logo BrailleNet

8th European e-Accessibility Forum
User-driven e-Accessibility

31/03/2014, 09:00 - 18:00 - Registration from 8:30
Cité des sciences et de l'industrie - Universcience - Paris

  Version française
Printable version

User Experience Evaluation: An essential step

Christian BASTIEN (Metz)
University of Lorraine-Metz

Speaker's information

Christian Bastien

Christian BASTIEN is Professor at the University of Lorraine-Metz. He holds a Master in Psychology of Cognitive Processes, a postgraduate degree in Ergonomics and a PhD in Cognitive Psychology with a speciality in Ergonomics from the Université René Descartes in Paris. Christian is currently Professor at the University of Lorraine in Metz where he directs the PErSEUs Laboratory (Psychologie Ergonomique et Sociale pour l’Expérience Utilisateur, EA 7312). After working as an engineering expert at the Institute for Research in Computer Science (INRIA), he held a lecturing post at the University René Descartes. Christian was a researcher at the Computer Research Institute of Montreal. His research focuses on the use of interactive software and the Web, user-centered design methods and ergonomic assessment methods.



Today we tend to talk of User Experience (UX) rather than ergonomics. Regardless of the term used, if integrated from the outset, User Experience allows better productivity, improved efficiency of products, lower costs of training and support and, above all, increased user satisfaction which leads to increased sales and higher incomes.

While certifying the ergonomic quality of a product or of a software application is not possible, following standards for the implementation of a user-centric approach can be relatively straightforward.

A single set of specifications can produce any number of different design solutions services, products and interfaces. It is therefore not only the quality of specifications that will determine the quality of an end product; the product needs instead to be evaluated by its end-users. To do this, there is not one single technique or approach.

Evaluation methods can be divided into four main groups:

  1. Evaluation methods based on models

    Evaluation methods based on models take into account the objectives of the proposed product and the steps the user must take to access the service. Such methods are rarely used for reasons of complexity and implementation.

  2. Evaluation methods based on inspection methods

    Evaluation methods based on inspection methods determine the extent to which a given interface follows a given set of recommendations. This helps to identify design choices that may go against recommendations. This evaluation method is used to analyse products early on in the design process and does not require the product to be finished.

  3. User testing

    User testing can identify interaction problems by putting people in real situations. Participants must perform tasks for which the product was developed and resulting data on performance can be supplemented by questionnaires. This evaluation method can be very expensive. With the Internet, it is possible to conduct user testing remotely. The University of Lorraine has developed a remote testing method using EVALYZER, a small plugin installed in a browser that can capture interactive behavior. This can also be supplemented by video taken via a webcam on remote servers. This tool can facilitate automate testing, but can also enable automated data analysis.

  4. Questionnaires

    Questionnaires can include anything from a dozen to a hundred questions. Their content should be worded so as to be easy to understand and answers are best formulated in terms of agreement or disagreement

In conclusion, to be able to properly assess the quality of a product from the user perspective, a hybrid approach combining several methods is needed.

video of the presentation by Christian Bastien (in French)

Retour à la liste des articles