JELIS ? Journal of Education in Library and Information Science ...

Initiation, Cultivation, Separation and Redefinition: Application of Kram?s mentoring framework to doctoral education in Information and Library Science

Cassidy R. Sugimoto

School of Library and Information Science, Indiana University Bloomington, 1320 East 10th Street, LI 013, Bloomington, IN 47405-3907. Email: sugimoto@indiana.edu

This work seeks to describe the current state of doctoral education in Library and Information Science (LIS) using Kram?s (1983) mentoring framework, which classifies mentoring into four linear stages: initiation, cultivation, separation, and redefinition. The data was gathered through 30 interviews and surveys of more than 200 faculty members in LIS. The results provide an exploratory description of doctoral education utilizing Kram?s framework. The results are particularly informative for the producers and consumers of doctoral education.

Keywords: mentoring, doctoral education, Kram, interviews, questionnaires

Introduction

Doctoral advisors are seen as the most influential mentors to LIS doctoral students during the dissertation process (Sugimoto, 2010) and the advisor-advisee relationship has been shown to be the most critical element in doctoral education (Heinrich, 1991, 1995; Heiss, 1970; Zhao, Golde, & McCormick, 2007), but the behaviors and actions of participants in the LIS doctoral mentoring relationship remain under-examined. As faculty advisors are the ?gate keepers to the scholarly profession? and the ?socializing agents of the discipline? (Girves & Wemmerus, 1988, p. 171), understanding this mentoring process provides understanding not only of the current state of doctoral education, but also the future of the profession and discipline. (Please note that here and throughout the text, mentor and advisors are equated; see Sugimoto (2010) for a thorough discussion of this distinction and empirical evidence that LIS doctoral students consider their advisors as mentors.)

Kram?s (1983) mentoring framework has been widely cited and used in academic and professional literature to provide a platform for discussing the process of mentoring (e.g., Ahuja, 2002; Kaplowitz, 1992; Phillips, Carson, & Carson, 1994). Kram gathered data from 18 relationships in a corporate setting and, from this data, created a four-phase framework to describe the mentoring process. This framework starts with the concept of initiation, in which the mentoring process begins through mutual engagement in the mentoring relationship. The key element of this stage is fantasy?the mentee fosters ?a strong positive fantasy? that the mentor will be able to provide the ?support and guidance? that she needs (Kram, 1983, p. 614). The mentor?s fantasy involves seeing the mentee as ?someone who can become an object for the transmission of the senior manager?s values and perspectives on the world? (Kram, 1983, p. 615). During the initiation phase, these fantasies are (ideally) positively reinforced by met expectations.

The second phase is cultivation, in which these fantasies are constantly tested and evaluated. During cultivation, the mentee begins to acknowledge his ?growing sense of competence? and his ability to more effectively ?navigate? through the new organizational system (Kram, 1983, p. 616). The mentor becomes empowered by her ability to cultivate a new member into the norms of the organization.

The separation phase, often marked by ?turmoil, anxiety, and feelings of loss? is the stage in which the mentor and mentee gain a sense of ?independence and autonomy? as the relationship becomes ?a less central part of each individual?s life? (Kram, p. 618). This separation is often initiated by structural and psychological changes. Although premature separation can cause psychological stress, in many cases the separation stage provides a feeling of pride for the mentor as they watch their mentee successfully navigate the culture for which they have been prepared.

In the final phase, redefinition, the mentor and mentee redefine the roles that they previously had. In the ideal case, this stage is marked by collegiality and friendship (Kram, 1983, p. 620), although a hierarchy still exists. However, in some cases, as the boundaries in the hierarchical system are challenged, feelings of resentment and hostility can manifest.

This framework may be suitable for the doctoral process, in which initiation could be seen as the process of selecting an advisor, cultivation could be interpreted as the disserting phase, and separation could occur at the point when the student completes the doctorate or decides to withdraw from a particular program (Maack & Passet, 1994).

This work utilizes Kram?s framework for an exploratory analysis of the doctoral mentoring process in LIS. The objective of this work is to describe the relationships between advisors, committee members, and doctoral students during the doctoral program and immediately following graduation, framed by Kram?s four phases. The results of this research describe the current state of mentoring in LIS doctoral education and provide recommendations for ways to improve the quality of this education. This research should prove valuable to creators and consumers of doctoral education and to those interested in the trajectory of the field.

