Exploring Possible Sociocultural Bias on the SCAN-C

Abstract

Sociocultural bias on the SCAN-C (R. W. Keith, 2000) was investigated with 20 Anglo American and 20 Latino American 8-year-old children from low- and mid-high-socioeconomic status (SES) backgrounds. Univariate and repeated measures analyses of variance (ANOVAs) failed to reveal any significant differences between the groups when clustered by ethnicity and SES. The Latino American participants' scores were analyzed for dialectal variations, and the ANOVA analyses were repeated using the corrected scores. No significant interactions were observed. Classification analyses revealed that 10% more Latino American children than Anglo American children fell into the borderline-to-disordered category based on SCAN-C composite scores; these classification differences were most apparent on the Filtered Words subtest (with a difference of 25%). When scores with dialectal rescoring were considered, the classification distribution for the Latino American children more closely matched that of the Anglo American children. Given the increased likelihood of Latino American children scoring in the borderline-to-disordered category, caution should be used in interpreting SCAN-C results for Latino American children. Dialect scoring should be applied when Latino American children fall in the borderline-to-disordered category.