The resent study com ares the nonverbal intellectual performance as meas d b y the Raven's &lowed Progressive Matrices of 68 Chicano and 28 Anglo third-grade boys. An important aspect of the study was the attempt to control for problems related to SES, language status of the subjects, and cultural content of the testing instrument. The null hypothesis of no difference was rejected as the -4nglo group scored significantly higher than did the Chicano group. $he results were discussed in terms of to* (omega square), a measure of statistical association. Because only 4% of the variance of the Raven's score8 could be accounted for by ethnicity, it was concluded that the ethnicity variable does not effectively predict intelligence.