The purpose of this study was to empirically test the construct validity of Gómez' Ethnicity and Public Recreation Participation (EPRP) Model. The EPRP Model was tested in the field on 311 Puerto Ricans in Massachusetts. Confirmatory factor analysis, path analysis, and fit indices were employed to test the model. The results indicated that all scales possessed good validity and reliability, and that the model met various goodness of fit criteria. Theoretical and practical implications of the EPRP Model are discussed.
I am extremely grateful for the comments, time, and patience from the anonymous reviewers and the editors. Many thanks to the Michigan State University faculty: Drs. Daniel Stynes and Frank Boster for their help in assessing unidimensional measurement, and Drs. Michael Chubb, Joseph Fridgen, and John Schweitzer for their help with item construction and wording.
Notes
a Numbers corresponding to items are used in consequent tables;
b Shaded items passed tests for internal consistency and parallelism, and used in CFA;
c factor loadings reported only for final items in CFA.
a Predicted Correlations in bold, top right;
b Factor loadings.
a z-test examines if errors are within sampling error of zero;
b if z-value outside of ± 1.96 interval, it indicates a large error at the .05 level of significance.