Literature Review

Very few quantitative measures have been developed to assess academic mentoring (Johnson, Rose & Schlosser, 2007). Rather, much of the literature on mentoring describes the qualities, characteristics, and responsibilities of a good academic mentor. For example, mentors should provide assistance in networking (COSEPUP, 1997; Illes, 2002; Maack & Passet, 1994), training in grant-writing (Benderly, 2003; Dixon-Reeves, 2003; Maack & Passet, 1994), personal and professional support and encouragement (Ashford, 1996; Clark et al., 2000; Dixon-Reeves, 2003; Heinrich, 1991; Kartje, 1996; Long, 1987; Maack & Passet, 1994), opportunities to engage in research and presentations (Benderly, 2003; Clark et al., 2000; Dixon-Reeves, 2003; Maack & Passet, 1994), timely and constructive feedback (Hartnett, 1976; Heiss, 1970; Lipschutz, 1993; Lovitts, 2001), and sponsorship for desirable positions (Clark et al., 2000; COSEPUP, 1997; Lipschutz, 1993). There should be high levels of interaction between mentors and mentees (Ashford, 1996; Benderly, 2003; Gerholm, 1990; Girves & Wemmerus, 1988; Hartnett, 1976; Weiss, 1981) with the student being treated as a junior colleague (Bargar & Duncan, 1982; Girves & Wemmerus, 1988; Kartje, 1996; Lipschutz, 1993; Maack & Passet, 1994) and, through all these activities, the mentors should impart the norms and expectations of the discipline (Austin, 2002; COSEPUP, 1997; Illes, 2002; Lipschutz, 1993; Maack & Passet, 1994; Phillips, 1979).

Few theories or frameworks have been created out of these admonitions. Of those which have been created, Johnson, Rose and Schlosser (2007) identify four which have shaped much of the discussion around academic mentoring: Levinson and Darrow?s (1978) development of the theory of life structure, Kram?s (1983) delineation of mentoring into functions and phases, Hunt and Michael?s (1983) identification of the holistic elements of a mentoring relationship, and O?Neil and Wrightsman?s (2001) ?Sources of Variance Theory.? Many mentoring frameworks, including three of those stated above, use linear phases to describe the mentoring process: many start with an initiative phase, commence with incubation or cultivation, are followed by some type of termination, and end with a transition (Hall, 1976; Hunt & Michael, 1983; Levinson & Darrow, 1978; Kochan & Trimble, 2000; Kram, 1983; Merriam & Thomas, 1986; Wallas, 1926). Van Dyne (1996) describes informal mentoring using a linear flow model, but provides multiple options for the transition period, including friendship, ongoing mentorship, collegiality, or no relationship.

Disciplinary differences in graduate study

When examining mentoring, it is important to consider disciplinary differences. Many scholars have used anthropological metaphors to describe differences in disciplinary cultures (Becher, 1989; Becher & Trowler, 2001; Clark, 1977; Donald, 2002; Huber & Morreale, 2002), with one scholar writing that ?disciplinary communities can be seen as cultures, in which norms and habits of interaction are taken for granted and [are] invisible to insiders? (Zhao, Golde, & McCormick, 2007, p. 265). Zhao, Golde, and McCormick (2007, p. 265) describe the ways in which these cultural differences ?manifest themselves in the policies and practices of doctoral education, for example, how research is funded, and what the dissertation looks like, how a dissertation topic is selected, and how students and faculty interact.? In a study across multiple disciplines, Zhao, Golde, and McCormick (2007, p. 276) tested their theory of disciplinary differences in advisor behavior and student satisfaction and found ?pronounced disciplinary differences in the way doctoral students approach the choice of an advisor, and also in the way the advising relationship is conducted.?

As the dissertation (both as process and product) can be ?viewed as reflecting much of our academic and intellectual culture? (Isaac, Quinlan, & Walker, 1992, p. 242), it is important to recognize the disciplinary differences which govern the ?conceptual development, preferred approaches, common practices, and expectations?of the doctoral dissertation? (Isaac, Quinlan, & Walker, p. 244). These findings reinforce the need for studies specifically aimed at particular disciplines. Previous large-scale analyses of doctoral education have not included LIS (e.g., Bowen & Rudenstine, 1992; Walker, et al., 2009). LIS is an interesting avenue of study as it is often cited as an interdisciplinary field, incorporating elements from humanities, social sciences, and natural sciences. Investigating the practice of mentoring within LIS may have applications for emerging interdisciplinary fields, as well as other social science disciplines.

Methods

The sampling frame for this study was the population of all full-time faculty members from ALA-accredited schools in the United States and Canada, excluding adjunct professors, doctoral candidates, lecturers, instructors, emeriti professors, and visiting professors.

Questionnaires

Two sub-populations (?advisors? and ?advisees?) were chosen for inclusion in the electronic questionnaire. Advisors (n=374) were defined as tenured professors (at the rank of associate or full) from doctoral degree-granting ALA-accredited schools. It was assumed that these professors had the highest potential for serving as advisors to doctoral students. Advisees (n=311) comprised all assistant professor faculty members from ALA-accredited schools. It was assumed that these faculty members were most recently in the doctoral process and would be best able to provide accurate reflections on their experiences. These individuals do not represent all doctoral graduates from LIS program, but these graduates were chosen as representative of one type of successful graduate?those who have successfully navigated the doctoral process and have chosen a career trajectory in academe.

Two questionnaires, one for advisees and advisors, were designed in parallel, pilot tested with several faculty members and built in an electronic medium using Qualtrics survey software (http://www.qualtrics.com). Each question was specifically matched to Kram?s framework to provide responses consistent with this framework; for a matrix of each question and its relationship with Kram?s framework, please see Sugimoto (2010). Individual emails with links to the questionnaire were sent to the 648 potential respondents. The questionnaires were opened on January 29, 2009 and closed on March 4, 2009. Quantitative data from the questionnaires were exported to Excel and SPSS and analyzed using predominantly descriptive statistics. The answers to open-ended questions from the questionnaires were exported to NVivo for coding and analysis.

