by
A proposal for a dissertation to be submitted
in partial fulfillment of the requirements
for the degree of
Doctor of Philosophy
University of Alabama
February 1997
Conventional wisdom holds that many technologically skilled doctoral students leaving their programs are motivated by financial concerns, and it is certainly true that the growing importance of information and communication technologies such as the Internet has led to an increase in the number of high-paying jobs for individuals with skills in the evaluation, organization and facilitation of access to information. However, no systematic study has been undertaken to determine whether there are aspects of the doctoral experience other than limited monetary rewards that influence the decision of these students to leave their programs.
To better understand the reasons for LIS doctoral student dropout, this study will use Brenda Dervin's sense-making methodology (Dervin, 1992a) to help identify communication breakdowns or "gaps" in the students' experiences that may lead to their departure from graduate study. The results will be used to make suggestions for better recruitment and retention of doctoral students, and will serve as a starting point for additional research into doctoral student retention in other disciplines.
Questions to be explored in future research could include:
Superficially, the methodological debate appears to present a clear choice between two opposing views. On one side are the champions of quantitative methods, logical positivists who believe that there is an absolute truth "out there," and that proper experimental design and statistical analysis will enable the researcher to elicit that truth in order to generalize from a sample to the larger population. On the other side are the defenders of qualitative or naturalistic measures, interpretive researchers who believe that human behavior cannot be reduced to discrete variables, and that the behavior of a sample, no matter how carefully chosen, cannot necessarily be generalized to a complete population. At first glance, this argument often appears to revolve around beliefs about the usefulness of "hard" data generated by the quantitative measures, versus that of "soft" data gleaned from qualitative work. However, a closer look reveals a much deeper epistemological difference underlying this debate.
Recent critical reviews of research in the areas of information needs and uses, and in communication studies (another field with a keen interest in the loss of its students to industry) have identified a paradigmatic shift in the views of what constitute appropriate research methods for studying human behavior (Cole, 1994; Dervin & Nilan, 1986; Hale, 1991; Schement & Ruben, 1993). In all of the social and behavioral sciences (as well as the so-called "hard" sciences), the idea that an absolute truth exists is being questioned by theorists and researchers, and this view is being replaced with a constructivist view that sees reality as a personal and subjective construction.
Dervin and Clark (1987) assert that statistical data is most valuable for the purposes of identifying the magnitude of a problem or phenomenon, whereas qualitative data is more useful in identifying the underlying causes. Underlying this difference is the difficulty in undertaking a quantitative study when the independent variables are unknown-the situation faced by most researchers into information-seeking behavior. Qualitative research allows the identification of variables that can then be further evaluated through descriptive and analytical statistical research.
Traditional LIS research has been built on the "information as thing" model articulated by Buckland (1991a; 1991b), a conceptualization that allowed researchers to count information interactions (number of transactions, number of records retrieved, etc.) as a measure of "successful" information use. This system-oriented perspective, which begins by assuming that the information, like truth, exists "out there" and must be moved into the users somehow, becomes increasingly difficult to support when one observes how little it has done to further the understanding of how and why people use libraries, online databases, schools, or other systems designed to "inform" (Dervin, 1991; Dervin, 1992b).
In the 1970s, Dervin (1976; 1977) began proposing that research based on understanding "communication" rather than "information" would provide a better theoretical base for research into library problems. Her sense-making methodology proposes a conceptualization of information as a user construct, rather than as a discrete external "thing." Dervin is not the only researcher who has called for such an epistemological shift. Her review article with Nilan in the 1986 Annual Review of Information Science and Technology (Dervin & Nilan, 1986), and Cole's (1994) review of the development of information as a subjective construct, note a growing dissatisfaction with the results of traditional information needs and uses research, and the lack of a clear and consistent theoretical base for that research. Cole, in particular, reviews the work of a growing number of researchers shifting toward a more constructivist approach. Overall, information needs and uses research has been moving away from methods that attempt to generalizes from sample to a larger population, and toward methods that focus on the situatedness of users' information needs.
In her discussion of the sense-making methodology, Dervin (1992a) tells us that this move need not be to a radical constructivist (or phenomenological) position, or a claim that generalizable results are not possible. Rather, she argues that from research on local constructs, the researcher can derive theoretical principles that can then be tested in other settings. She sees the primary difference in her approach as using "situatedness" as the generalizable aspect, rather than looking for across-time-and-space generalizations. Miles and Huberman (1994) also discuss the use of qualitative research to generalize not from the sample to the population, but rather from the sample to theory.
While other researchers have developed methodologies for understanding the information seeking process-most notably Kuhlthau's (1991) Information Search Process model Dervin's approach differs from most because it does not presuppose a knowledge of the user's information needs; instead, the methodology seeks to elicit information on both the needs and the uses of systems, institutions, and resources designed to inform. As a result, sense-making provides an excellent theoretical starting point for this examination of the doctoral student experience, a situation in which the researcher seeks to understand both the nature of the students' information needs and their ability meet those needs (make sense) through their doctoral program experiences.
Interpretive and qualitative theories, when operationalized in research projects, are often criticized for failure to account for reliability and validity. To answer these charges, one must examine the concepts of reliability and validity themselves. Dervin (1994) and Miles and Huberman (1994) note that these concepts as traditionally conceptualized rely upon the same epistemological presuppositions discussed above in the context of positivism. Reviews of traditional research methods texts (Katzer, Cook, & Crouch, 1991; Miller, 1991; Powell, 1985) provide these traditional definitions, while texts with discussions of qualitative methodologies (Babbie, 1989; Glazier & Powell, 1992; Miles & Huberman, 1994) provide some alternative views of the concepts.
In the traditional view, "reliability" is seen as the repeatability of results. This can take two forms: inter-instrumental reliability, or the question of whether the same methodology used in the same setting will generate the same results consistently; and inter-observer reliability, or the question of whether two researchers using the same methodology will obtain the same results. Fidel (1992) notes that these conditions of reliability cannot be met using a case study methodology, the success of which depends upon the ability of the observer to elicit information that another researcher might miss, information that is often unique to a specific time and place. Babbie (1989) notes that the inherent subjectivity of field research causes problems in reliability, but says that gathering data from multiple sites may present a way to overcome this weakness.
Miles and Huberman (1994) suggest that we need to reconceptualize the concept of "reliability" in the context of qualitative research, offering a variety of questions for the researcher to answer in developing a study: Is there meaningful parallelism between sites? Is there clear connection to theory? Has the researcher clearly indicated his or her role within the study, acknowledged biases, and ensured that the participants are fully aware of the purpose of the study and nature of the inquiry? Were data collected across a valid sample? And Dervin (1994) asserts that rather than striving for identical replication, researchers should attempt to achieve consistency in the time-space frame of situationality; that is, will individuals with similar information needs exhibit similar behavior in comparable situations?
There are two primary issues, then, that need to be addressed in order to ensure reliability in the results of this study. The first relates to sampling, as selection of representative subjects and the retrieval of results that can be generalized across those subjects is critical. Miles and Huberman note that in qualitative research, sampling tends to be purposive rather than random. Acknowledging not only that one researcher cannot study the entire universe, but also that the researcher cannot study any single segment that represents the entire population leads to a process where subjects or sites are selected based on their representativeness within a given situation.
Glaser and Strauss, in their influential work The Discovery of Grounded Theory (1967), advocate the use of "theoretical sampling" to generate information about elements and their connections within a theoretical framework. While this project will not be a strict application of the grounded theory approach (given that grounded theory demands that no variables be predefined), the selection of representative "slices" of the population to be studied will be consistent with the proposed methodology in this project. Miles and Huberman also recommend that sampling be linked to theory; accordingly, the students selected will need to represent varying populations, but comparable "situatedness." If sense-making theory holds true in this case, there should be reliably consistent results across the various programs based on their similar situational aspects.
Related to the concept of reliability, as well as to that of validity, is the issue of the Hawthorne Effect. That is, will the presence of an observer result in skewed results, results that may be consistent--and therefore on their face reliable--but in fact inaccurate and therefore invalid? (Miles and Huberman discuss the idea of the "broken thermometer"; it will produce results that are completely consistent, and thus reliable, but are dependent on that one instrument, and are in no way valid or generalizable to other situations.) It is difficult to compensate for this factor completely; however, the use of a qualitative, interview-based methodology can help the researcher to determine the extent to which the study itself has influenced the results in a way that a more rigid and quantitative study would not allow.
To address the topic of validity in qualitative research, the concept must be broken down into two separate constructs: internal validity (often called "criterion validity"), which relates to the ability of the research to accurately identify and measure a given variable, and to show causal relationships between independent and dependent variable; and external validity (also called "predictive validity") which addresses the extent to which the results in the study can be generalized to other settings.
When evaluating the internal validity of qualitative research, the question must go beyond the traditional "are the criteria clearly identified and causal relationships found." Paris (1988a; 1988b) notes that one problem with the application of criterion validity in social sciences is that precise operationalization and measurement are more difficult, if even possible, when examining human behavior. Miles and Huberman suggest that the question of internal validity should be reformulated as "Are these findings credible? Do they make sense? Do they have truth value?" They identify specific understandings that can come from qualitative research: descriptive (what happened in certain situations), interpretive (what the situation meant to participants), theoretical (what are the key concepts and their relationships), and evaluative (observer and participant judgments of worth or value). Babbie (1989) , in fact, believes that field research is in many ways more 'valid', especially in terms of internal validity, because of the richness of information received and the ability of the researcher to tap into a depth of meaning in the constructs.
In the context of this study, the data obtained from the sense-making interviews will assist in identifying the key concepts and variables that affect student perception and understanding of LIS doctoral programs. In a manner consistent with Glaser and Strauss' grounded theory, the research will be used to identify these variables, not just to measure them. The lack of research into this topic makes it particularly important that this study take an exploratory stance rather than a presumptive one. An interpretive framework such as sense-making helps to ensure that data gathered will be fully explored not just from an outside researcher's perspective, but with the active solicitation of the research subjects' views. Sense-making, in fact, specifically acknowledges the respondents as co-theorists in the research process (Dervin, personal communication, 1995).
In addressing the topic of external validity, the epistemological presuppositions of qualitative research must be revisited: specifically, the idea that the researcher can generalize from our results not to a complete population, but rather to a theoretical structure, one that can then be retested in other settings. Paris (1988b) notes that the case study and other qualitative methods establish external validity not based on a statistical generalization from sample to population, but rather on analytical generalization, allowing theory-building from observations, and retesting of that theory to verify and refine it. In particular, theoretical sampling encourages the likelihood of valid generalization by looking at a range of settings, carefully chosen to reflect a variety of aspects of the population being studied.
The attempt of this study, then, is not to generate results that can be applied in toto to all doctoral programs in all fields. Rather, it is to generate a theoretical framework that can identify the key aspects in student perceptions of and problems with LIS doctoral programs. By generating such a framework, identifying key variables in perception and experience of these programs, and looking for relationships between these variables, the researcher can begin to formulate ways for retesting and refining that theory in a variety of settings.
The true test of the reliability and validity of the results generated from an exploratory qualitative study come in projects based upon the initial results. An excellent example of this process is the work done by Kuhlthau in comparable educational settings (Kuhlthau, 1991). While she began with substantially qualitative work (a case study of a single group of high school students), she then used the results of the study to generate instruments for further testing of the variables identified in other settings, as well as for further testing of the students in the initial group over a longer period of time. The addition of both additional sites and a longitudinal view provided sufficient proof of the reliability and validity of her results to allow the development of a model that can be used in a wide variety of settings. Ideally, this research will eventually be extended in comparable ways. For the short term, however, the primary goal of this research will be the identification of the key variables and concepts in the area of LIS doctoral student attrition, with an eye towards the suggestion of changes in recruitment, program requirements, curricular design, or other aspects of the doctoral student experience to reduce that attrition rate.
The study will focus on a very specific group of students-doctoral students who have recently dropped out of their programs in LIS. It must be emphasized that such a study cannot and should not be used to generalize about LIS doctoral programs, or about doctoral students as a population. It can, however, be used to begin to identify those factors that appear consistently, across students and schools, to most influence the student's views of the program and the likelihood that they will complete their program. The goal is the development of a theory of LIS doctoral attrition that can then be tested in later studies.
Future research will also seek to expand the scope of this research to a wider variety of students and settings, and will be used to support or revise theoretical hypotheses about doctoral student expectations and experiences that are developed in this study. Additional research can explore whether the factors most influencing LIS doctoral students are also of importance in the experiences of doctoral students in other academic fields.
LIS: The field(s) known variously as "Library & Information Studies," "Library & Information Science," "Library Science," "Information Science," "Information Studies" et al. For the purpose of this study, only students in doctoral programs at schools or departments whose masters degree has been accredited by the American Library Association will be considered.
Doctoral Students: For the purposes of this study, doctoral student refers to students who have completed their master's degrees, and are enrolled in a course of study to obtain a Ph.D. in LIS.
ABD: An acronym for the phrase "all but dissertation," ABDs are defined as students who have completed all of the requirements for the doctorate with the exception of the dissertation.
