of mind (Loving-kindness, Compassion, Sympathetic Joy, Equanimity – APM) Administrators’ Learning Motivation (ALM) Level of Administrators’ Practice of Four Sublime States of Mind (Loving-kindness, Compassion, Sympathetic Joy, Equanimity – APM) Administrators’ Work Responsibility (AWR) Level of Administrators’ Learning Motivation (ALM) Administrators’ Work Responsibility (AWR) 0.629 0.853 0.543 0.734 0.793** 0.264** 0.733** 0.737** 0.857** Legend: **p < 0.01 The analysis of multiple correlation and beta weight (β) revealed that the administrators’ achievement motivation (β = 1.021) was the only variable that significantly influenced their servant leadership at the α = 0.01 significance level, accounting for 65.30 percent of the variance. The most important variable affecting the administrators’ servant leadership was their achievement motivation, which accounted for 62.90 percent of the variance. The variables affecting administrators’ achievement motivation were their practice of the four sublime states of mind (lovingkindness, compassion, sympathetic joy, equanimity), and their learning motivation, which accounted for 85.30 percent of the variance. The variable affecting administrators’ practice of the four sublime states of mind (loving-kindness, compassion, sympathetic joy, equanimity) was their accounted for 54.30 percent of the variance. The variable affecting administrators’ learning motivation was their work responsibility, which accounted for 73.40 percent of the variance. The results of analyses to determine the direct and indirect effect size of the independent variables on administrators’ servant leadership are shown in Table 3. Table 3. Direct/Indirect Size Effect of Independent Variables and Administrators’ Servant Leadership The levels of administrators’ servant leadership was directly influenced by their achievement motivation. It was also indirectly influenced by their learning motivation through their achievement motivation, while their work responsibility was influenced by their practice of the four sublime states of mind and their learning motivation. The direct and indirect effect size of these variables on administrators’ achievement motivation are shown in Table 4. Independent Variables Levels of Administrators’ Servant Leadership Effect Size Direct Indirect Total Administrators’ Achievement Motivation (AAM) Administrators’ Learning Motivation (ALM) Administrators’ Work Responsibility (AWR) Administrators’ Practice of Four Sublime States of Mind (Lovingkindness, Compassion, Sympathetic Joy, Equanimity – APM) 0.793 - - - - 0.582 0.653 0.210 0.793 0.582 0.653 0.210

126 Table 4. Direct/Indirect Size Effect of Independent Variables and Administrators’ Achievement Motivation Independent Variables Levels of Administrators’ Servant Leadership Effect Size Direct Indirect Total Administrators’ Learning Motivation Administrators’ Practice of the Four Sublime States of Mind (Lovingkindness, Compassion, Sympathetic Joy, Equanimity) Administrators’ Work Responsibility 0.733 0.264 - - - 0.823 0.733 0.264 0.823 The level of administrators’ achievement motivation was directly influenced by their learning motivation and their practice of the four sublime states of mind. The level of administrators’ achievement motivation was indirectly influenced by their work responsibility through their practice of the four sublime states of mind, along with their learning motivation. Discussion The research results show that educational institute administrators in the region surveyed viewed their servant leadership at the ‘high’ level, with slight variation among administrators. This may be due to their roles and duties. Professional standards require that all school administrators perform regular academic activities related to professional development for their educational administration, aiming at developing educational personnel, learners, and communities. Therefore, administrators with achievement motivation are those working with purpose, willing to reach their goals with enthusiasm despite the troubles and obstacles. These administrators attempted to find ways to solve problems by using their ability to be successful. This resulted in their servant leadership being at the ‘high’ level. The results are consistent with those of Chaiyapheuk (2007), Meemheuy (2008), Lengpanich (2011), Nithananon (2014) and Phongpanphaengpanga (2014) who studied administrators’ servant leadership and found that the servant leadership of the educational institute’s administrators they studied was at a ‘high’ level. The pathway analysis results revealed the following. 1) Administrators’ achievement motivation was the variable that explained 62.90 percent of the variance of their servant leadership. Practice of the four sublime states of mind (loving-kindness, compassion, sympathetic joy, equanimity) and administrators’ learning motivation did not directly influence servant leadership, but indirectly did so through the administrators’ achievement motivation. Administrators’ achievement motivation did not have a direct influence on their servant leadership, but indirectly influenced it through their practice of the four sublime states of mind, and their learning motivation driven by their achievement motivation. 2) Administrators’ achievement motivation was accounted for by their learning motivation and their practice of the four sublime states, which explained 85.30 percent of the variance. Administrators’ work responsibility did not have a direct influence on the level of their achievement motivation, but indirectly influenced achievement motivation through their practice of the four sublime states of mind and their learning motivation. 3) Administrators’ work responsibility was the variable that best accounted for their practice of the four sublime states of mind, explaining 54.30 percent of the variance. The variable accounting for administrators’ learning motivation was their work responsibility, explaining 73.40 percent of the variance observed. This is because achievement motivation is an important factor that helps administrators to perform activities towards goals and solve problems appropriately. This also motivates administrators to work according to good standards, which is the highest goal for assessing a person’s ability level (Singh, Granville & Dicka, 2002). The reason why administrators’ practice of the four sublime states of mind and their learning motivation influenced their achievement motivation may be that people with learning motivation always desire to learn new things. They are also curious and alert to new things. As a result, these

