Date of Degree
Education | Sociology
College; Education; Ethnicity; Graduation; Measurement; Sociology
The way in which college graduation is modeled matters with regard to the conclusions that a researcher is able to draw from the data. This dissertation explores different approaches to modeling degree pursuit and graduation that have implications for how researchers should model graduation. These implications include measuring degree pursued at entry to and exit from college to account for changes in level. Associate students who transfer to the baccalaureate level in particular are important to measure because of how different their outcomes are compared to associate students who stay at the associate level. Further, a variety of ways of measuring graduation are explored, from graduation as an alternative among others (such as dropping out or transferring) to binomially measured graduation (on-time versus delayed) to a typology of delayed graduation and finally graduation measured as time to degree. Additionally, decision trees are used to guide the testing of interaction terms, including interactions of ethnicity and a variety of measures of high school preparation and early college performance. The results indicate that the effects of these measures vary by ethnicity and interactions of this type should be included in any study of educational stratification. In particular, effect of high school GPA for different ethnic groups is explored. The policy implications of this for admissions decisions is discussed.
Wallace, Andrew Middleton, "The Measure of a Man: The Role of Measurement in Shaping Our Understanding of College Graduation along Ethnic Lines" (2015). CUNY Academic Works.