Only one extensive review of the literature on LIS doctoral study has been published; that study (Abrera, 1987; Abrera, 1988) covered the literature from 1926 to 1980, and found that the majority of recent literature on the topic consisted of descriptive studies, many focusing on analyses of doctoral dissertations, and a number surveying doctoral graduates. In her review, Abrera divides the literature on the topic into three segments, separated by the appearance of what she calls a "landmark study": Danton's "Doctoral Studies in Librarianship." (1959)
Abrera defines the pre-Danton period as 1926 (the year the first doctoral students were admitted to Graduate Library School of the University of Chicago) to 1958. During this time, only eight publications on LIS doctoral study were published, none of which discussed students, graduates, or dissertations. Instead, all focused on the value and role of the doctoral degree in the field of LIS.
The Danton study itself was a quantitative study of multiple aspects of doctoral study--dissertations, program objectives, major fields of study, factors inhibiting attainment of program objectives, student withdrawal rate and time to completion, positions held by graduates, and the contribution of doctoral study as a whole to the profession. This range of topics is extensive, and the methodology used is unclear (Danton does not describe his instrument or surveying methods in the study), which makes it difficult to determine how accurate his results were. Of particular interest in his study, however, is his discussion of the doctoral student attrition problem, something ignored completely by the majority of literature on this topic. His examination of number of students admitted vs. those awarded the degree during the period from 1948 to 1958 showed ratios ranging from 8:1 at Chicago to 11.9:1 at Illinois. While some of those admitted during the period in question would probably go on to complete their doctorates at a later date, Danton notes that those figures "point to an unmistakable high attrition rate," and asserts that "unless the causal factors--e.g., the lack of substantial fellowship aid-should change we shall have to continue to expect a small proportion of doctoral graduates in relation to the total numbers who begin study at this level." (p. 442) Unfortunately, Danton provides no convincing evidence to support this causal connection between fellowship availability and completion rates. Danton also tells us that in other fields, attrition rates have been reduced by a reduction in the number of required courses, and the setting of fixed time limits for completion of the degree. Again, however, he provides no evidence to support this claim.
In the post-Danton literature from 1959-1980, more quantitative studies began to appear. Those studies followed in Danton's footsteps in their analysis of dissertations and graduates as well as assessment of programs. The overall number of publications on the topic of LIS doctoral study increased dramatically during this period, as well, with 61 items being identified and included in the Abrera's review. The focus of the non-quantitative studies done during this time was on descriptions of programs, opinion papers on the role of the doctoral program, and analyses of dissertations. Nowhere in Abrera's review of the literature, however, does she identify non-quantitative studies of doctoral students (as opposed to graduates) that investigated the phenomenon of attrition.
Of particular note in the post-Danton period is Marco's (1967) investigation of the programs themselves. His research, commissioned by the Research Committee of the American Library Association's Education Division, surveyed the deans and directors of the eight doctoral programs in existence at that time. Marco's study focused on the requirements for admission to doctoral study, curriculum and requirements for the degree, and the image or status of the program, particularly in the eyes of other academics on the campuses where the programs were located. While some of the respondents did comment on the needs of the doctoral students in their programs, no reference of any kind is made to the time-to-completion or attrition rates in these programs.
Since Abrera's literature review was published, only a few studies related to LIS doctoral study have appeared. Perhaps the most negative is Houser's "examination of the dissertation process in the context of scientific scholarship." (1982, p. 95) Like Houser's earlier work on the topic of librarianship as an academic discipline, this article criticizes a "lack of theory in the discipline which dissertations should be testing or which would guide the research conducted in the dissertations," and embarks on a review of literature related to the Ph.D. in library science. (p. 97) In that review, he expresses concern that the failure of LIS education--and, consequently, LIS dissertations--to meet his criteria of a scientific discipline or profession means that "dissertation production in library science [is] intellectual fraud" (p. 105).
Houser's criticism of library science as "unscientific" and even fraudulent in its pretensions has its earlier roots in his book The Search for a Scientific Profession (Houser & Schrader, 1978). In it, the authors evince a belief that as the study of information is increasingly undertaken by disciplines such as psychology and anthropology, "library science will come to have less and less claim as a social science and as a scientific profession." In that event, they imply, doctoral study and its resulting research would have little or no value.
This remarkably bleak view of the value of LIS education and research was strongly criticized by one of the most highly respected scholars in the field, Jesse Shera (1979). His scathing response to Houser and Schrader's work begins with the following:
This book has so many faults errors, misinterpretations, and misunderstandings that the reviewer is in a quandary as to where best to begin, or, indeed, whether he should begin at all. Nevertheless, a review is justified on the ground that it could be a potentially dangerous book. Its heavy burden of scholarly impedimenta might lead the unwary to believe there is important substance between its covers. In fact, rather than being scholarly, the book is sophomoric and polemical. (p. 310)Shera points out that Houser and Schrader's critique is based on a very small and selective sample of the literature, and ignores many examples that contradict their claims. Certainly this is true of Schrader's review of the literature on LIS doctoral study, since he asserts that "the literature about doctoral programs [between 1930 and 1980] is virtually nonexistent" (1982, p. 103), while Abrera (1987; 1988) was able to identify 61 published items in the period from 1960-1980 alone. It seems unwise, therefore, to rely on Schrader's assessment of the literature as an indicator of the state of LIS doctoral study.
Another relatively recent study that has been widely cited by scholars in the field since then is Bobinski's analysis of the development, status, and future of LIS doctoral education (Bobinski, 1986). At that writing, the number of institutions offering a doctoral degree in LIS had grown from eight to twenty four, creating an urgent need for an updated version of the original research done by Danton and Marco. Much of Bobinski's data is drawn from the annual statistical reports compiled by the Association for Library and Information Science Education (ALISE).
Using the ALISE statistical data, which included items such as number of degrees granted annually, as well as HEA fellowships awarded annually, Bobinski was able to identify a correlation between financial support and completed doctorates. While this certainly suggests that financial support is a factor in LIS doctoral attrition, without in-depth qualitative research to verify the connection it is difficult to claim a causal relationship.
Bobinski's data was supplemented by responses to a letter sent to deans and directors of LIS programs, providing more detail on the admissions requirements and courses of study for their doctoral students, on perceptions of quality of their program (their own views, as well as their perceptions of the way the program was viewed by others on campus), and on recent or projected changes in the program.
In his analysis, Bobinski states that "doctoral study in library and information science seems to be in a state of transition," with no new programs having been established in the preceding ten years, three about to open, and three about to close. In fact, there has been a great deal of change in the field of library education during the ten years since Bobinski's study was published. While several library schools have closed their doors, others have changed their names, taken on new areas of study, and even merged with communication, management, education or computer science programs; Powell (1995) identifies twenty-seven doctoral programs in the US and Canada as of 1995, and all of those programs are still in existence today. With the growing emphasis on the study of information and technology in the LIS curriculum, it is clear that the state of library education today is far different than it was in the early 1980s, and these changes are sure to have had an effect on the students enrolled in--or leaving--doctoral programs in the field.
