Evaluation of Demographic and Language Predictors of Main Concept Production in Spanish/English Bilingual Discourse Using Nicolas and Brookshire Stimuli

Rivera, Ana and Edmonds, Lisa A. (2014) Evaluation of Demographic and Language Predictors of Main Concept Production in Spanish/English Bilingual Discourse Using Nicolas and Brookshire Stimuli. [Clinical Aphasiology Paper]

PDF 533-1151-1-RV(Rivera-Edmonds).pdf Download (100kB)

Abstract

Hispanics are the fastest growing minority group in the US, and stroke incidence for Hispanics is higher than that of non-Hispanic caucasians (Dong et al., 2012). Since stroke is the leading cause of aphasia, and the majority of Hispanics are bilingual Spanish/English speakers, there is a growing need for appropriate assessment methods for bilingual aphasia. One challenge in assessing bilinguals is their diverse language backgrounds, including differences in proficiency and use across languages. Previous studies have reported correlations between variables such as frequency of use and self-rated proficiency and performance in language tasks such as picture-naming accuracy (Edmonds & Donovan, 2012; Gollan et al. 2007), verbal fluency (Muñoz & Marquardt, 2008, Langdon et al., 2005; Elgamal et al., 2011), and discourse informativeness (Edmonds, 2013). In a study investigating discourse in 83 English/ Spanish bilingual adults, Edmonds (2013) reported varied patters of correlations for informativeness and efficiency measures (WPM, %CIUs, CIUs/min ) in English and Spanish. Naming accuracy and overall proficiency were significantly correlated with all measures of informativeness regardless of language; whereas, percent usage and age of exposure varied across languages (Edmonds, 2013). The metric of %CIU is often used to evaluate informativeness in discourse, but %CIU does not address the completeness of the discourse. I.e., one can achieve high %CIUs by providing correct information on half the picture, thus, targeting only half of the main concepts (MC). Alternatively, many MCs can be discussed with lower %CIU due to repetitions, circumlocutions and reformulations. In order to understand this dynamic better in bilinguals with varying degrees of proficiency across langauges, we adapted Nicholas and Brookshire’s (1993, 1995) MC analysis to this population. Our research questions were: 1) Is there a relationship (correlation) between %CIUs and MCs? 2) What self-reported participant variables (e.g., proficiency ratings, usage) and tested language variables (discourse variables, confrontation naming) correlate to English and Spanish MCs? 3) Of the significantly correlated variables, which contribute most to regression models of MCs in English and Spanish?

Actions (login required)

View Item