7
http://top.sagepub.com/ Teaching of Psychology http://top.sagepub.com/content/15/4/186 The online version of this article can be found at: DOI: 10.1207/s15328023top1504_2 1988 15: 186 Teaching of Psychology Maria Emma Garcia, Richard W. Malott and Dale Brethower A System of Thesis and Dissertation Supervision: Helping Graduate Students Succeed Published by: http://www.sagepublications.com On behalf of: Society for the Teaching of Psychology can be found at: Teaching of Psychology Additional services and information for http://top.sagepub.com/cgi/alerts Email Alerts: http://top.sagepub.com/subscriptions Subscriptions: http://www.sagepub.com/journalsReprints.nav Reprints: http://www.sagepub.com/journalsPermissions.nav Permissions: What is This? - Dec 1, 1988 Version of Record >> at MEMORIAL UNIV OF NEWFOUNDLAND on November 25, 2014 top.sagepub.com Downloaded from at MEMORIAL UNIV OF NEWFOUNDLAND on November 25, 2014 top.sagepub.com Downloaded from

A System of Thesis and Dissertation Supervision: Helping Graduate Students Succeed

  • Upload
    dale

  • View
    214

  • Download
    1

Embed Size (px)

Citation preview

Page 1: A System of Thesis and Dissertation Supervision: Helping Graduate Students Succeed

http://top.sagepub.com/Teaching of Psychology

http://top.sagepub.com/content/15/4/186The online version of this article can be found at:

 DOI: 10.1207/s15328023top1504_2

1988 15: 186Teaching of PsychologyMaria Emma Garcia, Richard W. Malott and Dale Brethower

A System of Thesis and Dissertation Supervision: Helping Graduate Students Succeed  

Published by:

http://www.sagepublications.com

On behalf of: 

  Society for the Teaching of Psychology

can be found at:Teaching of PsychologyAdditional services and information for    

  http://top.sagepub.com/cgi/alertsEmail Alerts:

 

http://top.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

What is This? 

- Dec 1, 1988Version of Record >>

at MEMORIAL UNIV OF NEWFOUNDLAND on November 25, 2014top.sagepub.comDownloaded from at MEMORIAL UNIV OF NEWFOUNDLAND on November 25, 2014top.sagepub.comDownloaded from

Page 2: A System of Thesis and Dissertation Supervision: Helping Graduate Students Succeed

McKeachie, W. J. (1986). Teaching tips: A guidebook for beg~nning Notes college teachers (8th ed.). Lexington, MA: Heath.

National Commission on Excellence in Education. (1983). A 1. 1 am grateful for the helpful comments of Charlc, L. Brewer natlon at risk: The imperative for educational reform. Washington, and an anonymous reviewer. DC: Author. 2. Requests for reprints should he sent to Diane 1-. Halper~i,

National Governors' Association. (1986). Time for results: The Department of Psychology, California State Un~\,ersity, 5500 governors' 1991 report on education. Washington, DC: Author. University Parkway, San Bernardino, CA 92407.

A System of Thesis and Dissertation Supervision: Helping Graduate Students Succeed

Maria Emma Garcia Richard W. Malott Dale Brethower Western Michigan University

Twenty-nine psychology graduate students participated in a thesis and dissertation supervisor^ system. The system included weekly meetings, task specification, feedback, and incentiues. Regular participants completed significantly more tasks when academic credit depended on task completion than when it did not. In comparison with a college-wide control group, a higher percentage of the participants in the supervisor?, group defined the topic, read relevant articles, collected data, wrote a draft, and completed their projects. In comparison with a psychology control group, the graduating participants completed M A theses of similar qtlality.

Of all students entering graduate programs, as many as one half drop out before obtaining their M A or PhD degrees (Berelson, 1960; Cartter, 1965; Heiss, 1970; Jacks, Chubin, Porter., & Connolly, 1983; Knox, 1970; Lun- neborg & Lunneborg, 1973; Naylor &Sanford, 1982; Sells, 1975; Von Zur-Muehlen, 1977; Wright, 1964). Of those who drop out, at least 25% do so after completing their courses and before finishing their theses or dissertations (Berelson, 1960; Elfatouri, Garcia, & Malott, 1988; Garcia & Malott, 1988; Moore, 1985).

