Cross-cultural adaptation (CCA): after all, the questionnaire is good to be used? - A pilot study of the applicability of the European Drug Abuse Prevention Trial (EU-Dap) questionnaire in Brazilian middle-school adolescent

Thursday, 21 August 2014: 11:15 AM
Kahtnu 2 (Dena'ina Center)
Mariangela C Prado, DrPH , Universidade Federal de São Paulo, São Paulo, Brazil
Daniela R Schneider, PhD , Universidade Federal de Santa Catarina, Florianópolis, Brazil
Adriana Sanudo, MS , Universidade Federal de São Paulo, São Paulo, Brazil
Zila M Sanchez, PhD , Universidade Federal de São Paulo, São Paulo, Brazil
INTRODUCTION: The accuracy of data in epidemiological studies employing self-report questionnaires developed in another socio-cultural language context depends on the success of CCA. The CCA is a complex, expensive and time consuming process developed in successive steps aimed at achieving validity and reproducibility, but there is no consensus on how best to accomplish it. The aim of this study is to evaluate the reproducibility and semantic validity of the translation into Portuguese of the Eu-Dap questionnaire

METHODS:  A pilot study using a mixed method design was carried out in three Brazilian cities in 2013. A total of 3835 students aged 11 to 14 years old were enrolled. Self- report data about the use of licit and illicit drugs, normative beliefs, knowledge and opinions about substances, social and personal skills were investigated through a Brazilian version of the EU-Dap questionnaire. Missing data analysis was performed among 3543 questionnaires. Reproducibility analysis through Kappa was investigated among 256 students and Semantic Analysis was carried out among 36 students in five focus groups

RESULTS:  Low reproducibility after 21 days was observed in most of the questions. The lowest scores (Kappa 0.16 to 0.47) was achieved in the normative beliefs construct. Highest reproducibility was found in socio demographic questions (Kappa 0.67 to 1.00). Missing analysis showed the difficulty in understands of questions about drug use and the length of application time leading to up to 78% of missing data in final questions. Focus Group pointed to serious flaws in understanding the questions on normative beliefs.

CONCLUSIONS: Triangulation of data collection and analysis (qualitative- quantitative methods) was crucial in the demonstration of the lack of applicability of the questionnaire in the Brazilian context. The reason for this is not clear, it suggests possible flaws in the English version or the influence of Brazilian context level factors

See more of: Youth, Alcohol, and Substance Abuse Studies: Free Papers
See more of: Free Papers
<< Previous Abstract | Next Abstract >>