Testing Very Preterm Black Children to Examine Potential Assessment Bias in Two Language Tests.

Purpose: The lack of appropriate assessments affects populations presumed to be most at risk for speech and language concerns, one of them being children with a history of preterm birth. This study aims to examine whether bias is present in the currently available tests for Black children under age three to examine potential cultural bias in the Communication and Symbolic Behavior Scales (CSBS) and the MacArthur Bates Communicative Inventory (CDI). Method: This study uses data from the Early Development of NICU graduates' study (EDeN), a National Institutes of Health-funded, ongoing prospective, multicenter cohort investigation of very preterm (VPT) infants. A total of 54 (18 Black; 36 White) children born VPT were included in this study. The outcome measures were the Black and White children's CSBS cluster scores and their CDI: Words and Gestures (WG) scores. T-tests and a mixed model approach were used to compare the scores of the CSBS and CDI: WG based on matched pairs. Spearman rank correlation was used to examine the relationship between the CSBS and CDI: WG scores. Results: There was no significant difference between CSBS scores of Black and White children. There were no significant differences between the Black and White children's total scores on the CDI: WG based on the parent responses. There was a weak correlation between the CSBS and the CDI: WG Gesture subtests scores for the total group and a moderate correlation for Black children. Conclusion: The comparison of scores on the CSBS and the CDI: WG showed no significant difference between the Black and White children, indicating that there may not be evidence to support bias for the premature children in this study. The correlation between the gestural subtest of the CSBS and the CDI: WG showed that direct observation by clinicians was similar to the observations of the parents.