Students who graduate take longer than the time speci- fied by M A and PhD programs- 1 to 2 years for the M A degree and 4 to 5 years for the PhD degree ("Graduate Deans," 1988; Von Zur-Muehlen, 1977). They take 8 to 9 months longer than expected to get the M A degree (Dillon & Malott, 1981; Garcia & Malott, 1988; Snell, 1965), 12 to 24 months longer to get the PhD after entering with the BA, and 15 months longer to get the PhD after entering with the M A (Berelson, 1960; Garcia & Malott, 1988; Wilson, 1965).

Thesis and dissertation requirements increase attrition and delay completion of graduate degrees. Students often fail to complete theses and dissertations or take longer than expected- 1 year for the M A thesis and 2 years for the PhD dissertation (Berelson, 1960; Garcia & Malott, 1988; U.S. Department of Labor, 1986). Several studies illdicate that students fail to complete their theses and dissertations or take longer than expected because of lack of supervision and effective incentives (Berelson, 1960; Rlanton, 1981; Dillon, Kent, & Malott, 1980; Dillon & Malott, 1981; Grant, Dillon, & Malott, 1980; Garcia & Malott, 1988; Malott, 1972, 1986; Wilson, 1965; Zoia, 1981). T o address this problem, some investigators use supervisory systems involving detailed task specification, frequent deadlines, weekly guidance and feedback, and incentives consisting of letters of recommendation. Dillon et al. (1980) found that letters of recommendation combined with weekly feedback increased task completion. Dillon and Malott (1981) found that students com~leted 91% of their tasks when thev received instructions, feedback, and letters of rccommen dation. Compared with a control group, students in Dillon and Malott's system had lower attrition rates and higher graduation rates, graduated in shorter times, allJ produced theses of equal quality.

Our original purpose was to replicate Dillon and Malott'b (1981) study using ,t larger control group j ~ n c l u d ~ n ~ nonpsychology students). a qual~t) evaluator ~ndependent from the students' committee members, and ( ontracts that zpec~fied required tasks In early ctagcs of o u ~ researcli. however, the average t a d iompletlon was lower than th t average task complet~on In Dlllon and Malott's \tudv, consequently, we added othcr Inc t ntl\e\

186 Teaching of Psychology at MEMORIAL UNIV OF NEWFOUNDLAND on November 25, 2014top.sagepub.comDownloaded from

Page 3: A System of Thesis and Dissertation Supervision: Helping Graduate Students Succeed

Method points per task. These contracts had blank spaces for the

This study involved within-group and between-group comparisons. In the within-group comparisons, we exam- ined the effect of several incentive conditions on task completion of students in a supervisory group. In the between-group comparisons, we compared the perform- ance of students in the supervisory groups with the per- formance of students in two control groups. We used a control group from the College of Arts and Sciences (the college-wide group) to compare thesis progress and amount of supervision. The other control group was from the Department of Psychology (the psychology group); we used it to compare thesis quality. There was no overlap between the students in the supervisory group and the students in the two control groups.

Subjects and Setting

The supervisory group consisted of 29 graduate students from the Department of Psychology at Western Michigan University (4 students preparing M A projects, 18 preparing MA theses, and 7 preparing PhD dissertations). Richard W. Malott, the second author, required participation of all his students. In addition, 7 students of other advisers partici- pated.

The college-wide control group consisted of 53 of the 55 students who answered a survey mailed to 930 graduate students. (Two students answered the survey but chose not to participate.) We randomly selected 930 students from a list of arts and sciences students who had taken at least one course in the preceding year. The sample consisted of 46 MA students and 7 PhD students. Sixteen students in the college-wide group were from the Department of Psycholo- gy. The psychology control group consisted of all 22 psy- chology students who submitted their theses, specialist projects, or dissertations to the Graduate College in the last year of this study. Students participated in this research with their written consent.

It is possible that slight overlap existed between psychol- ogy students who participated in the college-wide control group and in the psychology control group. But even a 100% overlap of students between the two control groups would not affect the validity of their separate functions; the college-wide control group dealt with rate of progress and amount of supervision, and the psychology control group dealt with thesis quality.

Within-Group Comparisons

Standard Procedures

At the beginning of each semester, each student received performance contracts, instructions, and a schedule of supervisory and advisory meetings. In addition, every week during the semester, each student received feedback about progress.