Interviews

The final question on the questionnaires asked individuals if they would be willing to participate in a follow-up interview. Contact information for 23 advisees who had received degrees within the field of LIS (self-identified) and 33 advisors was received. These respondents were emailed on March 31, 2009 with a request to participate in a 30-minute follow-up interview. A reminder email was sent on April 5, 2009. The first 30 individuals to respond to the request were selected for the interview phase of the study (although these individuals were split equally between the advisor and advisees, no explicit stratification was done during recruitment/selection).

Loosely-structured interviews were conducted over the telephone with each of the 30 respondents. Twenty-one interviews were conducted between April 13 and April 17, 2009, and the remaining 9 interviews were conducted between April 27 and May 1, 2009. Respondents were emailed one day before their scheduled interview and were given a list of themes that would guide the interview conversation (for this instrument and the questionnaire instrument, see Sugimoto, 2010). The interviews lasted approximately 30 minutes. Recordings of the telephone interviews were downloaded and imported into NVivo for transcription and analysis. Coding followed a mixed inductive and deductive approach. For the deductive coding, the elements of Kram?s mentoring model were chosen as concepts around which to organize the verbal statements. Additional inductive open coding was also conducted, in which ?codes are suggested by the researcher?s examination and questioning of the data? (Kelly et al., 2007, p. 1037). As these concepts arose out of the data itself, it required iterative listening and (re)coding of the recordings to ensure that each transcript was fully coded across all concepts. As Strauss (1987) describes, coding is used to ?fracture data, rearrange it into categories, and facilitate the comparison of data within and between categories? (c.f. Kelly et al., 2007). This process is complete when ?saturation has been reached and all relevant utterances have been classified? (Kelly, et al., 2007, p. 1037).

The data were then organized in an Excel spreadsheet in which each column represented a distinct concept, each row represented a distinct participant, and each cell represented the relevant utterance. When analyzing the results, all utterances in a column were evaluated, in order to provide a balanced report of the opinions and themes across all participants. Throughout this article, ID numbers will be used to identify a single participant. This method allows the reader to associate comments with unique participants and ensures that a single participant voice is not allowed to dominate the analysis.

Results

Ninety-three complete and six partial questionnaires were received for the advisee group, for a response rate of approximately 32%. One hundred seven complete and 33 partial questionnaires were received for the advisor group for a response rate of approximately 37%. One of the questions on the advisee questionnaire asked respondents to indicate whether or not they had received their doctoral degree in LIS. To focus on the scholarly development of LIS students, only the 75 advisees (75%) who indicated they had received their degree within the field of LIS were included in the analysis. All respondents of the advisor questionnaire were included (n=140).

The response rates were consistent with those found in Cook, et al.?s (2000) meta-analysis of response rates for Internet-based surveys from 56 surveys in 39 studies (34.6%) and Shannon and Bradshaw?s (2002) comparison between mail and electronic surveys (there was a 32% response rate for electronic surveys). The response rate was lower than the average found by Baruch, Holtorm and Brooks (2008) in their meta-analysis of 1607 studies published between 2000 and 2005 (55.5%). However, Baruch, Holtorm and Brooks found that reminders were associated with a significantly lower response rate than those studies without reminders. Therefore, reminders were not utilized in this study.

For the interview respondents, the advisees selected represented 14 different current institutions of employment and 10 different doctoral institutions (at which they had received their degrees). The advisors selected represented 9 different current and 12 different doctoral institutions. In the set of 30 interviewees, 19 unique current and 19 unique doctoral institutions were represented (31 unique institutions across both groups and types).

Initiation

In the questionnaire, advisors were asked to describe the practices at their institution for assigning/choosing dissertation advisees. The majority of the responses (out of n=107) indicated that the choice of advisor was entirely the choosing of the student (n=35) or ?negotiated with faculty upon initiation by the student? (n=23). Some respondents indicated that it was a cooperative or mutual process (n=18) with very few (n=6) indicating that the advisor chose or that advisors were assigned to the student. However, many respondents (n=29) noted a practice of assigning an initial advisor to guide the student upon arrival at the institution. This advisor was assigned based on a match between the student?s research statement and the faculty research area. Some respondents indicated that this advisor remained until the end of coursework or the comprehensive/qualifying exams, but most noted that the advisee was free to choose another advisor at any time.

