The PhD Supervisory Relationship Gap in the Accounting Discipline: A Study in a Malaysian Public University

This study examines the nature and quality of supervisory relationship of the PhD supervisors and PhD supervisees based on questionnaire survey on 100 PhD supervisors and PhD supervisees. The results show that PhD supervisee group tend to agree that having common research interest and professional reputation of the supervisors as an important factor whereas the PhD supervisors perceived the PhD supervisees’ past research and work experience as a major influence in choosing their supervisee. The results also show that there is a significant different in supervisory selection criteria between PhD supervisors and supervisees in terms of common research interest and PhD supervisee’s/supervisor’s work habit and personality. A key finding in this study is that the selection criteria of supervisory relationship characteristics vary between the PhD supervisors and supervisees. Therefore, it is recommended that PhD supervisees need to be aware of the supervisory selection criteria of their potential supervisors in order to be accepted by the later party. PhD supervisees also need to be aware of the supervisory relationship characteristics expected by their PhD supervisors need order to complete their study successfully. International Research in Education ISSN 2327-5499 2014, Vol. 2, No. 1 http://ire.macrothink.org 74


Introduction
The government, particularly in Malaysia, and the educational institutions of higher learning are striving to attract more students to embark on research supervision. Research supervision in education refers to a process of fostering and enhancing learning, research and communication at a higher level (Laske & Zuber-Skerritt, 1996). It is considered the highest and advanced level of teaching in the education system (Connell, 1985). Research supervision is intensive and interpersonally and often regarded as a critical factor to the success of graduate supervisees (Freeman, 1992;Dye, 1994;Dinham & Scott, 1999;Johnston, 1999). Dinham and Scott (1999) noted that "The student-supervisor relationship has the potential to be wonderfully enriching and productive, but it can also be extremely difficult and personally devastating".
In general, research supervision often involves two parties: the supervisors and the supervisees. These parties often have different perception on research supervision (Johnston, 1999). The supervisees often perceived an ideal supervision is a supervision that could assist them in their study where they learn about research and how to conduct research apart from achieving their personal and professional goals. Supervisors on the other hand often perceived that their supervision could provide guidance and create research situations. In addition, supervisors could conduct research projects with supervisees to enhance their learning, research as well as reputation (Laske & Zuber-Skerritt, 1996).
In most cases, supervisor is given the role to lead the supervisee in his/ her study process (Bernard and Goodyear, 1992). Throughout this process, a supervisor is expected to facilitate the supervisee and assist the supervisees. However, where an episode requires long process such as PhD journey, conflict may arise and magnitude of relationship may change over time which of consequence, could affect the supervisory relationship. Motivated by this argument, this study aims to examine the quality and nature of supervisory relationship between PhD supervisors and supervisees.
The remainder of this paper is structured as follows. The next section provides a review of relevant literature. Section 3 provides the research questions underpinning this study and section 4 outlines the research design. The results are presented in section 5. A summary and conclusion are provided in the last section.

Literature Review
Research on research supervision is becoming an eminent research interest for the past decades (Dye, 1994;Dinham & Scott, 1999). Within this area, one important component that has been researched into is the supervisory relationship between the supervisors and supervisees. The supervisory relationship has become a critical component to achieve postgraduate qualification due to the recognition that it has become a central process for successful completion of postgraduate study (Freeman, 1992;Dye, 1994;Dinham & Scott, 1999). It refers to two ways interact process between the supervisor and supervisee in continuingly engage each other within the spirit of open-mindedness, professionalism, respect and collegiality. It is seen as an interaction process that involves two parties with the same converging and diverging interests (Dye, 1994). The balance of these interests, therefore, is critical in ensuring success completion of postgraduate study.
The interest in researching supervisory relationship is due to factors that attributed to the postgraduate supervisees' success rate such as supervisee's dissatisfaction and low completion rates. Studies have shown that supervisee's satisfaction with their research study plays an important role to their study success (Hockey, 1991;Acker et al., 1994). They argued that it is important to get the relationship off to a good start and to maintain effective relations through the varying stresses and challenges of a research degree. Supervisees who have just started their study are sometimes unsure what expectations that their supervisors have (Cornwall et al., 1977;Ronnested & Skovholt, 1992). Therefore, it is often important that the supervisor and supervisee are clear about the expectations they have of each other and comfortable about re-negotiating expectations during the research process.
The area of supervisory relationship has been researched into in the past decade and a number of studies have evidenced the importance of characteristics of supervisory relationship. The characteristics identified in the literature could impact the perception of the supervisors and supervisees and of consequence, create an expectation gap between the supervisor and the supervisee (Moses, 1984;Grant, 1999;Gurr, 2001;Armstrong, 2004;Pearson and Kayrooz, 2005). However, there is still a lack of understanding on this issue and how supervisees could strategise in order to maintain good supervisory relationship. Understanding how supervisory relationship is accepted and acted upon by supervisees is an area worthy of more serious research efforts, to extend our understanding of the research process. This study aims to further examine the perceived supervisory relationship among PhD supervisors and PhD supervisees.

Research Questions
This study has two objectives. They comprise: (1) To determine the supervisory selection criteria of the PhD supervisors and PhD supervisees.
(2) To examine the perceived supervisory relationship between the PhD supervisors and supervisees.
The most important decision that PhD supervisees need to make before embarking for their research study is the selection of their PhD supervisors. Most PhD supervisees have the choice to choose who they want to be supervised with although there are PhD supervisees who were not given the privileged to do so. PhD supervisees who have the choice to choose their PhD supervisors often rely on a set of unplanned criteria (Ray, 2007) which resulted to regrets and lack of motivation which eventuate to poor quality of research output. Such arguments indicated that PhD supervisees need to set a strategy in selecting their potential PhD supervisors in an objective manner.
Generally, PhD supervisees could determine the several qualities that they expect of their PhD supervisors. These qualities however may or may not be of equal significance to the PhD supervisees (Ray, 2007). Arguably, the selection criteria become the critical factors that would determine the matching degree between the PhD supervisees and PhD supervisors (Eggleston & Delamont, 1983;Wright & Lodwick, 1989). In ideal circumstance, the PhD supervisees should be able to determine the criteria and its importance that they would their PhD supervisors to have and consequently, select their PhD supervisors who fits best (Ray, 2007). In a situation where PhD supervisees who were not given the choice to choose their PhD supervisors, often the PhD supervisors were given the choice by the graduate school whether the potential PhD supervisees that they have been allocated to could be accepted. Therefore, the following research questions are developed: RQ1: What are the supervisory relationship criteria of the PhD supervisors and supervisees?
The importance of the supervisory relationship often refers in two perspectives, namely the role of supervisor and the standard of supervision being delivered (Pearson & Brew, 2002;McMorland et al., 2003). PhD supervisors are responsible to do the monitoring role in the research progress and to ensuring their supervisees would master appropriate research skills as well as completing their study. Where issues arise between the PhD supervisors and PhD supervisees, many universities provide opportunities for the supervisees to change their supervisors although they are encouraged to try and solve any issues occurring between the PhD supervisees and their supervisors. This is because the PhD supervisors may not be aware of the supervisees' concerns and once known, would be able to take appropriate corrective action.
Studies in the education supervisory literature have identified that often supervisees who did not complete their studies have different expectations of their supervisors (Graham & Grant, 1997;Kam, 1997;Delamont et al., 2000, Smeby, 2000Pearson & Brew, 2002). These studies argued that the different in the expectations caused an expectation gap between the supervisees and the supervisors. Of consequence, the expectation gap may affect how supervisees progress in their research dissertations (Wisker & Sutcliffe, 1999). Other studies, however, suggested that other factors such as supervisory intervention and power relationships between supervisors and supervisees could affect the outcome for supervisees in completing their research dissertations. This is consistent with Sharp and Howard where they identified several problems experienced by supervisees on the research journey in completing their dissertation (2000, pp. 164). They noted that: "'There are a number of ways in which research progress may be affected by what may be termed "personal" factors such as illness, loss of motivation, occurrence of other opportunities and the need to search for a job"' The notation above indicates that there is an expectation gap on the perceive importance of factors related to supervisory relationship. Further, most studies were conducted on a different setting that did not specifically focused on PhD. Of consequence, there is still a lack of understanding on this issue and how supervisees could strategise in order to maintain good International Research in Education ISSN 2327-5499 2014 supervisory relationship. This study aims to examine this issue by examining the following research questions: RQ2: Are the PhD supervisors and supervisees satisfied with their supervisory relationship?
RQ3: What do the PhD supervisors and supervisees perceived the impact of supervisory characteristics on the degree outcome? RQ4: What is the perceived impact of supervisory characteristics on supervisee's satisfaction and degree outcome of the PhD supervisors and supervisees?