Contracts. The basic data collection forms consisted of 39 weekly contracts (Garcia, 1987). The contracts specified standard tasks, proof of accomplishment, and possible

student to record additional tasks, points earned, and time invested per task. Each student received a set of 13 weekly contracts every semester. Each set corresponded to one of three research phases (generating, implementing, and writ- ing).

lnstr~ctions. Written instructions specified general procedures of the system. In addition, instructions included guidelines for using contracts, finding relevant articles, and preparing research presentations (Garcia, 1987). In the first week of each semester, Maria Emma Garcia, the first author, reviewed the instructions with the students. Then they adapted the contracts to their specific needs by re- scheduling, eliminating, or adding tasks. Finally, students took a written test on the instructions. Those who missed more than one question took a parallel test until showing mastery.

Supemiso y and adviso y meetings. Each PhD student received 16 hr of supervision a semester; 1 hr a week for 13 weeks from the adviser (Malott and ~rethower, second and third authors), and 1 hr per semester from each of the other three members of his or her PhD committee. Each MA student received 18 hr of supervision a semester: 1 hr a week for 13 weeks from a PhD student or an advanced MA student, ?h hr every other week from the adviser (second and third authors and one other faculty member), and 1 hr per semester from each of the other two members of her or his MA committee. In the weekly meeting, the supervisor reviewed the contract of the week, checked proofs of task completion, verified points earned, assigned new tasks, and discussed research problems. In the biweekly and semestral meetings, students only discussed research issues. All were individual meetings.

Feedback on progress. Students received a feedback form every week. Feedback included the weekly and cumu- lative percentage of contracted tasks completed by each student and by the group.

Incentive Conditions

This study involved four incentive conditions: (a) letter of recommendation alone, (b) letter of recommendation and a semester credit contingency, (c) letter of recommen- dation and a biweekly credit contingency, and (d) no incentive.

Letter of recommendation alone. The primary adviser (second author) wrote a letter of recommendation when- ever a student requested it. The letter included the percent- age of contracted tasks completed and rank of the student's performance in this condition.

Letter of recommendation and semester credit contingen- cy. Under this condition, we also used the letter of rec- ommendation procedure. In addition, students received no research credit if their cumulative percentage of contracted tasks completed was less than 85% by the end of the semester. To continue in the system, students had to register and pay for the lost credits the next semester. However, the semester credit contingency did not affect

Vol. 15, No. 4, December 1988 187 at MEMORIAL UNIV OF NEWFOUNDLAND on November 25, 2014top.sagepub.comDownloaded from

Page 4: A System of Thesis and Dissertation Supervision: Helping Graduate Students Succeed

GPAs or enrollment in graduate school. (Students got an Incomplete for research credits if their cumulative percent- age of contracted tasks completed was 85% or higher and they had not completed their theses or dissertations by the end of the semester. Academic credits replaced the Incompletes when students finished their theses or disser- tations and had them accepted by their committee and the Graduate College.)

Students conducting MA projects (rather than theses) got a failing grade if their cumulative percentage of con- tracted tasks completed was less than 85% by the end of the semester. To continue in the system, these students regis- tered for extra credits the next semester. The failing grade affected cumulative GPAs but not enrollment in graduate school.

Letter of recommendation and biweekly credit contingen- cy. Under the biweekly credit contingency, students earned credit only if their cumulative percentage of con- tracted tasks completed never fell below 85% for 2 weeks in a row. A student who failed the biweekly requirement could not participate in the system for the rest of that semester. As in the semester contingency, such a student also received no thesis or dissertation credit or received a failing grade for the master's project. To continue in the system, a student who had failed registered for the lost credits the next semester. We developed the semester and biweekly contingencies in collaboration with the reigstrar's office and with the Graduate College.

No incentive. Performance in this condition was not included in any letter of recommendation and did not affect academic credits, letter grades, or participation in the system.

Dependent Variables

The percentage of contracted tasks completed per week was the main dependent variable. Tasks consisted of at- tending meetings, writing, reading, collecting and graphing data, presenting the thesis in a seminar, and doing nonre- curring tasks, such as completing a human subjects form. Students performed each nonrecurring task once or twice during their study (Fulton & Malott, 1982). Students

earned 1 point for 1 hr of work, contingent on proof of task completion. Notes or signatures were proof of attendance at meetings; literature review forms were proof of reading h pages of an article or 20 of a book; tables and graphs were proof of data processing. Students and their supervisors specified in advance the evidence for completing these tasks.

We arranged weekly tasks so that, in the generating phase, students defined the topic, reviewed relevant litera- ture, conducted pilot studies, and wrote the research proposal. In the implementing phase, students recruited subjects, collected and graphed data, and measured accu- racy and reliability of data. In the writing phase, they wrote the final draft, had the oral examination, and submitted the thesis to the Graduate College. At the end of each - semester, each supervisee and each supervisor in the system answered an anonymous social-validation questionnaire to evaluate all aspects of the system.

Experimental Design

Different students received different combinations of incentive conditions depending on their performance and when they participated in the system. We collected data for three 13-week semesters (see Table 1).

In Semester 1, students participated in the letter condi- tion for 13 weeks, but the mean percentage of contracted tasks completed was lower than in Dillon and Malott's (1981) system. We assumed that the letter of recommenda- tion was not a sufficient incentive. Therefore, we also made academic credit contingent on task completion (the letter and semester contingency condition), for the first 5 weeks of Semester 2. Then, for comparison, we returned to the letter condition for the last 8 weeks of Semester 2. However, students who completed less than 85% of their tasks during the first 5 weeks continued with the letter and semester contingency until achieving the passing criterion.

Even with the added incentive of academic credit, some students failed to achieve the 85% passing criterion by the end of Semester 2. We assumed that the semester deadline allowed for too much procrastination. Therefore, we intro- duced and evaluated the letter and biweekly contingency in Semester 3. This condition was in effect for the first 4 weeks

Table 1. Number of Students Exposed to Different Experimental Phases

Conditions

Letter & Letter & No Letter & Letter Semester Letter Biweekly Incentive Biweekly

n (13) (5) (8) (4) (4) (5)

4 X X X X X X

1 X X X X X

2 X X X X X

1 X X X 4 X X X

2 X X

4 X X X X X

1 X X

1 X X X

6 x

Note. Three students involved in the supervisory system for less than one semester were not included. Numbers in parentheses indicate the number of weeks of each condition.

Teaching of Psychology at MEMORIAL UNIV OF NEWFOUNDLAND on November 25, 2014top.sagepub.comDownloaded from

Page 5: A System of Thesis and Dissertation Supervision: Helping Graduate Students Succeed

of Semester 3. Then, for comparison, we introduced 4 weeks of the no-incentive condition, after which we re- turned to 5 weeks of the letter and biweekly contingency.

Students in standard circumstances worked on their research and other academic requirements at the same time. (Standard circumstances involved conditions charac- teristic of most students in the system, not necessarily of graduate students in general.) Students in special circum- stances had exceptional matters interfering with work on their research. They had serious personal problems (e.g., extended illness) or were in at least two of the following conditions: finished coursework before the research (thereby decreased contact with the university), moved out of town, worked full-time in nonacademic jobs, or studied in graduate programs in other universities. (Five of the six students who ended in special circumstances began the system in standard circumstances.)

Between-Group Procedures

At the end of Semester 1, we mailed a survey and an informed consent form to each student in the college-wide group and in the supervisory group. For 4 months until the beginning of Semester 2, surveys were returned and tallied. The survey included questions about academic history (e.g., entry date), adviser's supervision (e.g., frequency of advisory meetings), and thesis progress (e.g., data col- lected).

We compared survey data on academic history with university records and found them to be consistent. Thesis progress and adviser's supervision were based only on survey data. In addition, we used academic information based only on university records (i.e., cumulative GPAs and credits taken per semester).

During the three semesters of this study, the Dean of the Graduate College completed an evaluation form after stu- dents from the supervisory group and the psychology group submitted their research to the Graduate College. With no knowledge of students' group membership, the Dean ranked the items on a scale ranging from lowest (1) to highest (4).

