Fresno test to measure evidence-based practice knowledge and skills for Portuguese undergraduate nursing students: a translation and adaptation study

Abstract

Background: The Fresno Test was originally identified as an instrument to assess evidence-based practice knowledge and skills through cognitive testing and performance assessment in medical students. Further studies have been recommended to establish the measurement properties of the Fresno Test in different learner populations. Objectives: To perform a cross-cultural adaptation of the Fresno Test for Portuguese undergraduate nursing students and to analyze the interrater reliability. Design: Cross-cultural adaptation study with interrater reliability assessment carried out in two phases during 2017–2018. Settings: One of the main nursing schools, Portugal. Participants: Fourth year undergraduate nursing students. Methods: The study was performed in two phases, firstly the cross-cultural adaptation (performed in five stages) and secondly the analysis of interrater reliability. Results: Stages I, II, III and IV of the cross-cultural adaptation proceeded smoothly and the expert panel produced and agreed upon the pre-final version of Adapted Fresno Test. In stage V (the pre-test stage), students reported a general understanding of the items, but they reported a lack of knowledge to answer the test. An expert panel subsequently agreed that modifications were needed to ensure the test was within the student’s competency level and to decrease risk of assessment bias. For phase II, 50 complete questionnaires were randomly selected to be rated by three independent nurses using the modified rubric to score the test. The overall interrater reliability was 0.826 with a range from 0.271 to 1.000 for each item. Conclusions: The Adapted Fresno Test presented in this paper is the first instrument translated for European Portuguese and adapted specially for undergraduate nursing students. Despite good interrater reliability, further validation studies with more robust samples are suggested to definitively establish psychometric properties beyond the interrater reliability.