Fielding Graduate University
Research Poster Session at National Session 2011
Digit Span with a Linguistically Diverse Latina/Latino Population: A Cross-Language Study -- Darlene Wheeler, PhD, Alumna (2010), School of Psychology
Variance in language is an important characteristic of the substantial Hispanic population residing in the United States. Insufficient consideration of that language diversity may bias clinical assessment and confound research. Cross-disciplinary research literature indicates there are group differences in performance on auditory digit span tasks based on language of administration (English > Spanish), perhaps, secondary to the increased number of syllables for digit words in Spanish over English – one component of phonological loop theory of working memory. This study included examination of how the language of administration (English or Spanish) of the Neuropsychological Assessment Battery (NAB) Digits Forward/Digits Backward Test affected performance for 5 language groups using the Bilingual Dominance Scale (Dunn & Fox Tree, 2009) to weight 3 bilingual groups. Further, whether adapting digit span tasks by developing alternative items more comparable in word length would reduce or eliminate the mean differences in performance among groups was explored. Study participants (N=74) were volunteers recruited from community settings. A repeated measures design using analysis of covariance (ANCOVA), statistically controlling for education, was employed to compare performance on the NAB versus adapted digit span tasks. Results were mixed with language and task types (NAB/adapted, digits forward/backward) affecting performance and interacting divergently. Significant main effects were found for language (English > Spanish) on digits forward and for education (more > less) on digits backward tasks. Mean group differences persisted on adapted digits forward and backward measures although they were reduced in the adapted condition and there was no significant difference between English and Spanish on the backward adapted measure. Insufficient power due, in part, to insufficient subgroup sample size; and the effect of repetition inhibition on adapted tasks may have contributed to this result. Language dominance, highly correlated with length of education in U.S., affects performance on regular digit span tasks (forward and backward) for bilingual groups (English-weighted and Balanced > Spanish-weighted) . Study participants reported using various sensory modalities and strategies to aid in digit recall. Comparison to English normative samples of scores on digit span tasks administered in Spanish is inappropriate. Further, when administered in English to those who are of indeterminate extent bilingual, any comparison should be made only with caution.