Results and Discussion

Within-Group Results

In all incentive conditions, students under standard circumstances completed a higher percentage of contracted tasks per week than did students under special circum- stances (see Table 2). Students in standard circumstances

completed more tasks under credit contingencies (1 semes- - ter contingency and 2 biweekly contingencies) than under no-credit contingencies (2 letter conditions and 1 no- incentive condition). We found a significant difference, t(14) = 4.06, p < .O1 (one-tailed), between the mean percentage of contracted tasks completed under the credit contingencies (M = 102.3) and under the no-credit contin- gencies (M = 76.9). (Statistical analysis is based on the average performance during applicable conditions of the 15 students in standard circumstances who were exposed to both credit and no-credit contingencies, including 6 stu- dents who withdrew after participating one or more semes- ters in the supervisory system.)

Students in standard circumstances completed fewer tasks during the first letter condition (63%) than during the second letter condition (89%). Student interviews and wer- . . formance data suggest that, during the first letter condi- tion, most students did not understand instructions about the point system. Instructions were that they could work more hours than swecified in the contracts and thus earn more than 100% of the points contracted for in a given week. Performance in the second letter condition was comparable to that obtained in Dillon and Malott's (1981) study (91%). However, even ignoring the first letter condi- tion (with its potential confounding of misunderstood instructions), a significant difference holds between the - percentage of contracted tasks completed under the re- maining no-credit contingencies (M = 88.0) and the credit contingencies (M = 102.3)) t(14) = 2.09, p < .05 (one- tailed).

Students in special circumstances comwleted 50% of the tasks when academic credit depended on task completion and 8% when it did not. These students achieved their highest percentage of contracted tasks completed (76%) during the phases with the contingency allowing the least procrastination, the biweekly contingency. By contrast, they completed only 31% of their tasks during the letter and semester-contingency condition, 26% during the letter- only condition, and 1% during the no-incentive condition (see Table 2). The small number of students in special circumstances preclude statistical tests.

All students liked all features of the system, according to their anonymous answers to the social validation question- naire. Because we used this questionnaire at the end of each semester, some students evaluated the system more than once. We received 94% of all the questionnaires distributed in the three semesters combined. A high percentage of respondents evaluated each aspect of the system with a modal ranking of 1 or 2 on a scale ranging from highest (1) to lowest (4): contribution of the research supervisor (98%), research presentations (94%), contribution of the adviser

Table 2. Mean Percentages of Task Completion by Phase

Phase

Letter & Letter & No Letter & Letter Semester Letter Biweekly Incentive Biweekly

Standard circumstances 63 101 89 97 86 108 ( n = 1 9 ) ( n = 1 4 ) ( n = 1 4 ) ( n = 9 ) ( n = 9 ) ( n = 9 )

Special circumstances 26 3 1 - 76 1 69 ( n = 1) ( n = 5) ( n = 4 ) ( n = 3 ) ( n = 2 )

Vol. 15, No. 4, December 1988 at MEMORIAL UNIV OF NEWFOUNDLAND on November 25, 2014top.sagepub.comDownloaded from

Page 6: A System of Thesis and Dissertation Supervision: Helping Graduate Students Succeed

(91%), contracts (90%), writing guidelines (88%), coordina- tion meeting (88%), structure of the supervisory meeting (88%), literature review forms (80%), experience as a super- visor (73%), and feedback form (70%). In addition, 89% of the supervisees said they wished to remain in the system, and 79% of the supervisors said in the same questionnaire that they wished to continue supervising.

Nine of the 29 (31%) students withdrew from the super- visory system. At the time of withdrawal, they said they could not invest as much time as required in the system. Five of these students entered the system in Semester 1; 3 withdrew at the end of Semester 1 and 2 at the end of Semester 2. One other student participated in the system only in Semester 2. The remaining 3 students participated in the system 4 weeks or less; thus, we did not include their data.

Befween-Group Results

By Semester 2, students in the supervisory group had made more progress than students in the college-wide group, including those students who eventually dropped out of the system. Thirty-six percent more of the students in the system had defined their topic, 25% more had read 13 or more articles. 11% more had collected data. and 16% more had written ;heir first draft.

Students in the system made this greater amount of progress, even though their mean cumulative GPA was lower (.44 points lower for M A students and .26 lower for PhD students), their time in the program from entry to the survev date was shorter (13 months less for M A students and 2$ months less for PhD students), and the number of academic credits taken simultaneously with research cred- its was higher (6.2 credit hours more for M A students and 2.6 more for PhD students). The better performance of students in the system, despite their less favorable academic conditions, suggests the power of the supervisory system in helping students move toward completion of their theses and dissertations.

