track students over all terms whether they are enrolled or not, and (3) exclude students explicitly enrolled in terminal master's degree programs.

A cohort of students consists of all those who first enroll in a degree program on a given date. At any future date, this cohort can be divided into four types of students:

  • (1)  

    those who have completed the degree;

  • (2)  

    those who have officially left the program without completing;

  • (3)  

    those who are still enrolled in the program; and

  • (4)  

    those who have "stopped out" and may return.

Over time, students move from groups (3) and (4) into groups (1) and (2). At some arbitrary date (e.g., eight years later), completion and attrition rates can be calculated based on the number of students in groups (1) and (2). Unfortunately, students at some institutions may remain indefinitely in groups (3) and (4), and researchers will need to decide whether to create a third category for these "noncompleters" or include them in the attrition rate.

Comprehensive tracking of graduate students will occur across a number of diverse graduate unit (e.g., programs, departments, colleges) within an institution. To design and mount an effective student tracking system at the doctoral level, an institutionally based committee or graduate office would be needed to provide data managers with specific information about important differences in the graduate programs that must be reflected in longitudinal student records. For example, does the doctoral program include a master's degree requirement on the way to the doctorate?

The data elements actually included in a longitudinal tracking system will include a minimal set of information about a student as well as optional information. Some elements will be determined by institutional information needs as well as reporting requirements imposed by external agencies, such as state and federal governments. The elements listed in the text box on the following page are offered for purposes of illustration without regard to other reporting requirements.2


A longitudinal tracking system can be used to develop several types of reports. Through manipulation of the data set, cohort analyses can be conducted, yielding the following types of derived measures:

  • number of students in a cohort
  • number/percent of students earning master's degrees by a particular semester
  • number/percent of students admitted to doctoral candidacy in a particular semester
  • number/percent of students in attendance in a semester (full-time or part-time or terminated)

In addition, the data base can be used to develop prescriptive reports which assist planners and institutional researchers to document variables contributing to attrition, such as the nature and timing of student


Coker and Feidel (1991) report that increasing emphasis on the assessment of institutional effectiveness has stimulated many campuses to examine existing data sources, surveys, and reports before proceeding to develop new measures.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement