Adaptation française et propriétés psychométriques de l’échelle d’utilisabilité perçue des sites web « Design-oriented evaluation of perceived usability (DEEP) »

Abstract

Introduction

With the ubiquitous and often unavoidable use of websites, it is necessary that they present the best possible usability to satisfy users. The Design-Oriented Evaluation of Perceived Usability (DEEP) questionnaire, comprising 19 items, assesses perceived usability on the basis of six genotypic criteria: content, structure and information architecture, navigation, cognitive effort, layout consistency, and visual guidance. After completing the DEEP, designers are then able to identify the main factors of the website that cause problems.

Objective

The aim of this study was to propose a cross-cultural adaptation of the DEEP in French and to evaluate its psychometric qualities.

Method

Four hundred and seventy users completed an online survey to give their opinion on the Onisep.fr website. The survey included the French translation of the DEEP (F-DEEP) and the Net Promoter Score questionnaire. A principal component analysis (PCA) with oblimin rotation, an internal consistency analysis, several confirmatory factor analyses (CFA) and a convergent validity study between the F-DEEP and the NPS were conducted.

Results

The PCA suggests a 6-factor structure, which partly corresponds to the 6 factors of the original questionnaire. The internal consistency is very satisfactory, with a Cronbach's alpha coefficient of 0.944 and a McDonalds omega of 0.945. The CFA with the most satisfactory fit indices corresponds to the first-order model including the 6 factors of the F-DEEP. All the correlations calculated between the F-DEEP factors and the NPS are positive and highly significant.

Conclusion

The F-DEEP has very good psychometric qualities, which means that it can be used by researchers and professionals in the field of ergonomic psychology of human-computer interaction to evaluate websites for French-speaking audiences. This questionnaire can also be applied to the evaluation of the usability of mobile applications.