127 people become achievement motivated and perform duties to succeed at their highest level of efficiency, which assures that the schools they work for will be successful because of their ability and efficiency. The results of this study are consistent with those of Innok (2013), who found that administrators’ achievement motivation directly influenced their academic leadership, similarly to administrators’ servant leadership. The reason why administrators’ practice of the four sublime states of mind influenced their servant leadership was that administrators who had the four sublime states of mind (loving-kindness, compassion, sympathetic joy, equanimity) are usually kind and willing to help other people. Recommendations Recommendations for Policy Makers The study showed that achievement motivation directly and indirectly influenced the level of servant leadership by educational administrators. Therefore, administrators of the Ministry of Education and the Office of Basic Educational Service Area should promote achievement motivation by educational institute administrators in order to develop their servant leadership, which will improve educational achievement of basic education. Recommendations for Practice Top-level administrators should encourage educational administrators’ to enroll in training that will provide motivation for higher achievement. This should be done in order to develop servant leadership that will enhance their teachers’ teaching quality, leading to better student learning achievements. Recommendations for Research Courses should be developed to enhance administrators’ achievement motivation, as it has been found to be a critical variable effecting their servant leadership, which could further improve the educational quality in schools. About the Author Dr. Phitzesa Phongpanphaengpanga is Lecturer in the Faculty of Education at Vongchavalitkul University, Nakorn Ratchasima, Thailand. References Boonruangrut, S. (2013). Teaching material on research statistics. Nakhon Ratchasima: Vongchavalitkul University. Chaiyapheuk, B. (2007). A study of servant leadership characteristics and school climate of infant Jesus Sister Schools. Master’s Project, in Educational Administration. Bangkok: Srinakharinwirot University. Retrieved from http://thesis.swu.ac.th/swuthesis/Ed_Adm/Bang-orn_C.pdf Greenleaf, R. (2002). Servant Leadership: A journey into the nature of legitimate power and greatness. New Jersey: Paulist Press. Innok, N. (2013). Variables affecting to a level of academic leadership of school directors. Doctor of Education Dissertation in Educational Administration. Nakhon Ratchasima: Vongchavalitkul University. Lengpanich, J. (2011). A structural equation model of servant leadership for basic school administrators. Doctor Philosophy Dissertation in Education Administration, Khon Kaen: Khon Kaen University. Retrieved from http://programevaluation.kku.ac.th/student/student_detail/517050030?locale=th Martin, M., Gregory, K., Stemler, S. (Editors) (2000). TIMSS 1999 technical report: International science report findings from IEA’s report of the third international mathematics and science study at the eighth grade. Chestnut Hill, MA: International Study Center, Lynch School of Education, Boston College. Retrieved from https://timssandpirls.bc.edu/timss1999i/pdf/T99_TR.book.pdf Meemheuy, T. (2010). Servant leadership, spirituality in organization and job performance of managers in a private food industry business in Bangkok. Thesis. Master of Sciences in Industrial Psychology. Bangkok: Kasetsart University. Retrieved from http://research.rdi.ku.ac.th/forest/Publish.aspx?PublishID=15430

