| The percentage of graduate degrees awarded to African American, Hispanic, and Native American students, as a group, remains low. Undergraduate research programs are one strategy currently implemented to increase underrepresented students' participation in graduate education. Results of several research and evaluation studies indicate that the programs are effective in enhancing participants' perceptions of and interests in graduate study, but there is a lack of understanding concerning how or why the programs are effective.; The current study employed structural equation modeling techniques to develop and test a model to investigate the relationships among program outcomes and underrepresented students' plans to pursue graduate study. Data for the study were taken from surveys administered to students in the 1988-2006 cohorts of one summer research program, resulting in a sample of just over 600 students. Results indicated that the model had good fit to the data with the initial data set, chi2 (14, N = 319) = 17.47, p = .23, CFI = 1.00, TLI = 1.00, RMSEA = .03, and SRMR = .04 and with the cross-validation data set, chi2 (15, N = 317) = 19.78, p = .18, CFI = 1.0, TLI = 1.0, RMSEA = .03 and SRMR = .04. More specifically, in the cross-validation procedure, seven of the eight paths tested were significant at p < .05, indicating support for the hypothesized relationships among faculty preceptor relationship, program satisfaction, increased knowledge of graduate school and research, outcome value of program participation, increased interest in graduate school, and plans for graduate study. The amount of variance explained by the model was, 24% in sample one and 22% in sample two. The findings have implications for program design, program replication, program improvement, and evaluation. |
