Abstract:

The mentoring relationship between a doctoral student and his or her dissertation supervisor is one of the most important factors in the student's successful completion of the degree. It is important for universities to be able to be able to identify which supervisors are effective and which are not in order to identify the best practices of the former and to improve the mentoring of the latter. North Carolina State University in the United States has developed a program that allows the Graduate School, which oversees doctoral education, to achieve both of these goals. The program applies a systematic method to pinpoint the most important parameters of effective mentoring and displays the results in a way that makes them easily usable. In addition, this method includes the ability to "drill-down" in the data, through organizational units of college, department and program all the way down to the individual faculty member and his or her students. In this way, one can determine which units, as well as individual faculty, have the best (and worst) records for mentoring students. We have identified four major areas of analysis: Contribution to Graduate Student Research, Effectiveness in Advancing and Graduating Students, Effectiveness in Enabling Doctoral Students to Complete Their Degrees, and Contribution to Interdisciplinary/Interdepartmental Graduate Student Research. Some primary parameters on which statistics are captured include number of advisory committees chaired or co-chaired for doctoral students, average time to degree of doctoral students for whom the faculty member served as a committee chair or co-chair, ten-year doctoral completion rate for students of the faculty member serving as chair or co-chair of their committee, and number of advisory committees for thesis and dissertation students which the faculty member served on in a department outside of his or her home department. To present this information in a useful way, a secure on-demand web application was developed. Data for all graduate faculty members is captured nightly from the Oracle PeopleSoft database at North Carolina State University and is further grouped by organizational structures such as college, department and program as well as faculty characteristics such as tenure rank, gender and number of years at the university, in order to form a data cube. The web application then directly accesses the data from the data cube in order to provide statistics in the four major areas of analysis. Data can be "drilled-down" so that the by a single click, one can observe the value of a parameter first from the university level, then proceeding down to the college, department and individual faculty member level. In this paper, we will discuss in detail the methodology used for the mentor analysis and how the web application was developed. We will also provide an example of how the application was used to determine mentoring areas in need of improvement as well as those which can serve as a model for the university.

Keywords:

Assessment, graduate education, web application.