Sample Selection
The sample for this study is drawn from academics in the Faculty of Accountancy of a public university in Malaysia that were/ are currently PhD supervisors and PhD supervisees studying in local or overseas universities. The choice of this sample is considered appropriate because of the adequacy of obtaining a number of a match PhD supervisors and PhD supervisees. This sample is also chosen to enable the researchers to identify the respective PhD supervisors and their supervisees so their responses could be matched.

Development of Data Instrument
The questionnaire survey is adapted from the work of Drysdale (2001) with appropriate modifications to fit the purpose and context of this study. Two sets of questionnaire are developed. One set of questionnaire is developed for respondents who were/ are supervising PhD supervisees. The second set of questionnaire is developed for respondents who were/ are supervised by PhD supervisors.

Questionnaire for PhD Supervisors
For the first set of questionnaire which is intended to be given to PhD supervisors as respondents, there are three sections. Section A relates to demographic profiles. Demographic information on each respondent includes gender, age, position and supervising experience. The respondents are asked to complete these variables in categorical form.
In this questionnaire also, the respondents are asked whether they had a choice in supervising the intended supervisees. If their answer is yes, the respondents are asked on the importance of five factors in their decision to supervise their supervisees. The factors are common research interest, supervisees' past research and academic experience, supervisees' work habits and personality, recommendation from colleague and recruited by supervisee. The respondents are asked to complete this question on a 5-point scale of 1 being 'very unimportant' to 5 being 'very important'. Finally, in Section A, the respondents are asked on their satisfactory level in supervising the named PhD supervisee using a 5-point scale of 1 being 'very satisfied' to 5 being 'very dissatisfied'.
Section B seeks respondents' responses concerning the supervisory characteristics and characteristics of supervision. This section has two parts. The first part involves examining the impact of identified characteristics on the PhD supervisors' relationship with their supervisees. There are 13 questions in this part which includes availability to the supervisee, promptness in returning messages to supervisee, expectation made known to supervisee at beginning of program, schedule to accommodate demands in supervising supervisee, availability to discuss academic issues with supervisee, support given to supervisee's academic problems, belief in supervisee's ability, guidance in preparation of proposal, guidance in writing thesis, amount of constructive feedback provided to supervisee's research proposal, amount of constructive feedback provided to supervisee's thesis, and open to ideas on the direction of supervisee's research. The respondents are asked to response using a 5-point scale of 1 being 'very negative impact' to 5 being 'very positive impact'.
The second part of Section B involves examining the impact of identified characteristics on the PhD supervisors with the supervisees' studies and degree outcome. This part consists of four questions which include knowledge on supervisee's research topic, supporting letters, regular meetings during proposal, generation or writing stages and scheduled meetings with supervisee to accomplish necessary tasks. The respondents are asked to response using a 5-point scale of 1 being 'very negative impact' to 5 being 'very positive impact'.
Section C relates to examining the satisfaction and impact of identified characteristics of supervisory relationship with the supervisee's study and degree outcome. There are two parts in this section. The first part focuses on the satisfaction on the characteristics in supervisory relationship. There are 11 questions in this part which include supervisee's accessibility, schedule in meeting demands of graduate school, ability to meet deadlines, enthusiasm in his/ her research, attitude towards their studies, competency in undertaking a research project, willingness to seek new information, willingness in sharing authorship on joint projects, motivation to complete their program on time and willingness to admit mistakes. The respondents are asked to complete this question on a 5-point scale of 1 being 'very dissatisfied' to 5 being 'very satisfied'.
The second part of Section C seeks to examine the impact of the characteristics identified on supervisees' study and degree outcome. There are 11 questions in this part. All the eleven questions are similar to the questions in the first part. However, the responses are on the impact of the characteristics of supervisory relationship towards the supervisees' study and degree outcome whereas in the first part of Section C, the respondents are asked to response the eleven questions towards the supervisory relationship. The respondents are asked to response using a 5-point scale of 1 being 'very negative impact' to 5 being 'very positive impact'.

