Validation of a questionnaire to measure people’s ability to assess claims about treatment effects in Spanish




Poster session 3 Friday: Evidence Tools / Evidence synthesis - creation, publication and updating in the digital age


Friday 15 September 2017 - 12:30 to 14:00


All authors in correct order:

Pérez-Gaxiola G1, Austvoll-Dahlgren A2
1 Cochrane México - Hospital Pediátrico de Sinaloa, Mexico
2 Knowledge Center for Health Services | Norwegian Institute of Public Health, Norway
Presenting author and contact person

Presenting author:

Giordano Pérez-Gaxiola

Contact person:

Abstract text
Background: Every day people are faced with claims about treatment effects through the mainstream media, social media, or from family and friends. Such claims may include advice on how to prevent illness, or about the effects of treatments, or public-health interventions. Enabling people to make informed decisions about health and healthcare by improving their ability to critically assess such claims is an important public-health initiative. The Claim Evaluation Tools (CETs) consists of a set of multiple-choice questions (MCQs) that can be used for assessing people’s ability to assess such claims.

Objective: To describe the psychometric testing using Rasch analysis of a sub-set of multiple-choice questions taken from the CETs database in Spanish-speaking populations in Mexico.

Methods: We used purposeful sampling and included 172 children (10-15 years old) and 268 adults. A set of 22 MCQs were translated from English to Spanish, and the final questionnaire was applied to adults as an online questionnaire and to children in two middle schools. We explored the overall Item-Person Interaction as well as individual item and person fit. We also tested the items for potential item differential functioning (based on gender, age and mode of administration), and for dimensionality and local dependency.

Results: The item-person interaction and fit to the model was satisfactory (Figures) with a mean fit residual of -0.0643 (SD 0.60). The mean location of person’s ability is 1.348. Based on t-tests the MCQs were found to be satisfactory unidimensional, and there was no important local dependency suggesting that there is no redundancy in the MCQs. The reliability was found to be satisfactory with a Person-Separation-Index of 0.7, and the MCQ’s ability to discriminate between respondents was good overall.

Conclusion: Based on the findings from the Rasch analysis, a final set of 18 MCQs with satisfactory fit to the Rasch model was selected. This is the first set of tools available in Spanish to measure people’s ability to assess treatment claims. The MCQs set is freely available for non-commercial use and can be used for educational or research purposes.