Factor structure models for the Post-Concussion Syndrome Scale with monolingual Spanish-speaking adults from Colombia.

PRIMARY OBJECTIVE: The Post-Concussion Syndrome Scale (PCSS) is a self-report questionnaire that measures post-concussive symptom severity and has been primarily normed on young Caucasian samples. This study aims to explore the factor structure models of a Spanish translation of the PCSS at a chronic post-traumatic brain injury (TBI) time point. RESEARCH DESIGN: Descriptive and exploratory research designs were utilized. METHODS AND PROCEDURES: The study consisted of a monolingual sample of Spanish-speaking adults from Colombia, with 100 subjects in the control group and 70 subjects in the TBI group. A t-test, chi-square, and MANOVA were calculated to compare group differences. Cronbach's alpha was calculated to investigate reliability. Confirmatory factor analysis compared item loadings onto an existing four-factor model. Exploratory factor analysis sought to identify a new factor model if the loadings did not fit. MAIN OUTCOMES AND RESULTS: There were no group differences in demographic variables. Internal consistency was acceptable. Model fit indices revealed a poor fit with the original four factors. Item loadings revealed a novel six-structure model. CONCLUSIONS: While the PCSS appears to capture general post-TBI sequelae, the underlying factors may differ due to cultural and linguistic differences in Spanish-speaking individuals. Clinical implications and future directions are further discussed.

Item analysis of the Spanish version of the Boston Naming Test with a Spanish speaking adult population from Colombia.

OBJECTIVE: The Boston Naming Test (BNT) is a widely used measure of confrontation naming ability that has been criticized for its questionable construct validity for non-English speakers. This study investigated item difficulty and construct validity of the Spanish version of the BNT to assess cultural and linguistic impact on performance. METHODS: Subjects were 1298 healthy Spanish speaking adults from Colombia. They were administered the 60- and 15-item Spanish version of the BNT. A Rasch analysis was computed to assess dimensionality, item hierarchy, targeting, reliability, and item fit. RESULTS: Both versions of the BNT satisfied requirements for unidimensionality. Although internal consistency was excellent for the 60-item BNT, order of difficulty did not increase consistently with item number and there were a number of items that did not fit the Rasch model. For the 15-item BNT, a total of 5 items changed position on the item hierarchy with 7 poor fitting items. Internal consistency was acceptable. CONCLUSIONS: Construct validity of the BNT remains a concern when it is administered to non-English speaking populations. Similar to previous findings, the order of item presentation did not correspond with increasing item difficulty, and both versions were inadequate at assessing high naming ability.