Cultural adaptation to Peruvian Spanish and psychometric validation of an instrument to measure scientific literacy in medical students

Abstract

Introduction. Scientific literacy comprises skills to explain phenomena scientifically, evaluate science and interpret its results. Validated tools in different languages are needed to ensure their correct measurement. Objective. To adapt and validate, to Peruvian Spanish, an instrument to measure scientific literacy skills in medical students. Methods. The cultural adaptation of the Test of Scientific Literacy Skills (TOSLS) into Spanish had six steps. First, an independent, simultaneous translation from English into Spanish was performed. Then, both translations were synthesised and back-translated, followed by comparison with the original version and back-translation. This was followed by a pilot test and a cross-sectional study with medical students in Lima, Peru. The last step involved psychometric evaluation using the Rasch model. The instrument consisted of 28 multiple-choice questions with only one correct answer. Results. 300 students were enrolled. The Kuder-Richadson coefficient was 0.742. The items showed unidimensionality and local independence (MADaQ3 = 0.054). Item 12 was the most difficult, while item 5 was the easiest. The infit and outfit statistics of the items were in the range of 0.7 to 1.3 logits. The test information function provided information with greater precision at skill levels between -1 and 0 logits. There was a correspondence between item difficulty and scientific literacy ability. Conclusion. The Spanish version retained the 28 items of the original TOSLS and had acceptable internal consistency. The items had acceptable properties, independently of students’ scientific literacy ability.