After Bobinski's study was published, Whitbeck (1991) undertook another descriptive survey of LIS programs offering doctoral degrees. At the time of his research 1987 there were twenty programs offering research-based doctoral degrees, and Whitbeck surveyed nineteen of those. In many ways, this survey was similar to its predecessors, including questions on admissions criteria and curriculum components. Unlike Bobinski, Whitbeck did include questions about the time taken from admission to completion of program; however, because he received data only on graduates and current students (the assumption being that "current students" were those actually enrolled in classes during the time of the survey), it is difficult to tell how the results would have been altered had all students admitted during a given time period been surveyed. He draws no conclusions from this data regarding attrition, and provides little data that would be useful in understanding that aspect of the programs.
One problem with using these studies of doctoral programs to shed light on the problem of doctoral student attrition is that they have focused almost exclusively on characteristics of the programs themselves, as reported by the faculty and administrators of the programs, and not on the students themselves. Reeling (1992), however, does examine characteristics of doctorate recipients in LIS. Spurred by a concern that "there may be developing a shortage of individuals with the doctorate who will be available to teach in library and information science programs in this country" (pp. 311-312), Reeling set out to determine trends in the number of doctoral degrees granted in LIS, to develop a profile of LIS doctorate recipients, to identify the most popular fields of study for doctoral students in LIS, and to determine what factors doctoral program administrators believed would be most influential in recruitment and retention in the immediate future.
Though her interest included retention of students, most of Reeling's data came from surveys of individuals who had completed their doctorate, and from administrators may have been reluctant to discuss their programs' failure to retain doctoral students. Thus, Reeling's study was able to identify some characteristics of students who successfully navigated the doctoral path, but reveals little about the students who failed to complete their programs.
In short, while there have been some excellent studies focusing on the characteristics of LIS doctoral programs, doctoral dissertations, and Ph.D. recipients, there has been almost no attention paid to the unquestionably large number of students who enter LIS doctoral programs with every intention of completing their studies, but end up part of the growing population of doctoral dropouts.
The broader topic of doctoral student attrition and retention, however, has been studied to a greater extent in the fields of higher education and communication. Perhaps because of the difficulty in identifying and surveying students who have left their programs entirely, however, much of this literature focuses on time to completion of the degree rather than on attrition. While Abedi and Benkin (1987) complain that "the literature on graduate students, and specifically on the time it takes doctoral students to complete degrees, can charitably be described as sparse" (p. 4), these studies do provide a starting point for an exploration of the doctoral student experience in LIS.
Benkin (1984), in her study of doctoral students in all academic and professional fields at UCLA, found that the two factors that seemed to differentiate ABDs from doctoral recipients were their reports of financial assistance and their reported relationships with their faculty members. However, this study failed to differentiate among fields of study, and thus it is difficult to ascertain whether those factors are equally important to doctoral students in all fields.
Another researcher who surveyed students in multiple disciplines at a single institution is Bodian (1987), who found that "career instrumentality" (defined as the extent to which doctoral students perceive degree completion as instrumental to the attainment of future career goals) was negatively correlated with attrition intent, but that no relationship appeared to exist between attitude toward financial situation and either career instrumentality or attrition intent.
One of the most significant studies of doctoral student characteristics, including discussion of both time to completion and student attrition, is Bowen and Rudenstine's In Pursuit of the PhD (1992). Their in-depth analysis of graduate education in the US led them to assert that
only about half of all entering students in many PhD programs eventually obtain doctorates (frequently after pursuing degrees for anywhere from six to twelve years). In sharp contrast it is common for completion rates in leading professional schools of business, law, and medicine to exceed 90 percent. And it is not just the plight of the ABDs. . . that has caused completion rates in PhD programs to be low; attrition has been high at all stages of graduate study. Moreover, attrition appears to have increased over the last three decades. (p. 105)Bowen and Rudenstine identify three discrete stages of doctoral study-before the second year of study ("pre-2nd Year"), from the start of the second year until the completion of all requirements other than the dissertation ("pre-ABD"), and after completion of all requirements but the dissertation ("ABD"). They note that most research done to date on doctoral student attrition has focused exclusively on the ABD group. However, in their analysis of doctoral students entering Ph.D. programs at six universities (Berkeley, Chicago, Cornell, Princeton, Stanford, and the University of North Carolina), they found that "more than twice as many students left these Ph.D. programs prior to achieving ABD status as left after achieving ABD status." (p. 111)
Notably, Bowen and Rudenstine found that completion rates for Ph.D. students varied more based on field of study than on any other variable they considered. An important aspect of their findings on this aspect of attrition was that these differences among completion rates in fields of study existed even when the outcomes were controlled for factors commonly seen as important factors in attrition, such as the availability of financial aid and the gender of the student. In contrast, Hanson (1992) found significant evidence of difference in attrition rates between men and women in doctoral study; these differences may, however, be attributable to differences in time periods studied, as well as to institutions included in the research.
While Bowen and Rudenstine provide some valuable statistics regarding doctoral student time-to-completion and attrition, they acknowledge the difficulty in translating these statistics into an understanding of student motivations, noting that "it is impossible to invoke this simple dichotomy [between 'voluntary' and 'involuntary' attrition]." Their response to this difficulty in ascertaining motives, however, is not to suggest more detailed qualitative research, but rather to extract what they can from the "objective" statistical data. They assert that "it is much more sensible, in our view, to focus on when attrition occurs, which can be determined objectively."
Perhaps the most insightful comments on doctoral student attrition come from Vincent Tinto, whose well-respected work on collegiate student attrition, Leaving College (1993), outlines the development of his theory of student persistence. While most of Tinto's research has focused on undergraduate students, the work includes an appendix entitled "Toward a theory of doctoral persistence" in which he explores the possible extension of his model to the graduate student experience. Tinto also discusses the relationship of attrition rates to fields of study. Noting that in doctoral study the primary reference environment is the program or department rather than the institution, he asserts that "one can reasonably expect the process of doctoral persistence in some fields to be much more similar across institutions than it might be among some fields of study within a particular university."