Furthermore, by Semester 2, only 15% of the MA stu- dents in the college-wide group had completed a thesis and no PhD student had completed a dissertation. However, by the end of this study, 50% of the M A students in the supervisory group had completed a thesis, 43% of the PhD students had com~leted a dissertation. and none of the students who drodDed out had comdeked a thesis or dis- . . sertation. Therefore, students in the supervisory system had completed 30% more research projects, even though their time in the program at the end of the study was still shorter than the time of students in the college-wide group by Semester 2 (1 month less for MA students and 17 months less for PhD students).

Students in the supervisory group and the college-wide group also differed in other characteristics. The supervisory group had 45% Black Americans, 30% White Americans, and 25% foreign students; however, the college-wide group had no Black American. 87% White Americans, and onlv 13% foreign students. Master's students in the supervisor^ - group averaged 2 years younger and doctoral students 8 years younger than did students in the college-wide group. Students in the supervisory group worked an average of 18

hr per week in nonacademic jobs, and students in the college-wide group worked an average of 25 hr per week.

The high rate of progress in the supervisory group might have resulted from the greater amount of supervision students received. In comparison with students in the college-wide group, 57% more of the students in the system attended at least one supervisory meeting every 2 weeks, 85% more had weekly task specification, 52% more re- ceived frequent feedback, and 86% more received incen- tives for working on their research.

One could argue that differences in progress resulted from characteristics of respondents to the survey, rather than from amount of supervision. For instance, curricular differences and possible different research requirements could have determined the lesser progress of the students in the college-wide group. However, 30% of these students were psychology students not involved in the supervisory system. Three psychology students made as little progress in their research as did students from other departments, suggesting that curricular differences might not have been the crucial variable.

We selected participants of the college-wide group based on their completion of the survey on thesis progress and supervision. University regulations restricted the use of student data to surveys and some university records, con- ditional on the student's written consent. This restrictiorl determined the self-selection bias in the college-wide sam- ple.

We surveyed the large college-wide group rather than a smaller one restricted to psychology. We did this in an effort to obtain a large enough sample to allow for matching the students of the supervisory group according to aca- demic and demographic characteristics. However, the low rate of survey return prevented such matching.

Although the college-wide group was a biased sample, ~t

is difficult to estimate the direction of the bias. One could argue that the respondents were the students who procras- tinated less on their theses and dissertations or thosc who procrastinated more.

The Dean of the Graduate College ranked the items ot the thesis evaluation form on a scale ranging from lou'csr ( 1 I to highest (4). The Dean assigned mean evaluations of 2.40 to the 5 theses completed in the system and 2.88 to the 1 i theses completed in the psychology control group. T h c quality of the 5 theses completed by the supervisory group represented the quality of the theses conducted in thc system. (Students in the system completed 3 dissertations, 1 thesis, and 3 M A projects in the 2 months following collection of thesis-quality evaluations.)

The supervisory system was reasonably inexpensive: C:o- ordination by student staff required 5 to 10 hr per week, including data processing, distribution of feedback anti materials, system evaluations, and problem solving. 111 addition, the second author spent 9.5 hr a week supervising 6 dissertations and 14 theses.

In conclusion, comparisons within the supervisory syi- tem suggested that students made more progress in their theses and dissertations when they received irlcentives t h a l ~ when they did not. However, students in special circum- stances seemed to require stronger incentives than regular students. Comparisons with the college-wick group u g -

190 Teaching of Psychology at MEMORIAL UNIV OF NEWFOUNDLAND on November 25, 2014top.sagepub.comDownloaded from

Page 7: A System of Thesis and Dissertation Supervision: Helping Graduate Students Succeed

gested that students made more progress when they re- ceived supervision and incentives than when they did not. These findings were consistent with previous studies on thesis and dissertation completion.

References

Berelson, B. (1960). Graduate education in the United States. New York: McGraw-Hill.

Blanton, J. S. (1983). Midwifing the dissertation. Teaching of Psychology, 10, 74-77.

Cartter, A. (1965). The decades ahead: Trends and problems. In E. Walters (Ed.), Graduate education today (pp. 223-245). Wash- ington, DC: American Council on Education.