Questionnaire for PhD Supervisees
The second set of questionnaire is intended for PhD supervisees as respondents. There are three sections. Section A relates to demographic profiles. Demographic information on each respondent includes gender, age and years of completing program. The respondents are asked to complete these variables in categorical form. The respondents are also asked whether they International Research in Education ISSN 2327-5499 2014 have a choice in determining who would supervise them. If their response is yes, the respondents are asked to identify the importance of five factors, namely, common research interest, supervisor's professional reputation, supervisor's work habits and personality, recommendation from colleague and recruited by supervisor. The respondents are asked to complete this question on a 5-point scale of 1 being 'very unimportant' to 5 being 'very important'.
This section also asked the respondents whether they have changed their PhD supervisors during their program and if yes, their main reason. Among the reasons that the respondents could select are change in research interest, fail to meet supervisor's expectations, supervisor could not meet professional needs, supervisor moved or went on sabbatical or personality conflict. The respondents are also asked on whether they consider their supervisor to be their role model. Finally, in Section A, the respondents are asked on their satisfactory level in being supervised by the named PhD supervisor using a 5-point scale of 1 being 'very satisfied' to 5 being 'very dissatisfied'.
Section B seeks respondents' responses concerning the supervisory characteristics and characteristics of supervision. This section has two parts. The first part involves examining the impact of identified characteristics on the PhD supervisees' relationship with their supervisors. There are 13 questions in this part which includes availability to the supervisee, promptness in returning messages to supervisee, expectation made known to supervisee at beginning of program, schedule to accommodate demands in supervising supervisee, availability to discuss academic issues with supervisee, support given to supervisee's academic problems, belief in supervisee's ability, guidance in preparation of proposal, guidance in writing thesis, amount of constructive feedback provided to supervisee's research proposal, amount of constructive feedback provided to supervisee's thesis, and open to ideas on the direction of supervisee's research.
The second part of Section B seeks to examine the impact of the characteristics identified on supervisees' study and degree outcome. There are 13 questions in this part. All 13 questions are similar to the questions in the first part. However, the responses are on the impact of the characteristics of supervisory relationship towards the supervisees' study and degree outcome whereas in the first part of Section C, the respondents are asked to response the eleven questions towards the supervisory relationship. The respondents are asked to response using a 5-point scale of 1 being 'very negative impact' to 5 being 'very positive impact'.
Section C seeks to examine the satisfaction and impact of identified characteristics of supervisory relationship with PhD supervisors towards their study and degree outcome. There are 11 questions in this part which include supervisee's accessibility, schedule in meeting demands of graduate school, ability to meet deadlines, enthusiasm in his/ her research, attitude towards their studies, competency in undertaking a research project, willingness to seek new information, willingness in sharing authorship on joint projects, motivation to complete their program on time and willingness to admit mistakes. The respondents are asked to complete this question on a 5-point scale of 1 being 'very dissatisfied' to 5 being 'very satisfied'. ISSN 2327-5499 2014

Data Collection Procedure
Data collection was conducted between the months of January to July 2010. Three hundred questionnaires were distributed to PhD supervisees and their respective supervisors. The data collection involved two stages. The first stage involved sending out questionnaire to the PhD supervisors. Supervisor of each supervisee was identified from the records obtained from the Human Resource and Development office. The questionnaires were sent out to the PhD supervisors personally by the researchers. A set of questionnaire was given to each respondent requesting them to complete within a month. The questionnaire was coded to ensure that the PhD supervisor's responses be matched his/ her PhD supervisee. For example: if a set of questionnaire is coded 1 for the PhD supervisor, then the other set for questionnaire is also coded 1 for the PhD supervisee being supervised by the PhD supervisor coded '1'.
The second stage involved sending out questionnaire to the graduated and ongoing PhD supervisees. The simple random sampling from a list of PhD supervisees were selected to participate in this study. The questionnaires were sent out to the PhD supervisees via mail. A set of questionnaire was sent with a formal letter identifying the purpose of such study and requesting them to complete and return the questionnaire using the self-addressed envelope provided. Throughout the 6 months period of January to June 2010, 73 PhD supervisors responded and 86 PhD supervisees responded. A preliminary screening of the questionnaire received back has revealed that only 51 matching sample has responded to the questionnaire survey. As such, the questionnaire received from unmatched sample was excluded from the study. Table 1 presents the demographic profile of the respondents which has been divided into supervisee or supervisor. The characteristics of the respondents were evaluated on the basis of the following criteria: gender, age, their position in the organisation. The results in panel A, Table 1 show that out of the 51 supervisee that has responded, 13 (25.5%) were males and 38 (74.5%) were females. This scenario has appeared to be consistent with the supervisor where females tended to dominate (82.3%).