Like Bowen and Rudenstine, Tinto identifies three stages of doctoral persistence: the first year of study, which he calls the transitional stage; the period leading to candidacy; and the completion of the dissertation. In both the first and second stages, the student's experience appears to be dependent on interactions with a wide range of faculty members; in the third stage, however, the focus shifts to the relationship with the advisor and a selected number of faculty making up the committee. (p. 237)
Tinto claims that "while finances matter, it is not clear that they matter in the same way at different stages of the doctoral completion process." While this does not contradict the findings of other researchers that finances play a key role in attrition, it certainly opens the door for more in-depth analysis of the range of reasons LIS doctoral students have for abandoning their course of study.
If Tinto is correct, it makes the need for research into doctoral attrition in LIS all the more necessary, since conclusions drawn from the research undertaken to date-all of which focuses on other fields, or on doctoral study across fields-could not be extended to LIS students.
Much of the research done on doctoral attrition during the past twenty years has focused on the available statistics relating to doctoral student characteristics and time to completion, with conclusions based on relationships between those sets of variables. In some cases, the data is supplemented by surveys sent to doctoral students, graduates, and occasionally drop-outs. However, the survey methodologies used, by nature of their structured format, often force the respondents to choose among factors predefined by the researchers, rather than allowing them to provide their own narrative explanations for their decisions.
In contrast, only a few researchers have taken a qualitative approach to studying the doctoral student attrition issue. Jacks, Chubin, Porter, and Connolly (1983), begin one of the few such studies stating that "we know little about [ABDs], their perceptions of the graduate training experience, and above all, their assessment of their 'failure' and subsequent career choices." Rather than approaching the issue from a statistical standpoint, they construct a "narrative portrait" (p. 74), interviewing twenty-five ABDs from eighteen departments at fifteen universities, and questioning them about their reasons for leaving their doctoral programs, the impact of that decision on their lives and careers, and their assessments of the value of the doctorate.
Like Bowen and Rudenstine (1992), Jacks et al. found that the greatest similarities among their respondents corresponded to fields of study, rather than age, gender, or other factors. However, they also identified financial pressure, a poor working relationship with the advisor and/or committee, substantive problems with the dissertation research, and personal or emotional problems as significant reasons cited by respondents for their departure from the programs.
Golde (1994) also used a qualitative approach to understanding doctoral student attrition, noting that the existing research on doctoral student attrition includes a number of critical gaps. In particular, she found that within the literature "the voices of students have been noticeably absent," expressing a concern that without the actual words of the students made available, it is impossible to "see the student's experience in its entirety." She also asserts that "because many of the investigations have been purely quantitative in nature, they have conceptualize [sic] attrition as a solitary event, rather than the consequence of a dynamic process." (1994, p.2)
In her doctoral thesis, Golde (1996) found that:
(1) The structure and timing of the various requirements for the doctorate affect student experiences, and consequently the attrition decision. (2) Many students become disillusioned in graduate school, because they had unrealistic expectations about graduate school and the academic life. (3) Relationships with advisors are critical, particularly in the sciences. Problematic relationships often stem from mismatched expectations and working styles. (4) The attrition decision was usually regarded as positive by those interviewed. Attrited students expressed few regrets about the decision to matriculate or to attrite. (5) Some women science students experienced graduate school differently than their male peers. (6) There are no consistent patterns in the reasons for attrition by attrition stage (early vs. late).While Golde's research provides an excellent starting point for better understanding of the doctoral student attrition problem, her research is focused on the interaction between the student and the department, and the identification of departmental factors, actions, and characteristics that influenced the students' choices. The possibility that the decisions were influenced in whole or in part by factors external to the academic setting is not fully explored. The voices of the students are an important aspect, but it is possible that these voices are not providing all of the information necessary to truly understand the complex decision-making process involved in leaving a doctoral program.
It is clear that while the topic of doctoral student attrition in LIS is an important one to the profession, little research has been done to shed light on this phenomenon. There is a pressing need to extend the research on persistence and attrition to the field of LIS, and to do so in a way that enables the identification of key variables in the attrition process.
Sense-making research involves selecting focal situations for study that present the subject with specific challenges, or "gaps" in understanding, and attempting to identify the problems or failures in an information system (which can be as broadly defined as a library or a school) that create barriers to information for the participants in that system. The sense-making methodology attempts to isolate what it is about an individual's experiences that leads them to choose a particular path through that system, and to discover how it is that individuals handle troublesome situations in their navigation of a system.
Dervin defines a troublesome situation as "any situation . . . where a person faces some kind of gap preventing movement ahead," and provides the following possible characterizations of such a situation:
Most importantly, sense-making does not assume that an outsider can tell by observing other people whether individually or through aggregate statistics how those people see themselves as blocked, or how they can be helped through their troublesome situations. Rather, it encourages the researcher to ask what Dervin calls "significant questions" about people's experiences, focusing on the blocks (what stops the person from accomplishing their goals), the questions (what information does the person seek), and the helps (what assistance would the person like to receive). These questions form the "sense-making triangle," allowing the researcher to circle the experiences of the individual and begin to make the difficult topic of human motivations approachable in a systematic way.
The basic sense-making research tool is a "time-line" interview, which asks the subject to reconstruct, step-by-step, a specific problem situation, focusing in turn on each of the sides of the sense-making triangle-the blocks encountered, the questions arising, and the helps sought. This interviewing process is highly detailed and time-consuming, but yields a richness of results that is difficult to match in any other form of interviewing. Because such interviewing is not always practical, however, the sense-making techniques have also been adapted to a number of other methods, including brief interviews, questionnaires, and document analysis.
Dervin addresses the concern that the focus on a relatively small number of individual experiences does not provide a sufficiently generalizable picture, saying:
Finally, sense-making assumes that people's stories, told in their own words, are as useful results of systematic research as counts of this and that. Statistics indicate the magnitude of achievements or problems-how many did this? How many failed to do that? The stories behind these counts make the numbers live and give them a reality that speaks eloquently. (Dervin & Clark, 1987, p. 36)As noted in the literature review, statistics have already been used effectively by many researchers to indicate the magnitude of the doctoral attrition problem, both in graduate study as a whole and in LIS doctoral study specifically. What remains to be discovered are the stories behind the LIS attrition numbers; the sense-making time-line interview method will be used to uncover those stories and provide a richer picture of the doctoral student experience as well as the doctoral program as an information system.
To be effective in studying the needs of a system's users, it is necessary to target specific problem situations for the user in structuring the interview. When sense-making research focuses on needs assessment, the interview focuses on what gaps were faced, questions were asked, and helps were sought, with an emphasis on the nature of the gaps encountered. In contrast, accountability research focuses on whether gaps were faced, questions were answered, and helps were sought, and places particular emphasis on the nature of the helps encountered and the role of the institution in providing those helps. In the context of this study, little is currently known about the nature of the "gaps" faced by students; as a result, a needs assessment approach will be used. Once these gaps have been identified, follow-up research can focus on the accountability aspects of the doctoral program.