Dillon, M. J., Kent, H. M., & Malott, R. W. (1980). A supervisory system for accomplishing long-range projects: An application to master's thesis research. Journal of Organizational Behavior Management, 2 , 2 13-227.

Dillon, M. J., & Malott, R. W. (1981). Supervising masters theses and doctoral dissertations. Teaching of Psychology, 8, 195-202.

Elfatouri, J. R., Garcia, M. E., & Malott, R. W. (1988). Success of psychology doctoral students: Attrition, discontinuation, and gradu- ation. Manuscript submitted for publication.

Fulton, B. J., & Malott, R. W. (1982). The structured meeting system: A procedure for improving the completion of nonre- curring tasks. Journal of Organizational Behavior Management, 3 , 65-72.

Cant, G. D., Dillon, M. J., & Malott, R. W. (1980). A behavioral system for supervising undergraduate research. Teaching of Psychology, 7 , 89-92.

Garcia, M. E. (1987). Behavioral research supervisory system manual. Unpublished manuscript, Western Michigan University, De- partment of Psychology, Kalamazw.

Garcia, M. E., & Malott, R. W. (1988). A solution to the all but dissertation phenomenon. Manuscript submitted for publication.

Graduate deans unhappy about time taken to finish degrees. (1988, March). University Affairs, p. 9.

Heiss, A. M. (1970). Challenges to graduate schools. San Francisco: Jossey-Bass.

Jacks, P., Chubin, D. E., Porter, A. L., & Connolly, T. (1983). The ABCs of ABDs: A study of incomplete doctorates. Improv- ing College and University Teaching, 31, 74-81.

Knox, W. J. (1970). Obtaining a Ph.D. in psychology. American Psychologist, 25, 1026-1032.

Lunneborg, C. E., & Lunneborg, P. W. (1973). Doctoral study

attrition in psychology. Research in Higher Education, 1,379-388. Malott, R. W. (1972). Contingency management in education.

Bridgewater, N J : Fournies. Malott, R. W. (1986). Self-management, rule-governed behavior,

and everyday life. In H. W. Reese & L. J. Parrot (Eds.), Behavior science: Philosophical, methodological, and empirical advances (pp. 207-228). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.

Moore, R. W. (1985). Winning the Ph.D. game. New York: Dodd, Mead.

Naylor, P. D., &Sanford, T . R. (1982). Intrainstitutional analysis of student retention across student levels. College and Univer- sity, 58, 143-158.

Sells, L. W. (1975). Sex, ethnic, and field differences in doctoral outcomes (Doctoral dissertation, University of California at Berkeley, 1975). Dissertation Abstracts International, 37, 637.

Snell, 1. L. (1965). The master's degree. In E. Walters (Ed.), Graduate education today (pp. 74-102). Washington, DC: Amer- ican Council on Education.

U.S. Department of Labor. (1 986). Outlook for college graduates and Ph.D.'s (Publication No. 029-001-02907-1). Washington, DC: U.S. Government Printing Office.

Von Zur-Muehlen, M. (1977). The relationship between graduate degrees awarded and enrollment by level and discipline. Ottawa: Statistics Canada.

Wilson, K. M . (1965). Of time and the doctorate. Atlanta, GA: Southern Regional Education Board.

Wright, C. R. (1964). Success or failure in earning graduate degrees. Sociology of Education, 38, 73-97.

Zoia, T. K. (1981). Completing long-term undergraduate projects: Some critical variables. Unpublished master's thesis, Western Michigan University, Kalamazoo, MI.

Notes

1. This study is based on the doctoral dissertation presented by Maria Emma Garcia, the first author, as part of the require- ments for her PhD degree in psychology.

2. We thank Laurel Grotzinger for her contribution to the eval- uation of thesis quality and her support of this research. We also thank Susan Hannah, Bradley Huitema, David Lyon, and Wayne Fuqua for their careful editing of early drafts of this manuscript.

3. Requests for reprints and more information about the super- visory system should be sent to Maria Emma Garcia, De- partment of Psychology, Western Michigan University, Kalamazoo, MI 49008.

Vol. 15, No. 4, December 1988 at MEMORIAL UNIV OF NEWFOUNDLAND on November 25, 2014top.sagepub.comDownloaded from