Toward a typology of Hispanic high school dropouts
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
A lag exists between the reduction in the rate of dropout from high school among Hispanic youth and their rate of growth among the school population. Although researchers have called for the need to discriminate among dropout types, no study to date has succeeded in classifying dropouts, especially Hispanic dropouts. This study used selected data for Hispanic cases from the database created for the National Education Longitudinal Study of 1988. All data except final dropout status was based on respondents' answers to the first phase of the survey, administered when they were in the 8 grade. Item selection was based on a theoretical dropout model by R. Rumberger. Variables found to be significant in research differentiating dropouts from persisting students were selected for the model. Groups were created using cluster analysis, and verified with confirmatory discriminant analysis. Three distinct groups emerged for Hispanic dropouts, with standardized test scores in reading explaining more than 60% of the variance among groups. Similar results were observed among the subsample of Mexican heritage dropouts, who accounted for more than 74% of the Hispanic sample and had more homogeneous characteristics.
Although this was an exploratory study, the results indicate that potential dropouts can be differentiated when they are grouped by similar characteristics. Cluster analysis may be useful for further discrimination among groups, with results used as a basis for developing a dropout diagnosis model similar to medical diagnostic models. A diagnostic tool targeting the educational needs of at-risk students could help schools develop more effective interventions before the potential dropouts enter high school.