In a response to a request for assistance in developing the instrument for this study that I posted to the sense-making researcher's mailing list, Dervin (personal communication, September 25, 1996) said:
To me, Liz's topic is a general importance far beyond LIS because, in fact, in all of the disciplines which touch on technologies many of the best doctorates drop out to go into industry. It is commonly thought to be a money issue but without much empirical evidence. The topic has even more generalizable importance as well because it deals with what I call path-changing -- moments of BIG gap-bridging when people change the roads they are on. It's important, I think, in studying such moments to free oneself as much as possible from pre-conceptions -- e.g. that money drives these changes, or that these changes are self-controlled, etc. --because these pre-conceptions lead one to "name" the world for respondents/informants.
My first reaction when I read Liz's topic is that she might consider not studying her topic directly because to do so catapults respondents into a world so constrained by expectations. Think, for example, of the difference between asking people about major times in their lives when they made path-changes, or perhaps times when they set out to do one thing and ended up doing another. Compare this with asking them what happened that led you to drop out of LIS?Based on that feedback, I developed a draft instrument that used two focal situations; the decision to attend graduate school, and a particularly difficult time or incident during graduate study. This allowed the indirect approach to the decision to leave (keeping in mind that the subjects would still be fully aware of the purpose and focus of my study), as well as making the instrument appropriate for expanding the study to students who had completed their programs if that appeared necessary during the course of the research.
After sending that draft instrument, and a question as to whether I should the specific "why you left graduate school" focal situation in the instrument, to the mailing list for comments and review, I received these additional comments from Dervin:
I've been pondering Liz's focal situation selection issue in general lately while writing up sense-making. Data collecting (including via observing) is always a problematic trade-off between ideal reach and procedural constraints (time, money, patience, memory). In Liz' case, the ends of the poles might be:
a) all important changings of life-paths versus
b) going for the specific set of material circumstances most pertinent to the issue at hand (e.g. the two focal situations of changing life paths to go into grad school and then out of grad school).
We always end up doing more of what is dictated by material constraint whether we are quantitative of qualitative. The issue is what set of ideal principles do we hang on to for methodological guidance.
In Liz' case, I expect her interview needs to be focused primarily on something like her two possible focal situations -- into and out of grad school. But Liz might use one or more means of reaching for the ideal, e.g.Based on Dervin's comments, and my review of additional sense-making studies and their associated instruments, I have constructed the following second draft of the survey instrument:
*Including a third focal situation -- the (or one) important life path change of respondent/informant's life
*Anchoring the two specific focal situations in a comparative set. Focusing on them in detail but always refer them back to the set. For example, starting with a life history of life path changes -- e.g. times when you were on one life path and changed, either because you wanted to or because of circumstances. Put these on a life time-line by year. Then add getting into and leaving grad school onto the time line. Then add to the in-depth questions about the grad school transitions some questions pursuing what makes this like/not like other life changes. (personal communication, January 5, 1997)
[write out timeline, anchoring events by year]
Okay, now that we have that timeline in place, I'd like to ask you some questions about events related to your doctoral study at [institution name here].
Situation #1
I want to begin by asking you some questions about the events in your life leading up to your decision to enroll in a doctoral program of study, and your experiences in that program. As we proceed, you may have some particularly strong memories, and I'd like to hear all of them. I have a questionnaire structure which guides how and when I ask you about different parts of your memories. By using this structure, we can compare your experiences with others', while still allowing you to recall your particular experiences.
As we go through this process, there may be times when you get a lot of ideas bottled up in your mind and you really want to say them all at once. If this happens, just let me know and we'll take time out for me to just listen to you, and them we'll return to the questionnaire structure afterwards.
To help both of us in this process, I would like, with your permission, to tape record the interview. This is only so I won't have to take as many notes while you are talking. I will use the tape only to help us reconstruct our interview, afterwards it will be erased. At no time will your name or other identifying characteristics be attached to the tape. Would this be ok with you?
What I'd like for you to do now is to choose an important, difficult, troublesome, or challenging situation that occurred in the past that related to your decision to begin a doctoral program in LIS. This situation could be a discussion with a family member, friend, or colleague; visiting a campus; reading an article or book; or simply an occasion when you were particularly focused on the idea of doctoral study. I'll give you a minute to get that situation in mind.
Okay, let's focus now on the situation you have in mind. Can you describe it for me?
Record response on green card--number 1.
When did this situation occur? Let's anchor it on the timeline.
[add Situation #1 to timeline]
When we face the challenges associated with path-changing situations, we see different things as making the situation difficult--sometimes we blame ourselves, people close to us, authority figures, bureaucracies, institutions, society, etc. When you had this situation in your mind [review situation] did you see yourself as blocked or hindered in some way?
___no ____yes
If yes, how did you see yourself blocked? If no, what made this situation particularly stand out in your mind?
Record response on blue card--number 1a
I'm also interested in knowing whether you saw this situation as being connected to or related to other important situations in your life at the time.
___no ____yes
If yes, what was that situation?
Record response on yellow card--number 1b (add to timeline if it does not appear already)
Did you see this situation as substantially different from other path-changing situations you've encountered in your life?
___no ____yes
If so, in what ways?
Record response on yellow card--number 1c1 (add to timeline if it does not appear already)
Did you see this situation as substantially similar to other path-changing situations you've encountered in your life?
___no ____yes
If so, in what ways?
Record response on yellow card--number 1c2 (add to timeline if it does not appear already)
Now I'd like you to go back in your mind and try to identify what the questions you had in your mind at the time were. By questions, I mean things that you wanted to find out, learn about, come to understand, unconfuse, or make sense out of. You need not have asked the question out loud, nor found an answer; we simply want to identify gaps in understanding that you faced at the time. These may not have been in your mind as questions, but rather as unclear aspects of the situation or your feelings. In these cases, what I need you to do is to translate that aspect into a question, or to simply talk about that aspect so that together we can translate it into a question.
Here's an example: Suppose I'm in a grocery store, and have just wheeled my cart into the produce section. That is my situation. My questions might be: Where are the avocados? How can that man wear purple pants in public? Why is that man spraying the lettuce? Are mushrooms still $3.00 a pound? I wonder if the corn is as good as it looks? Etc.Now let's look at your situation [review situation]. Think back, what questions did you have in your mind at this point in time?
Lay cards out, leading down from the timeline event.
Now that we have these aspects of your situation laid out, what I want you to do is to think about the events and questions to see if you'd like to add anything.
If time allows, begin in-depth analysis of questions now. (see question analysis section below). Otherwise, write out the timeline and questions for the subject on response forms, and give it to them to take with them. Explain that I'll be contacting them via telephone and/or electronic mail to analyze those questions in more detail.