Demographic Profile
Panel B, Table 1 presents the results showing the age group of the respondents for both groups. The results show that a majority of the supervisee were from the age group of 31-49(80.4%). Only 3 (5.8%) of the respondents were less than 30 years old. It is interesting to highlight that 7 (13.8%) of the respondents from supervisee group were above 50 years old. In contrast, all (100%) of the respondents from supervisor group were above 40 years old. The results in panel D, Table 1 show that in terms of years after completing PhD, slightly more than half of the respondents were still in the process of completing their PhD. The results show that 28 out of the 51 (55%) respondents in the PhD supervisee group have yet to complete their research study. The remaining 45% of the respondents in the PhD supervisee groups have completed their research study with 7.8% graduated more than 4 years ago.
Panel E, Table 1 shows that almost half of the respondents in the PhD supervisor group have had a lot of supervising experience. At least 45 (88%) had more than 6 years of supervising experience. 21 (41.2%) had between 6-10 years of experience and 19 (37.2%) had 11-15 years of supervising experience. The result has shown that only 6 (11.8%) had less than six years of supervising experience. Table 2 shows the descriptive statistics of supervisory relationship which has been divided into PhD supervisees or PhD supervisors. Panel A, Table 2 presents the choice in selecting their PhD supervisor or supervisee. The results reveal that majority of the respondents (94%) in the PhD supervisee group agree that they have the choice to select their supervisor. Only 3 (6%) of the respondents in the PhD supervisee group were not given the choice to choose their supervisor. In contrast, more than half majority of the respondents in the PhD supervisor group (59%) responded that they have not given the choice to choose their supervisee. The results provide an earlier sign of possible conflicting needs of supervisory relationship.  Table 2 presents the results showing whether the two groups were happy with their supervisory relationship. The results show an interesting finding where 44 out of 51 (86.2%) of the respondents in the PhD supervisee group were satisfied with their relationship with their supervisor. Such results are not consistent to previous studies that show most PhD supervisees were not satisfied with their supervisory relationship (Young et al., 1987;Hockey, 1991;Acker et al., 1994). Similarly, 88.1% of the respondents in the PhD supervisor groups were satisfied with their supervisory relationship. Specifically, there are 20 (39%) respondents in the PhD supervisee group who were very satisfied with their supervisory relationship. In contrast, only 3 (6%) of the PhD supervisors were very satisfied with the relationship.

Supervisory Relationship
The results in panel B, Table 2 show that only a small number of respondents in both groups who were not satisfied or neutral towards their supervisory relationship. There were 2 respondents from both groups agreed that they were not happy with the relationship. Only 7 (13.7%) and 6 (11.7%) PhD supervisees and PhD supervisors respectively, agree that they are neutral or dissatisfied with the relationship, results which are consistent to Young et al. (1987), Hockey (1991) and Acker et al. (1994).
The respondents were also asked whether they consider their PhD supervisors to be a positive role model. The results in panel C, Table 2 show that only 19 (37%) of the supervisee agree to the statement. The remaining 32 (63%) of the respondents in the PhD supervisee group disagreed that their PhD supervisors have provided them a positive role model. The results provide an indication that the PhD supervisees did not look highly on their PhD supervisors.