128 Mekkayai, P. (2007). The development of servant leadership training curriculum for student leaders of Kasetsart University. Master of Science Thesis in Industrial Psychology. Bangkok: Kasetsart University. Retrieved from https://www.regis.ku.ac.th/cpcmns/kugradNew/mis/gr_student.php?studentid= 47680145 Nithananon, M. (2014). Relationship between servant leadership and job performance of coordination teachers of the World Morality Revival Project. Master of Education Thesis in Education Administration. Bangkok: Kasetsart University. file:///C:/Users/relsec/Downloads/Fulltext%232_210845.pdf.pdf Phongpanphaengpanga, P. (2014). A model development of servant leadership for educational institution administrators of the Lower Northeastern region. Doctor of Philosophy Dissertation, Educational Administration. Nakhon Ratchasima: Vongchavalitkul University. PhraBrahmagunabhorn (Payutto). (2010). Buddhism based education (11th)Ed.). Bangkok: Pimsuay Company Limited. Roonjaroen, T. (2010). Professionalism in management and administration of education in the education reform era (7th Ed.). Bangkok: Khao Fang Printing House. Rueanthai, S. (2008). The relationship between servant leadership, team working, organizational citizenship behavior and job performance of managers at a private communication organization in Bangkok. Thesis. Master of Sciences in Industrial Psychology. Bangkok: Kasetsart University. Retrieved from http://digital_collect.lib.buu.ac.th/journal/Huso/v16n25-26/165-181.pdf Singh, K. Granville, M. & Dicka, S. (2002). Mathematics and science achievement: Effects of motivation, interest, and academic engagement. Journal of Educational Research, 95-(6), 323-331. Retrieved from http://dx.doi.org/10.1080/00220670209596607 Sinlarat, P. Teerapijit, S. & Chaodamrong W. (2010). The development of Thai education to international standards: the comparison between China, Vietnam, Japan, South Korea, Finland, and Thailand (2nd printing). Bangkok: Chulalongkorn University Printing House. Retrieved from http://jfklib.oas.psu.ac.th/ opac/BibDetail.aspx?bibno=1142783&page=2&recordsNo=15&Back=Back Sudman, S. (1976). Applied sampling. New York: Academic Press.

129 PUBLICATION ETHICS CATALYST STATEMENT OF PUBLICATION ETHICS “The publication of an article in a peer-reviewed journal is an essential building block in the development of a coherent and respected network of knowledge. It is a direct reflection of the quality of the work of the authors and the institutions that support them. Peer-reviewed articles support and embody the scientific method. It is therefore important to agree upon standards of expected ethical behavior for all parties involved in the act of publishing: the author, the journal editors, and the peer reviewer.” (Elsevier, Publishing Ethics Guidelines) DUTIES OF AUTHORS Original Work: Catalyst expects that manuscripts submitted to be considered for publication are the author(s) own original work, are not currently under review elsewhere, and have not been previously published in any language. Manuscripts should not “recycle” essentially the same material that has been published elsewhere. Turnitin software is used to check for plagiarism and/or previous publication, and Catalyst rejects articles that have substantial proportions of text that are copied from other sources. Authorship: The names of authors appearing on manuscripts should be limited to individuals who have made significant contributions to the article. Individuals who have provided technical assistance or support may be thanked by mentioning their contributions in the Acknowledgements. The corresponding author should ensure that all co-authors have seen and approved the final version of the paper, and have agreed to its submission for publication. Accuracy: Authors of original research papers should present an accurate account of their work along with an objective discussion of its significance. Underlying data should be represented accurately in the paper; adding data or removal of outlier results are unacceptable. Conclusions should be based on evidence presented in the paper, not on personal opinions. Errors: If significant errors/inaccuracies become apparent after an article has already been submitted, Catalyst expects authors to immediately contact the Editorial Staff so that appropriate corrections may be made in a timely manner. Acknowledgement of Sources: A research paper builds upon previously published work. Author(s) should paraphrase ideas or results that have been previously published elsewhere in their own words, along with citing these works in the paper and listing them in the references. Making statements of facts or ideas without citing evidence to back up these statements is not good research practice. Disclosure of Financial Support and Conflicts of Interest: Financial support received for research and the paper writing process should be disclosed in the acknowledgments, and any potential conflicts of interest must be disclosed to the Catalyst editorial team when a manuscript is first submitted. Protection of Human Subjects: Catalyst expects that experiments using human subjects have been conducted in compliance with international standards and research protocols, and that they received prior review and approval from an Institutional Review Board (IRB) or ethics committee. Such approval and the approving entity’s name should be specified in the Methods section.