The institutional differences in initiation were further explored in the follow-up interviews. One respondent described a two-tiered process at their institution, where all students are assigned to a single default advisor upon entering the program (ID641). The student was encouraged to choose a secondary advisor in their content area before beginning dissertation work, but retained the initial advisor as an administrative advisor. In another program, students were not assigned advisors upon arrival and were not instructed to choose advisors until they were ready to ?qualify? (ID506). Another respondent described the practice of sending out doctoral applications to the faculty for review, whereupon each faculty member selects one of three options: 1) reject; 2) accept and I will work with the person; or 3) accept, but I will not work with the person. The admission decision is based upon this feedback and the advisor is chosen from among the faculty members who indicated that they would work with the person. Although the respondent noted that ?there are times when people switch? he indicated that this was rare (ID398). The success rate of another matching system was mentioned by a respondent at a different institution who remarked that only ?5-10%? change advisors after the initial assignment (ID622). However, the matching system has consequences for institutions, such as having ?rejected eminently qualified students because we couldn?t figure out a match? (ID398). A similar practice was noted at another institution by a respondent who said, ?We will not accept somebody no matter how smart they are, no matter how academically qualified they are, unless there is some faculty member who is ready to work with them? (ID497).

In the questionnaire, advisees indicated the choice of selection was predominantly their own. The majority of the respondents (n=26) indicated that they chose their advisors, 8 indicated that there was a negotiation with student initiation and 2 indicated that the choice was mutual. Of those who reported being assigned to an advisor (n=10), many indicated that the assignment was due to their funding. A few students recalled being ?recruited? into the doctoral program by their eventual advisor (ID415), recruited from the institution?s master?s program (ID234; ID155), recruited from the profession as part of a funding opportunity (ID221) or working with eventual advisors on their application (ID478).

Advisors were asked to identify from among a list of characteristics on the questionnaires, which they considered important when initiating the relationship (Table 1).

Table 1. Advisor perception of important advisee qualities

Characteristic	Not important	Somewhat important	Very important	Total respondents
Student?s level of initiative	0	18	93	111
Match of the student?s intellectual interests with your own	2	21	90	113
Student is doing interesting research	0	35	77	112
You perceive the student will do a rigorous dissertation	3	34	75	112
Student?s academic record	15	55	42	112
You perceive the student will graduate in a timely manner	13	65	33	111
Reputation of the student within the department/discipline	18	63	32	113
Student?s personality	17	71	25	113
Student?s previous work experience	53	50	6	109
Student?s letters of recommendation for the program	61	42	10	113
Having money to support the student	65	44	4	113
Student?s standardized test scores	75	34	2	111

Note. Majority response for each characteristic shaded; characteristics ranked by level of importance.

The student?s level of initiative was ranked by the most respondents as ?very important.? Other items ranked ?very important? by the majority of respondents involved the research area and perceived research rigor of the student. Materials required for admission, such as standardized test scores, letters of recommendation, and previous work experience did not receive high rankings of importance. Being able to financially support the student was also ranked as ?not important? by the majority of the respondents. Of the additional responses provided by the respondents, ?curiosity? was the only characteristic mentioned by more than one individual.

Many interview respondents noted the importance of accepting students who worked in their area or were doing work that interested the advisor. One advisor noted: ?the people I work with have to, in a sense, convince me of a particular idea?that it?s worthy of dissertation work? (ID398). The respondent went on to report that he typically only accepts those working in his area. Other interview respondents talked about other characteristics of a successful student. One respondent said that what makes a student successful is ?being able to get involved intellectually with something you love and sticking to it; you can?t do a dissertation if you don?t love it? (ID342). One advisor noted that students should be ?hard-working, willing to work, and self-driven? (ID575); ?self-motivated? was noted by another (ID415).

Advisees were also asked to identify how important certain advisor characteristics are when choosing an advisor (Table 2).

Table 2. Advisee perception of important advisor qualities

Characteristic	Not important	Somewhat important	Very important	Total respondents
Intellectual interests match mine	1	18	53	72
Will make sure I do a rigorous dissertation	4	22	47	73
Reputation as a good researcher	5	20	47	72
Willing to take me	9	18	45	72
Is doing interesting research	8	21	44	73
Reputation as a good advisor	5	20	47	72
Knows the techniques and methods I will employ	6	32	34	72
Reputation for getting students through in a timely manner	15	20	38	73
Fosters a working environment I like in his/her research group	20	20	32	72
Recommended by other people	18	28	26	72
Can write a good recommendation letter to carry my career a long way	23	26	23	72
Reputation as a good teacher	23	29	20	72
Has money to support me	45	20	38	103

Note. Majority of response for each characteristic is shaded; characteristics are ranked by level of importance.

Those items which ranked the highest were the match of the advisor?s intellectual interests to the student?s, the advisor?s reputation as a good researcher and the perception that the advisor ?will make sure I do a rigorous dissertation.? These items and others ranked very important by the majority of respondents align well with the advisor responses. One difference between the lists is that the advisor listing of ?you perceive the student will graduate in a timely manner? as somewhat important and the advisee listing of ?reputation for getting students through in a timely manner? as very important. Having money to support the student was the only category which the majority of the advisees selected as not important.

Some respondents chose their advisors not just for what they could provide for them during the doctoral process, but also how they could help them shape their future. One noted that choosing their advisor ?was as much strategy as it was a topical spin,? describing their decision to choose someone who could advise them not as much on the dissertation work itself, but ?to advise me in areas that I saw as kind of future directions for the dissertation? (ID153). Another respondent choose an advisor based on ?the specific kinds of activities they were engaged in as scholars and the kinds of work and lifestyle attributes they exhibited? (ID74), in an attempt to emulate these behaviors.