Supervisory Selection Criteria
This section presents the response to the first research question (RQ1). RQ1 states that "What are the supervisory selection criteria of the PhD supervisors and supervisees?" Table 3 presents the results of the mean score for PhD supervisee group and PhD supervisor group. The results reveal that both groups have different opinion about the ranking of the factors. The PhD supervisee group agrees that having common research interest is a primary factor in choosing their supervisor (mean score=4.4510). In contrast, the PhD supervisor group perceived common research interest as the least important factor (mean score=1.4510).
In the eyes of the respondents of PhD supervisee group, the PhD supervisor's professional reputation is ranked second most important factor with a mean score of 4.3725. Whereas in the eyes of the respondents of the PhD supervisor group, they perceived the PhD supervisees' past research and work experience as a major influence in choosing their supervisee (mean score=3.9805). Both groups however agreed that to have a healthy supervisory relationship, previous experience working with each other prior to PhD research study is not an important factors in deciding their supervisory relationship (mean score=2.7059, ranked fifth and mean score=3.0196, ranked fourth respectively). Such results support the earlier results shown in panel B, Table 2. ISSN 2327-5499 2014

Satisfaction on Characteristics of Supervisory Relationship
This section presents the response to the second research question (RQ2) developed in this study. RQ2 states that "Are the PhD supervisors and supervisees satisfied with their supervisory relationship?" Table 4 presents the results of the mean score for PhD supervisee group and PhD supervisor group. The results show the extent of agreement between the two groups on their level of satisfaction on their supervisory relationship. The results reveal that both groups have different opinion about the ranking of the factors. The PhD supervisor group perceived the highest for the factor "my availability to this study" with a mean score of 4.5882, followed closely by the factor "I provide guidance on my supervisee's thesis" with a mean score of 4.4706 and the factor "My schedule could accommodate the demand of supervision" with a mean score of 4.4118.
Although the satisfaction level of supervisory relationship between the PhD supervisee group and the PhD supervisor group is somewhat similar, the different mean score given by both group reflect that the level of satisfaction on supervisory relationship of the PhD supervisee group is lower compared to the PhD supervisor group.
The results also show that the PhD supervisee group has the least satisfaction level on the following variable "Supervisor provide constructive feedback on my proposal" with a mean score of 3.7255, indicating that they were not happy with their PhD supervisors' performance in giving constructive feedback on their proposal. In contrast, the PhD supervisors have somewhat more higher satisfaction level that they have provided a good performance in giving constructive comments on their supervisees' proposal (mean score=4.3922), a finding that supports the results in panel B, Table 2. ISSN 2327-5499 2014

Impact of Characteristics of Supervisory Relationship on Degree Outcome
This section presents the response to the third research question (RQ3) developed in this study. RQ3 states that "What do the PhD supervisors and supervisees perceived the impact of supervisory characteristics on the degree outcome?". Table 5 presents the results of the mean score for the response of the impact of supervisory relationship characteristics on PhD supervisees' degree outcome. The results showing both PhD supervisor group and the PhD supervisee group agreement that there are few characteristics on the supervisory relationship that may impact the supervisees' degree outcome needs to be highlighted. Table 5, in most instances, both groups have no significant difference opinion about the supervisory relationship characteristics. However, it is interesting to highlight that there is significant different in terms of PhD supervisor being accessible to the supervisee. The PhD supervisor group believed that they find difficulty in accessing their supervisees (mean score=3.6275) whereas the PhD supervisee group felt that their supervisors could be accessible easily (mean score= 4.1569).

As shown in
In addition, the PhD supervisee group felt that their PhD supervisors' schedule is less in accommodating the demand of supervision (mean score=4.0588). In contrast, the PhD supervisor group believed that their work schedule could still accommodate the demand of supervision (mean score=4.4706). Such results provide an indication that there is an expectation gap between the PhD supervisor group and PhD supervisee group, results that are consistent to previous studies in the supervisory relationship literature.
The PhD supervisee group, on the other hand perceived the highest for the factor "Supervisor return messages promptly" with a mean score of 4.1176, followed by the factor "Supervisor is easily accessible" (mean score=4.0784) and "Supervisor discussed expectation at first meeting" (mean score=4.0784). Although the satisfaction level of supervisory relationship between the PhD supervisee group and the PhD supervisor group is somewhat similar, the different mean score given by both group reflect that the level of satisfaction on supervisory relationship of the PhD supervisee group is lower compared to the PhD supervisor group.
The results also show that the PhD supervisee group has the least satisfaction level on the following variable "Supervisor provide constructive feedback on my proposal" with a mean score of 3.7255, indicating that they were not happy with their PhD supervisors' performance in giving constructive feedback on their proposal.
The PhD supervisee group, on the other hand perceived the highest for the factor "Supervisor return messages promptly" with a mean score of 4.1176, followed by the factor "Supervisor is easily accessible" (mean score=4.0784) and "Supervisor discussed expectation at first meeting" (mean score=4.0784). ISSN 2327-5499 2014