Situation #2
Okay, let's move on now. What I'd like for you to do this time is to choose a situation that was particularly important, challenging or problematic during your doctoral program. The situation could be an interaction with a colleague or professor; a difficult decision you needed to make regarding your course of study; or, again, simply an occasion when you were particularly focused on the idea of doctoral study. I'll give you a minute to get that situation in mind.
Okay, let's focus now on the situation you have in mind. Can you describe it for me?
Record response on green card--number 2.
When did this situation occur? Let's anchor it on the timeline.
[add Situation #2 to timeline]
Okay, let's focus now on the details of this situation. When you had this event in your mind, did you see yourself as blocked or hindered in some way?
___no ____yes
If yes, how did you see yourself blocked? If no, what made this situation particularly stand out in your mind?
Record response on blue card--number 2a
I'm also interested in knowing whether you saw [read situation] as being connected to or related to other important situations in your life at the time.
___no ____yes
If yes, what was that situation?
Record response on yellow card--number 2b (add to timeline if it does not appear already)
Did you see this situation as substantially different from other path-changing situations you've encountered in your life?
___no ____yes
If so, in what ways?
Record response on yellow card--number 2c1 (add to timeline if it does not appear already)
Did you see this situation as substantially similar to other path-changing situations you've encountered in your life?
___no ____yes
If so, in what ways?
Record response on yellow card--number 2c2 (add to timeline if it does not appear already)
Now let's look at your situation [review situation]. Think back, what questions did you have in your mind at this point in time?
Record questions on white cards; number questions 2d1-2dnn.
Lay cards out, leading down from the timeline event.
Now that we have these verbal pictures or photographs of your situation laid out, what I want you to do is to think about the events and questions to see if you'd like to add anything.
If time allows, begin in-depth analysis of questions now. (see question analysis section below). Otherwise, write out the timeline and questions for the subject on response forms, and give it to them to take with them. Explain that I'll be contacting them via telephone and/or electronic mail to analyze those questions in more detail.
Situation #3
Okay, let's move on to the final situation now. What I'd like for you to do this time is to choose a situation that was particularly important, challenging or problematic related to your decision to leave your doctoral program. Again, the situation could be an interaction with a colleague or professor; a difficult decision you needed to make regarding your course of study; or, again, simply an occasion when you were particularly focused on the idea of doctoral study. I'll give you a minute to get that situation in mind.
Okay, let's focus now on the situation you have in mind. Can you describe it for me?
Record response on green card--number 3.
When did this situation occur? Let's anchor it on the timeline.
[add Situation #3 to timeline]
Okay, let's focus now on [read situation]. When you had this event in your mind, did you see yourself as blocked or hindered in some way?
___no ____yes
If yes, how did you see yourself blocked? If no, what made this situation particularly stand out in your mind?
Record response on blue card--number 3a
I'm also interested in knowing whether you saw [read situation] as being connected to or related to other important situations in your life at the time.
___no ____yes
If yes, what was that situation?
Record response on yellow card--number 3b (add to timeline if it does not appear already)
Did you see this situation as substantially different from other path-changing situations you've encountered in your life?
___no ____yes
If so, in what ways?
Record response on yellow card--number 3c1 (add to timeline if it does not appear already)
Did you see this situation as substantially similar to other path-changing situations you've encountered in your life?
___no ____yes
If so, in what ways?
Record response on yellow card--number 3c2 (add to timeline if it does not appear already)
Now let's look at your situation [review situation]. Think back, what questions did you have in your mind at this point in time?
Record questions on white cards; number questions 3d1-3dnn.
Lay cards out, leading down from the timeline event.
Now that we have these verbal pictures or photographs of your situation laid out, what I want you to do is to think about the events and questions to see if you'd like to add anything.
If time allows, begin in-depth analysis of questions now. (see question analysis section below). Otherwise, write out the timeline and questions for the subject on response forms, and give it to them to take with them. Explain that I'll be contacting them via telephone and/or electronic mail to analyze those questions in more detail.
Let's start with this first question [read question]:
I'd like to know what you were trying to do by asking this question.
Example: For example, in the grocery store example that we used before, when I entered the produce department and asked if the mushrooms were still $3.00 per pound what I might have been trying to do was decide whether or not to get mushroom for myself because I like them, but I was worried about paying too much. I also might have been curious because even though I don't like mushrooms, it is interesting to me that some people will pay up to $3.00 per pound for them, like maybe that guy in the purple pants.When you had this question in your mind [review question] did you see yourself as blocked or hindered in some way?So, think back to when you asked [review question] in your mind. What was it that you were trying to do by asking this question? It's okay to take time to think.
Were you able to find an answer to this question? ___yes ___no
If yes:
Was it a complete or a partial answer?If no:
How did you get the answer?
How did the answer help you?
What do you see as having prevented you from getting an answer?
How do you think the answer could have helped you?
For this study, situations will be analyzed for their movement state ("the way in which the person sees his/her movement through time space being blocked,"), as well as the situation embeddedness ("the extent to which the person sees the situation as related to other situations"). (Dervin, 1983, p.60)
Gaps, the most important aspect of the triangle for the purposes of this study, will be analyzed for valence focus (whether the gaps involved a good, bad, and/or neutral path), ease of answering, importance of answering, source of answers, and answering success. In addition, the responses themselves will be used to identify descriptive categories for analysis.
Uses, or helps, will be categorized based on information gathered in the interview practice. I anticipate that categories will include several that have been encountered and utilized in other sense-making studies, including:
All interviews will be analyzed using each of the measures described above, and compared for consistencies and differences among themselves, as well as for their support or refutation of conclusions drawn by other researchers into doctoral attrition.
I expect to submit a first draft of the dissertation to the committee in August or September of 1997, with the intention of making revisions and submitting a final version in preparation for a December defense.
Abedi, J., & Benkin, E. (1987). The effects of students' academic, financial, and demographic variables on time to the doctorate. Research in Higher Education, 27(1), 3-14.
Abrera, J. B. (1987). Doctoral programs, dissertations, and graduates: Characteristics of the literature, 1926-80. Journal of Education for Library and Information Science, 28(1), 38-51.
Abrera, J. B. (1988). Doctoral programs, theses, and graduates in library and information science in the United States: An analysis of the published literature, 1960-1980. Occasional Papers Number 183. Urbana. Graduate School of Library and Information Science.: Illinois Univ.
Babbie, E. (1989). The practice of social research (Fifth ed.). Belmont, CA: Wadsworth.
