Should Hispanic persons from different countries of origin be combined into one sample for analysis?

Hispanic respondents to 21 nationwide surveys were grouped on the basis of country of origin into Mexican Americans (n = 402), Puerto Ricans (n = 119), and Other Latin Americans (n = 105). Analysis showed no significant over-all differences among subgroups. No significant subgroup differences in work attitudes were found for women and only one for men. Analysis of subgroup attitudes across age, education, occupation, family income, and personal income also gave no significant differences. These results provide tentative evidence that Hispanic respondents from different countries of origin may be aggregated into a single sample for analysis. There may be no more reason to analyze their responses separately by country of origin than there is to do so for persons of European or African origin.