Abstract
There has been a considerable amount of interest in graduation rates of baccalaureate degree programs in the educational literature. There has also been some attention given to graduation from doctoral programs and from associate degree programs. However, there is almost no literature on the graduation rates for Master’s degree programs. This report uses a method of analysis known as survival analysis or event history analysis to examine the time to event for two different events leading to completion of a Master’s degree. One event is the time to preliminary examination, and the other event is the time to final thesis defense or final examination. The data compare three different Master’s programs in a psychology department of one of the largest campuses of one of the largest public university systems in the United States.
TopIntroduction
There has been a considerable amount of interest in graduation rates for baccalaureate degrees in the educational literature, partially caused by an increasing emphasis on performance-based accountability (Alexander, 2000). One overt manifestation of this emphasis was the Federal Student Right-to-Know and Campus Security Act of 1991 that required disclosure of completion or graduation rates (Astin, 1997). Formulas have been provided for cut-off points at four, six and nine years after freshman entry (Astin, 1997). This federal legislation was justified in that the perspective of the federal government has been that graduation rates represented a way for external reviewers to compare different institutions to each other (Underwood & Rieck, 1999). The graduation rate they decided upon has been the six-year graduation rate (Astin, 2005-2006).
This six-year perspective in use for baccalaureate degree programs has been modified for community colleges awarding associate degrees to a standard of 150% of normal time. In this method, normal time for an associate degree is considered to be two years, so 150% of this would be three years (Bailey, Calcagno, Jenkins, Leinbach, & Kienzl, 2006).
The six-year rate is also used for doctorate-granting universities, and there has been attention paid to predicting graduation rates for the doctoral degree (Goenner & Snaith, 2003-2004). Graduate faculty are often concerned with students who leave a department, and in some cases consider those students to be failures of the department (Pauley, Cunningham, & Toth, 1999). This means that the higher this failure rate, the greater is the perception that the university is failing to meet student needs (Pauley et al., 1999). A useful study examined the attrition and retention of students in a non-traditional Ed.D. program at a single university (Pauley et al., 1999). This study is useful for other universities who are also examining degree programs other than the standard baccalaureate and doctorate.
The current study was motivated to examine the retention and degree completion performance of three different types of masters degrees in one department of the largest campus of the California State University system which itself is the largest senior system of higher education in the United States (The California State University, 1999). We have been unable to locate any published literature on degree completion rates of masters programs. Our study is an attempt to aid others who may want to examine master’s degree programs and are also having a difficult experience in locating directly relevant literature.
Analysis Background
The general approach taken in the analysis is that the important measure is time to degree. This is because programs with high doctoral completion rates are those with shorter times to earn the degree (de Valero, 2001). The analytic problem with using time-to-event data with graduate school admission/graduation data is that some of the students who were admitted may not have graduated by the time of cessation of data collection. These data points are termed “censored” data because the date of graduation is unknown (it may never occur). There is a method of analysis that is able to deal with data with this characteristic known as survival analysis (Fisher & Anglin, 1987; Singer & Willett, 1991). This method has been applied to examining drug abuse treatment program retention (Fisher, Lankford, & Galea, 1996) and to employment retention of healthcare providers, among others (Fisher, Pearce, Statz, & Wood, 2003). An early introduction of this method to studying student dropout advocated using survival analysis as a way to explore important unanswered questions in educational research (Willett & Singer, 1991). This method has been applied to study the retention of undergraduate students (DesJardins, Ahlburg, & McCall, 1999; Murtaugh, Burns, & Schuster, 1999). We are not aware of it ever being applied to graduate school education.
Key Terms in this Chapter
Final Examination: An assessment of mastery of knowledge, concepts, theories, and practices in a specific area of study.
Retention: The ability of a program or university to keep students from leaving a degree program to pursue some other activity.
Censored Data: In survival analysis, the outcome event of interest is dichotomized (0 = event did not occur or 1 = event did occur); censored data refer to events where the event did occur (that is, those students coded as 1 for having achieved graduation during the period under study).
Master’s Degree: A post-baccalaureate degree in a field of study, generally conceived of as a two-year program of study.
Survival Analysis: An analytical approach originally developed for use in studies of cancer survivorship, it looks at the time to some event (traditionally, death in cancer studies; graduation in this study).
Clinician: An individual trained in an area of expertise, such as mental health or medicine, who delivers services directly to individuals.