Many questionnaire respondents indicated how previous relationships led to the advisor-advisee relationship, such as having the advisor as an instructor or doing research together before initializing the advisor-advisee relationship. Interview respondents reported similar experiences of working with an advisor on research or taking a course from the potential advisor before asking the advisor to take on the formal advising role (ID217; ID69; ID175). One advisor noted the importance of working previously with students before accepting them as advisees, saying: ?it?s a way for them to understand kind of my expectations?for me to kind of see if these are people that I want to work with? (ID499).

Cultivation

Meetings and proximity

In the questionnaire, advisors and advisees were asked about the frequency of meeting, before and after coursework was completed (Table 3).

Table 3. Frequency of meetings

	Time Period	Never	Once a year	Once a term	Once a month	Twice a month	Once a week	More than once a week
Advisor response(n=108)	Before the end of coursework	0 (0%)	2 (2%)	26 (24%)	43 (40%)	17 (16%)	20 (19%)	0 (0%)
	After the end of coursework	0 (0%)	0 (0%)	8 (7%)	43 (40%)	34 (31%)	21 (19%)	2 (2%)
Advisee response (n=72)	Before the end of coursework	1 (1%)	1 (1%)	18 (25%)	19 (26%)	12 (17%)	14 (19%)	7 (10%)
	After the end of coursework	0 (0%)	0 (0%)	7 (10%)	25 (35%)	22 (31%)	12 (17%)	6 (8%)

Note. Shading denotes plurality responses for each category.

In all cases, the largest percentage of respondents reported meeting once a month, with high levels of agreement between the groups. Examination of the difference between before and after coursework provides evidence that the frequency of meetings increases during the program.

Interview respondents reinforced the idea that frequency of meetings was not stable throughout the process. One advisor noted: ?the closer they get to defending?we get more and more meetings until they?re practically living in your office? (ID160). Another advisor remarked: ?once the dissertation is going we?re more or less in constant contact if they have questions? (ID415). The advisor went on to say that frequent contact is not only important for the student, but also for the advisor: ?I don?t want anyone to drop 700 pages on my head without my having had some influence on it? (ID415).

One advisee recalled how the frequency of meeting sped up her completion, as she was forced to have some sort of product for each meeting (ID155). An advisor reinforced the idea of holding students accountable by frequent meetings saying: ?I?ve never had it be successful where I haven?t seen them pretty often, like every two weeks during the proposal and dissertation phases?that?s such an unstructured phase of life that I find for most students, if they know they have to come in in two weeks and have to face me, they will have done something in those two weeks and so that just keeps them moving at a reasonable pace? (ID497).

Interview responses indicated that frequency of meetings was often dependent on other aspects of the relationship, such as the individual needs of the student (ID500) and whether or not the advisor and advisee were working together on a research project or teaching together (ID497). One advisor noted that collaborative relationships bring two people in physical proximity more often because ?both need something from it? and that ?when you have an advisee that?you just never see, I think it?s very easy to let them fall off the radar screen? (ID488). This experience was noted by another advisor who described working with a student from a distance ?more complicated? (ID478). One advisor recalled two ways in which she encouraged distant students to finish. In one case, the advisor would call the student every six months and say, ?Are you working on your dissertation?? (ID342). For another student, the advisor opened up their home as a space for the student to come for a week ?to just eat and write? (ID342). This advisor said:

?I think the hardest thing for a doctoral student is to be away, particularly doctoral students who are away and have full time jobs?the further away a student is, I think the harder it is to get through the dissertation simply because you are not getting the feedback and you are not around and other things tend to assume greater importance in your life. Doctoral students aren?t supposed to have lives.? (ID342)

Another faculty noted that ?if somebody is a physical presence, you know they?re there and they?re working? and went on to say that ?they?re probably more likely to finish? (ID507). This respondent offered the final warning: ?Don?t leave without the Ph.D. or it?s the kiss of death? (ID507).

Initiative

Initiative was one dominant component that emerged from discussions of cultivation. As one advisor said: ?I think the environment that we live in these days is such that we are all so busy that unless we are bothered by somebody coming and talking to us we are probably going to keep our nose down and keep going on doing what we have to do? (ID488). The theme of initiative was explored in the questionnaires, which asked the respondents to indicate who initiated instances of information exchange between advisor/advisees. As shown (Error! Reference source not found. 4), the majority of the respondents felt that information exchanges were equally initiated (61% of advisors and 46% of advisees selected equal initiation). Of the advisors, 30% of respondents indicated that the student was more the initiator, while 8% indicated that the advisor was more of the initiator. Of the advisees, 53% of respondents indicated that the advisee was more the initiator, while 1% indicated that the advisor was more of the initiator.

Table 4. Initiation of information exchanges

	Always the student	?	?	Equal initiation	?	?	Always the advisor
Advisors (n =98)	1 (1%)	8 (8%)	21 (21%)	60 (61%)	6 (6%)	2 (2%)	0 (0%)
Advisees (n=68)	9 (13%)	16 (24%)	11 (16%)	31 (46%)	1 (1%)	0 (0%)	0 (0%)

Note. ? denotes that the option was blank on the questionnaire; shading denotes plurality of responses for each question.