Impact of Characteristics of Supervisory Relationship on Supervisee's Satisfaction and Degree Outcome
This section presents the response to the last research question (RQ4) developed in this study. RQ4 states that "What is the perceived impact of supervisory characteristics on supervisee's satisfaction and degree outcome of the PhD supervisors and supervisees. Table 6 presents the results of the mean score for PhD supervisee group and PhD supervisor group. The results show the impact of supervisory relationship characteristics on PhD supervisees' satisfaction and degree outcome. The results reveal that both groups have different opinion about the impact of supervisory relationship characteristics.
Specifically, the PhD supervisee group perceived the highest for the factor "My enthusiasm about my research" with a mean score of 4.7059, followed closely by the factor "My willingness to admit mistake" with a mean score of 4.6479. However, the PhD supervisor group perceived variables of "Supervisee's attitude about study" with a mean score of 4.3529 and "Supervisee's willingness to seek new information" with a mean score 4.2353 as the most important supervisory relationship characteristics.  ISSN 2327-5499 2014 Such results provide indication that there is an expectation gap between the PhD supervisor group and PhD supervisee group. The results show that both groups have different ideas on the importance of the supervisory relationship characteristics in determining supervisees' satisfaction and degree outcome. On the other hand, the PhD supervisors felt that their availability to their supervisees and the respect for supervisees' privacy/ personal issue are not important supervisory relationship characteristics in determining supervisees' satisfaction and degree outcome.

Summary and Conclusion
This study contributes to the literature in several areas. It provides a contribution to the understanding of the supervisory relationship in the PhD context. Understanding supervisory relationship could assist current and future PhD supervisors and supervisees to determine what are the characteristics of supervisory relationship that could leads to problems which eventuate to expectation gap. By tackling these problems, would hopefully reduce the expectation gap between the PhD supervisors and their supervisees.
This study examines the supervisory selection criteria of the PhD supervisees and PhD supervisors in selecting their supervisors and supervisees. The results show that PhD supervisee group tend to agree that having common research interest is a primary factor in choosing their supervisors. However, the PhD supervisor group perceived common research interest as the least important factor in selecting a supervisee. The PhD supervisees also perceived professional reputation of the supervisors as an important factor whereas the PhD supervisors perceived the PhD supervisees' past research and work experience as a major influence in choosing their supervisee.
The results in this study also show that PhD supervisors and PhD supervisees have different satisfaction level on the supervisory relationship characteristics. PhD supervisors were satisfied with their availability to their supervisees, provided guidance on their supervisee's thesis and being their schedule that could accommodate the supervision demand. PhD supervisees, on the other hand were more satisfied with their supervisors returning messages promptly, their supervisor being easily accessible and their supervisor discussed expectation at first meeting.
A key finding in this study is that the selection criteria and satisfaction of supervisory relationship characteristics vary between the PhD supervisors and supervisees. Therefore, it is recommended that PhD supervisees need to be aware of the supervisory selection criteria of their potential supervisors in order to be accepted by the later party. PhD supervisees also need to be aware of the supervisory relationship characteristics expected by their PhD supervisors need order to complete their study successfully.
This study is not without limitations. First, in this study the sample of the supervisee groups was taken from academics in the accounting discipline of a public university in Malaysia. Hence, it is not certain if the findings of this study can be generalized to other academic discipline or even accounting discipline from other universities.
Second, the findings in this study are based on the questionnaire developed from Drysdale (2001) with some modifications. There are other types of questionnaire instrument that could be adapted in the study. Using other questionnaire instrument may be able to provide different conclusion.