Graduation Year


Document Type




Degree Granting Department

Interdisciplinary Education

Major Professor

Dedrick, Robert.


confirmatory factor analysis, El Salvador, validation, cross-cultural adaptation, HIV/AIDS knowledge and attitudes


This study translated, cross-culturally adapted and validated an instrument's scores for use in public high schools in San Salvador, El Salvador. The original instrument consisted of items developed by the Centers for Disease Control and Prevention (CDC) to assess HIV/AIDS knowledge and five dimensions of attitudes (Abstinence, Peer-pressure, Condom use, Drug use, and Threat of HIV) in grades 7 to 12 in the United States. Items were translated into Spanish using the back-translation method. The instrument was cross-culturally adapted using guidelines proposed by Guillemin, Bombardier, and Beaton (1993). A cross-culturally equivalent version of the original instrument was obtained using three different Salvadorian review panels and two pretests with Salvadorian high school students. An expert panel of HIV Salvadorian professionals validated the content and established its cultural acceptability for public school use.

A total of 483 students from 30 randomly selected public high schools in El Salvador participated in a series of validation studies. Confirmatory factor analysis of the translated instrument was used to evaluate the factorial validity of the five-factor attitudinal model. As part of the validation process, the translated Abstinence and Condom use subscales from the CDC were correlated with similar translated subscales from Basen-Engquist et al.'s (1999) published study as a measure of concurrent validity. Finally, internal consistency reliability (Cronbach's alpha) was determined with 483 students and test-retest reliability was obtained with a subsample of 39 students. Six major conclusions were: (1) The methodology used was successful in cross-culturally adapting the instrument. (2) HIV/AIDS content was rated as culturally acceptable and valid for use in public high schools of El Salvador.

(3) The reliability of the scores from the knowledge section was moderate (test-retest reliability coefficient = .49 and coefficient alpha = .57). (4) Reliability (coefficient alpha) of the five attitudinal subscales was inconsistent: .55 (Peer-pressure), .58 (Abstinence), 0 (Condom use), .24 (Drugs), and .30 (Threat of HIV). (5) Confirmatory factor analysis provided support for a 4-factor attitudinal model (Peer-pressure, Abstinence, Drug use, and Threat of HIV). (6) Concurrent validity of the translated CDC Abstinence subscale was strong. The results provided support for the methodology to cross-culturally adapt an instrument. The psychometric properties from the knowledge section and the attitudinal component related to abstinence were acceptable but additional research is needed before the Spanish instrument can be used with confidence in El Salvador.