Interview respondents were asked to indicate the situations and points of the doctoral process in which either participant (mentor/mentee) was responsible for maintaining the relationship. The majority of the respondents indicated that the relationship was advisee-driven. One advisee recalled setting her own due dates for products and her advisor noting her ?self-driven? nature. Other advisees made comments that their chair let them ?call the pace? (ID234) and ?the whole relationship was sort of guided by me? (ID246). Another advisee noted that it was ?all proactive? on her part, recalling that her advisor was ?gracious and helpful whenever I asked him for anything? but that he expected his students to be ?adults? and, if they needed something, they needed to ask (ID86). This sentiment was echoed by another advisee who reported: ?I was never denied any help or assistance, but I did have to take that initiative? (ID221). One student reported his experience of taking the initiative and the role he thought this played in the doctoral process:

?I would say that most of the initiative came from me and they were mainly there to shape the initiative that I was taking?I think it?s extremely important?in my experience doctoral students who expect a lot of affirmative guidance from faculty members often either don?t get it or they just kind of flounder. I think it?s pretty important for the doctoral students to be the one taking the initiative and trying to push the process along.? (ID283)

Advisors reinforced this sentiment with comments such as ?I?ve never really had an advisee, I don?t think, where I?ve actually had to push them? (ID497), ?I think primarily doctoral work should be student driven? (ID495), and ?my students pick me?they drive the relationship? (ID415). As noted by one advisor: ?I?m expecting them to rise to the challenge of being an independent researcher?too much hand-holding yields Ph.D.s who may not be as productive as faculty members, because they?ve had too much sheltering, too much hand-holding? (ID499). This was echoed by another advisee who stated:

?a Ph.D. program is a self-selection process?it is largely an advisee-driven process, so if the advisee isn?t organized enough and doesn?t have their wits about them enough, hasn?t figured out the game enough to know to at least some extent what they need to do, if they aren?t motivated enough to make it happen, then maybe that?s part of the natural selection process, maybe they wouldn?t survive as a faculty member anyway?people know what junior faculty life is like and that coddling people?isn?t necessarily helpful in the doctoral program, that people have to be organized and self-motivated enough to make it through.? (ID641)

Separation

Responses from the questionnaire and interviews identified many potential points at which the formal relationship between an advisor and advisee can be terminated. Those at the initiation of the advisee include the advisee switching to another advisor, the advisee successfully graduating, or the advisee dropping out of the program (because this study focused on successful graduates, this option received only sparse treatment from the commentaries of advisors). The ways reported in which an advisor terminated the relationship were almost entirely due to the advisor leaving the institution. Although there were cases of committee members ?quitting? the student, it did not appear common for an advisor to terminate the relationship with their advisee.

On the questionnaire, advisees were asked if they changed advisors during the course of their dissertation; 25% (n=18) of the advisees indicated that they had changed advisors. The most frequently listed reasons were that the advisor left the university (n=6), the advisor died (n=4), or there were personal issues between the advisee and advisor (n=4). Also listed were a change in research interests (n=3), and retirement of the advisor (n=3).

The same themes were reinforced by the interview responses. Respondents noted advisor-driven separation in terms of the chair retiring (ID2), going on sabbatical (ID42), or passing away (ID155). Advisee-driven separation was most frequently noted in connection with a change in research interests. Advisees made comments such as ?it just wasn?t a fit for the kind of research I ended up doing? (ID283) and ?after a time it became clear to me that it just wasn?t going to work?we had a fundamental disagreement about how to move forward in the research that I wanted to do? (ID153). One advisor recalled that his ?old-fashioned approach? and ?methodical approach? (ID506) was a deterrent to many students. Other respondents reported that their advisors switched roles during the process, moving from the role of advisor to that of committee member (ID74).

The degree to which switching advisors occurred and was encouraged differed between institutions. An advisee from another institution remarked that ?there?s no hard feeling usually if you switch advisors,? saying most students at his institution did eventually switch from their initial advisors (ID153). Other advisees reported that the practice was not particularly encouraged, saying ?there was some expectation that the person who you started with is who would be the person who would eventually become your committee chair? (ID74).

One theme that was frequently discussed was the need to counsel students out of doctoral programs. Many respondents indicated that this was not done frequently enough, such as the following respondent who noted: ?We don?t have a point where we say, ?You?re not doing it, you?re out of the program??I don?t think we?re very good about counseling people out of our program? (ID495).

Some respondents noted the mechanisms in place at their institutions for counseling people out. One advisor explained the process at her institution in which the students are reviewed after the first year saying, ?it?s a graceful time where everybody can just stop or you?re at a point where you could, if they haven?t got a master?s degree, they could finish the requirements for a master?s degree and leave after another year or so? (ID497). However, the respondent went on to say that ?it?s hardly ever that anybody actually leaves at that point? (ID497).