Benkin, E. M. (1984). Where have all the doctoral students gone?: A study of doctoral student attrition at UCLA. (Doctoral dissertation, University of California, Los Angeles).
Bobinski, G. S. (1986). Doctoral programs in library and information science in the United States and Canada. Library Trends, 34(4), 697-714.
Bodian, L. H. (1987). Career instrumentality of degree completion as a factor in doctoral student attrition. (Doctoral dissertation, University of Maryland).
Bowen, W. G., & Rudenstine, N. L. (1992). In pursuit of the PhD. Princeton, NJ: Princeton University Press.
Buckland, M. (1991a). Information and Information Systems. New York: Praeger.
Buckland, M. K. (1991b). Information as thing. Journal of the American Society for Information Science, 42(5), 351-360.
Cole, C. (1994). Operationalizing the notion of information as a subjective construct. Journal of the American Society for Information Science, 45(7), 465-475.
Danton, J. P. (1959). Doctoral study in librarianship in the United States. College and Research Libraries, 20(Nov.), 435-453.
Davis, C. H. (1990). On qualitative research. Library and Information Science Research, 12, 327-328.
Davis, C. H. (1991). Response to Guba's letter. Library and Information Science Research, 13, 87-88.
Dervin, B. (1976). Strategies for dealing with human information needs: Information or communication. Journal of Broadcasting, 20(3), 324-333.
Dervin, B. (1977). Useful theory for librarianship: Communication, not information. Drexel Library Quarterly, 13(3), 16-32.
Dervin, B. (1983, May). An overview of sense-making research: Concepts, methods, and results to date. Paper presented at the International Communication Association annual meeting, Dallas, TX.
Dervin, B. (1991). Comparative theory reconceptualized: From entities and states to processes and dynamics. Communication Theory, 1(1), 59-69.
Dervin, B. (1992a). From the mind's eye of the 'user': The sense-making qualitative-quantitative methodology. In J. D. Glazier & R. R. Powell (Eds.), Qualitative research in information management (pp. 61-84). Englewood, CO: Libraries Unlimited.
Dervin, B. (1992b, October). Information as non-sense; information as sense: The communication technology connection. Paper presented at the American Society for Information Science, Pittsburgh, PA.
Dervin, B. (1994.) Reliability and validity in sense-making. October 1994.
Dervin, B., & Clark, K. (1987). ASQ: Asking signficant questions. Alternative tools for information need and accountability assessments (ERIC Document Reproduction Service No. ED 286 519). Peninsula Library System.
Dervin, B., & Nilan, M. (1986). Information needs and uses. In M. E. Williams (Ed.), Annual Review of Information Science and Technology (pp. 3-33).
Fidel, R. (1992). The case study method: A case study. In J. D. Glazier & R. R. Powell (Eds.), Qualitative research in information management (pp. 37-60). Englewood, CO: Libraries Unlimited.
Futas, E., & Zipkowitz, F. (1991). The faculty vanishes. Library Journal(September 1), 148-152.
Germeroth, D. (1990, November). Lonely days and lonely nights: Completing the doctoral dissertation. Paper presented at the Annual Meeting of the Speech Communication Association, Chicago, IL.
Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for qualitative research. New York: Aldine.
Glazier, J. D., & Powell, R. R. (1992). Qualitative research in information management. Englewood, CO: Libraries Unlimited.
Golde, C. M. (1994, November). Student descriptions of the doctoral student attrition process. Paper presented at the 19th Annual Meeting of the Association for the Study of Higher Education, Tucson, AZ.
Golde, C. M. (1996). How departmental contextual factors shape doctoral student attrition. (Doctoral dissertation, Stanford University).
Guba, E. G. (1991). A letter to the editor. Library and Information Science Research, 13, 85-86.
Hale, M. L. (1991). Paradigmatic shift in library and information science. In C. R. McClure & P. Hernon (Eds.), Library and information science research: Perspectives and strategies for improvement (pp. 336-346). Norwood, NJ: Ablex.
Hanson, T. L. (1992, October). The ABD phenomenon: The "at risk" population in higher education and the discipline of communication. Paper presented at the 78th Annual Meeting of the Speech Communication Assocation, Chicago, IL.
Hawley, P. (1993). Being bright is not enough: The unwritten rules of doctoral study. Springfield, IL: Charles C. Thomas.
Houser, L., & Schrader, A. M. (1978). The search for a scientific profession: Library science education in the U.S. and Canada. Metuchen, N.J.: The Scarecrow Press.
Houser, L. J. (1982). The PhD dissertation in library science. Library Research, 4, 95-107.
Jacks, P., Chubin, D. E., Porter, A. L., & Connolly, T. (1983). The ABCs of ABDs: A study of incomplete doctorates. Improving College and University Teaching, 31(2), 74-81.
Katzer, J., Cook, K. H., & Crouch, W. W. (1991). Evaluating information: A guide for users of social science research (Third ed.). New York: McGraw-Hill.
Kuhlthau, C. C. (1991). Inside the search process: Information seeking from the user's perspective. Journal of the American Society for Informaton Science, 42(5), 361-371.
Marco, G. (1967). Doctoral programs in American library schools. Journal of Education for Librarianship, 8(Summer), 6-13.
Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: An expanded sourcebook (2nd ed.). Thousand Oaks, CA: Sage.
Miller, D. C. (1991). Handbook of research design and social measurement (Fifth ed.). Newbury Park, CA: Sage.
Paris, M. (1988a). Library school closings: Four case studies. Metuchen, NJ: Scarecrow Press.
Paris, M. (1988b). Thoughts on the case study. Journal of Education for Library and Information Science, 29, 137-140.
Powell, R. R. (1985). Basic research methods for librarians (First ed.). Norwood, NJ: Ablex.
Powell, R. R. (1995). Research competence for Ph.D. students in library and information science. Journal of Education for Library and Information Science, 36(4), 319-329.
Reeling, P. G. (1992). Doctorate recipients in library science: How they compare with doctorate recipients in other disciplines. Journal of Education for Library and Information Science, 33(4), 311-329.
Schement, J. R., & Ruben, B. D. (Eds.). (1993). Information and behavior: Vol. 4. Between communication and information. New Brunswick, NJ: Transaction Books.
Shera, J. H. (1979). Education for librarianship-an assessment and a perspective: A review article. Library Quarterly, 49(3), 310-316.
Sternberg, D. (1981). How to complete and survive a doctoral dissertation. New York: St. Martin's Press.
Tinto, V. (1993). Leaving college: Rethinking the causes and cures of student attrition (2nd ed.). Chicago: The University of Chicago Press.
Whitbeck, G. W. (1991). Doctoral programs in library and information science: A descriptive study. Journal of Education for Library and Information Science, 31(4), 314-328.