130 DUTIES OF EDITORS Confidentiality: The editors of Catalyst use a “double-blind” peer review process, where neither authors nor reviewers know each other’s identity. The editors endeavor to protect the identity of author(s) and reviewers throughout the review process. Unpublished materials from a submitted manuscript must not be quoted or referenced by an editor without the express written consent of the author. Information or ideas obtained through peer review must be kept confidential and not used for personal advantage. Objectivity: Decisions on publication are made objectively by the Catalyst Administrative Board after reviewing the submitted manuscript and the peer reviews. The importance of the article’s contribution to the existing research in its field, the quality of articulation of the argument, and the strength of evidence provided are critical factors in the decision to publish. Conflicts of Interest: Editors are not allowed to publish or co-publish articles that appear in Catalyst, and they must recuse themselves from the review process when they have a conflict of interest or personal stake in the publication of a research article. DUTIES OF REVIEWERS Confidentiality: Reviewers should respect the confidentiality of the review process. They should not discuss aspects of the work under review with other researchers until after an article is published. Unpublished materials disclosed in a manuscript under review must not be quoted or referenced by a reviewer without the express written consent of the author, requested through the editor. Information or ideas obtained through peer review must be kept confidential and not used for personal advantage. Objectivity: Manuscripts should be reviewed objectively in the context of the reviewer’s expertise in the field. The importance of the article’s contribution to the existing research in its field, the quality of articulation of the argument, and the strength of the evidence provided are critical factors in reviewing the quality of a manuscript. Personal opinions without backing evidence should not be used as criteria for review decisions. Conflicts of Interest: If the reviewer realizes, after receiving a manuscript for review, that he or she has been involved in the research described, knows the researcher(s) involved in the research, or for any reason cannot give an objective review of the manuscript, the reviewer should inform the editors and decline to review it. Conflicts of interest may include competitive, collaborative, or other relationships or connections with any of the authors, companies, or institutions connected to the paper under review. Acknowledgment of Sources: Reviewers should point out important relevant published work that has not been cited by the authors. A reviewer should also call to the editor’s attention any substantial similarity or overlap between the manuscript under consideration and any other published paper of which they have personal knowledge. SUBMISSIONS Unsolicited manuscripts are welcomed. Submissions must be in the form of completed articles; partially completed articles are not considered. Articles published elsewhere will not be considered. Full manuscripts of abstracts published in any form will be considered, but editorial judgment will be exercised to accept or reject.

131 The journal’s policy is not to reconsider an article that has been rejected by its referees. Papers that primarily feature negative results are also unlikely to be accepted, as are those which focus on promoting a concept by selectively choosing positive results. Papers rejected by the editorial team are not open to reconsideration. Rejection may be based on a number of considerations including relevance, interest to readers, timeliness, adoption of a writing style that is considered offensive or partisan political, outside the space considerations stipulated by the journal, or failure to follow the specified guidelines.

132 Catalyst Publishing Guidelines Following are some general guidelines for Catalyst article writing and submission. 1. Manuscripts for Submissions 1.1. Manuscripts should be no more than 8-12 pages in length 1.2. Manuscripts should be written in correct and standard academic English. 1.3. Manuscripts should be single-spaced. 1.4. Manuscripts should use Calibri font size 11. 1.5. Manuscripts should contain minimal formatting (bold and italics commands are acceptable). 1.6. Manuscripts should not contain editorial controls of any kind. 1.7. Manuscripts should also contain a 150-200 word abstract of the article, a list of at least 3-5 key words, and a brief descriptor of the author(s). 2. Suggested Referencing Format 2.1. In order to maintain a consistent look for Catalyst, use of the American Psychological Association (APA) publication guidelines is mandatory. While authors may be accustomed to formats that are specific to their particular disciplines, the journal has greater consistency if one format is followed, and the APA format provides this consistency. 2.2. Catalyst follows the APA style for endnotes (preferred), in-text citations and references. 2.3. Referencing Samples: Please refer to http://www.apastyle.org/ for questions about referencing, and serialization in the APA format. 3. Submissions Procedure 3.1. Manuscripts should be in MS Word format, and should relate to one of the relevant disciplines. 3.2. Manuscripts should adhere to the Catalyst Publishing Guidelines; failure to comply with the guidelines may result in the rejection of a submission. 3.3. Manuscripts may be submitted through the online submission system found in “submission request.” 3.4. Manuscripts should be submitted by no later than December 31 for the April issue, April 30 for the August issue, and August 31 for the December issue. Nonetheless, even if a submission is received before the stated dates, and is accepted, the Catalyst Administrative Board reserves the right to decide on the time of publication. 4. Review Process 4.1. Manuscripts will be checked for originality and prescreened by the editorial staff to see if the article’s structure, content, and formatting complies with journal standards. 4.2. Manuscripts will undergo a double-blind review process that takes about 4-6 weeks (depending on the response rate of individual reviewers). 4.3. Authors will be informed of reviewers’ comments as soon as they are available. 4.4. Authors will be given about 1 month to revise their papers (should that be necessary) and should return the revised version by the 15th of November/March/July (or earlier).