Another respondent talked about counseling students out and also the possibility of counseling students ?in? if they were not finding their way in the program, saying: ?I think you could counsel a student in from the cold, but I?m not sure you?d want to? (ID153). The respondent noted ?If a student can?t figure out [how to succeed] for themselves, isn?t motivated enough to figure it out themselves, then once the mentorship relationship ends, they?re just going to sink again? (ID153). Another advisor reinforced the idea that some students are better off being counseled out, saying: ?I think it?s a hard thing to do, to counsel people out of a program, but probably in the long run, it?s the best thing you could do for them? (ID488). One respondent noted that much of this has to do with the career trajectories of their students, where ?the end goal is to try to help people into a life that they are comfortable with? (ID641).

As this study was predominantly focused on successful graduates, the point of separation occurred at graduation and the student?s physical move to a new location as a faculty member.

Redefinition

To assess how students redefine their relationships following graduation, the questionnaire asked advisors to describe (on average) their relationship with their advisees post-graduation and for advisees to describe their post-graduation relationship with their advisors. Choices included: 1) we are friends, 2) we are colleagues, 3) we are collaborators, 4) I am a mentor to my advisee (post-graduation)/my advisor continues to serve as my mentor, or 5) we have no relationship. Respondents were allowed to select multiple options.

The largest percentage of advisors (90%) perceived their advisees as colleagues after graduation (Table 5). More than 50% of respondents reported that their advisees were considered friends and collaborators and indicated that they remain in a mentor role to their advisees. Very few advisors (n=3) reported having no relationship with their advisee following graduation.

Table 5. Post graduation relationships

	Friends	Colleagues	Collaborators	Mentor	No Relationship
Advisors (n=93)	50 (54%)	84 (90%)	52 (56%)	50 (54%)	3 (3%)
Advisees (n=70)	46 (66%)	43 (61%)	16 (23%)	19 (27%)	9 (13%)

Advisees were most likely to see their advisors as friends after graduation, with 66% selecting this response. The next highest category was the perception of the advisor as a colleague, with 61% selecting this response. Less than 25% of the advisees saw the advisor as a collaborator. However, a few interview respondents talked about collaborating after graduation (ID175, ID74, ID155, 217). One advisee recalled: ?My advisor was extremely supportive and she continues to be?she was asked to edit a special issue of a journal recently and she contacted and asked me if I would be interested in co-editing with her? (ID175). Another student noted collaborating with her advisor on a grant application, but says that she did not collaborate during her program. She remarked: ?it seems like it takes?the senior faculty members?a few years to sort of recognize you as a colleague?you have to prove yourself as an independent person? (ID155). Another advisee also noted that she did not collaborate until after graduating (ID217).

In interviews, the uniqueness of each advisee was also noted, with one advisor saying: ?There is high variability across my advisees, in terms of whether they need/want me to remain in a mentor role after graduation. Some of them still ask for advice frequently, while others are more independent and we interact as peers? (ID497). Another advisee reported that, while she collaborated with her advisor during her doctoral program, her ?research agenda has matured in a different direction? so she doubts they will continue to collaborate (ID234). Another doctoral student noted the shift in the relationship post-graduation saying: ?We still like each other and still like to work together, but I have become independent? (ID69).

On the questionnaire, 27% of the advisees reported that the advisor continued to serve as a mentor. A lack of mentoring was lamented by one respondent who noted: ?I believe the biggest disservice done to doctoral students is that many of them are shoved out of the door with the thought that they can teach, research, and pursue service with little or no mentoring once we have a professional position? (ID42). However, other reported position mentoring, with one advisee noting how her advisor was ?following me until I get tenure?she?s keeping an eye on me to make sure I know what I?m doing? (ID 175). Other advisees noted the unique relationship of moving into a faculty position saying, ?you?re often being mentored and mentoring at the same time? (ID488).

Discussion

Initiation

Kram?s conceptualization of initiation involves mutual ?fantasies? (p. 615), in which the mentee believes a senior mentor can provide the support and guidance they need and the mentor believes the mentee to have some degree of potential. Kram describes a ?balance of initiative? from both parties. The examination of the Initiation stage in the doctoral process shares some similarities. There is certainly a degree to which both parties believe in the fantasy of a shared experience?many of the characteristics that are important upon initiation are potential rather than realized characteristics (such as the student?s ability to create a rigorous dissertation). The balance of realized to potential characteristics known before initiating the relationship varies a great deal depending on the structure of the program?an advisee may be assigned upon entry into the program or may be chosen after a set time period or academic milestone. Some respondents described how they must have previous knowledge of the advisee either as a student or research assistant before accepting them as an advisee. In situations such as these, there is a higher degree of known characteristics, although the fantasy of the successful mentoring relationship still exists. Programs may want to consider strongly whether the application process and program structure allows mentors and mentees the optimal ability to discover desired characteristics before engaging in the formal mentoring relationship, thereby reducing the chances of unobtainable fantasies.

A primary divergence from Kram?s conception is in initiative: whereas Kram identifies a balance of initiative, the findings from this study support a much more advisee-driven model, where the advisee is responsible for soliciting mentorship and the advisor serves in a passive initiation role?either accepting or declining the offer. The exception to this is in the cases where the advisor is assigned before entering the program. However, even in that case the student is initiating to some extent by applying to the program and, in many cases, explicitly stating the advisor(s) with whom the student would like to work.

Cultivation

Kram describes the Cultivation phase as a period in which ?the positive expectations that emerge during the initiation phase are continuously tested against reality? (p. 616). The findings from this study provide evidence that a successful cultivation phase requires frequent in-person contact between the parties. Although there were exceptions, advisors were better prepared to provide guidance and advisees were more prepared with deliverables when regular, in-person meetings were observed. Many respondents described the unstructured nature of the doctoral process and the way in which students floundered in the phase between the end of coursework and the dissertation defense. Frequent meetings were a way in which structure could be imposed on this nebulous stage.

As with the Initiation phase, initiative was a strong theme in the Cultivation phase. Respondents noted that the advisee was responsible for maintaining and propelling the relationship forward. Many respondents noted that the function of the doctoral process was to teach students to become researchers and future faculty. They mentioned that strong ?coddling? in this stage was actually detrimental to the advisee. The advisees should be developing themselves as independent scholars and be fully prepared to engage as junior faculty members upon graduation.

Separation

Kram?s Separation phase ?occurs when both managers recognize that the relationship is no longer needed in its previous form? (p. 620). Kram characterizes this period as one replete with ?turmoil, anxiety, and feelings of loss? (p. 618). Except in the cases where an advisor/advisee relationship is terminated before a successful graduation, this seems an inaccurate portrayal of the doctoral mentoring experience. The formalized separation of graduation seems to make it a less tumultuous experience in that both parties are aware of an explicit separation point in which the formal relationship will no longer exist. Although some respondents noted that the mentoring continued in the post-graduation phase, they were much more likely to characterize each other as friends and colleagues once the student successfully graduated.

Separation was also very much under the control of the advisee. There were few reports of an advisor terminating the relationship except for external reasons (another job opportunity) or death. An advisee was far more likely to terminate the relationship mid-studies and engage another faculty member. In addition to controlling the decision to stay with a particular advisor, it also appeared to be under the advisee?s control whether or not they continued in the program. Many respondents spoke of how difficult it was to counsel students out of a given program, remarking that students were never terminated from a program unless they made the decision to quit.

Redefinition

Kram describes the Redefinition phase as one in which the mentee becomes a peer to the mentor and the relationship becomes primarily a friendship (p. 620). Kram describes the diminished importance of career mentoring at this stage, in favor of psychosocial mentoring. This can be seen in the fact that the majority of advisors and advisees conceptualized the relationship as one of collegiality and friendship post-graduation.

In addition, this study identified an additional role played by the participants following graduation: collaborator. Although the relationship of colleague and collaborator may reduce some of the hierarchy of the mentor-mentee relationship, more than 25% of the advisees continued to see the mentor in a position to offer guidance and support following graduation. This may provide a type of role modeling as the student engages in their own mentoring; in this case, the student becomes the master through emulation of mentoring behaviors.

Visualization of Kram?s framework

While Kram?s model provides an adequate structural model for exploring the doctoral mentoring process, it does not provide depth into how these phases are negotiated or the factors that play into success at each of these stages. Mentoring in doctoral education can be seen to be structured by the programmatic phases of doctoral education. Figure 1 portrays these phases, annotated by the actors that are most dominant for each step of the process. The blue boxes are processes of Initiation, Cultivation, and Redefinition (from Kram?s model). Separation is indicated by red boxes.[1] The Separation processes are those items which provide movement from Cultivation to another stage of the process (either to Redefinition, back to Initiation, or terminate the process). The items in the green rounded box indicate those items which were heavily discussed by respondents but are not necessarily part of the programmatic process of doctoral education. In many cases, a doctoral student could graduate without engaging in these items. Institutions should evaluate their programs to the extent to which they embed these items into programmatic elements?for example, requiring an interdisciplinary minor, requiring teaching/research practica, providing opportunities for students to share ideas and receive feedback from each other, and providing the student with opportunities for engagement and networking.

Figure 1. Doctoral mentoring process/actor model

As is shown in Figure 1, the advisee is the primary driver through the majority of the processes. Doctoral programs may want to consider the implications of this and determine if there are certain processes for which the advisor, committee members or schools should play more dominant roles. For example, programs mostly become involved in separation when the student has successfully completed all the requirements of a program and the advisor predominantly becomes involved only for reasons not related to the student. The student is therefore given the majority of control in determining whether or not to continue in a program.

Determination of whether a student should be in control of the process and in which stages other actors should be involved is largely left to the philosophy of the school and individual advisors. However, to enable students to move more successfully through doctoral education, it is recommended that the school make the process and expectations explicit to the students. Many students may enter a doctoral program with the conception that they will be heavily guided. When this fantasy is unrealized, many programs may unwittingly allow the student to flounder until the student drops out. This has at least two implications: programs may be losing otherwise talented students who were not prepared for this unmet expectation and programs may be expending valuable resources maintaining students who will be unable to finish (or may finish only after a prolonged time). These issues could be greatly mitigated by in

lemonade diet steve jobs action figure chris jericho rose bowl johnny weir quadrantid meteor shower osu football