Remerciez-le!

Remerciez @Admin pour avoir partagé cet document gratuitement, de la manière la plus simple, en partageant sur les réseaux sociaux.

Programme 3 Artificial intelligence, cognitive systems, and man-machine interac

Programme 3 Artificial intelligence, cognitive systems, and man-machine interaction. Ergonomic Criteria for the Evaluation of Human-Computer Interfaces • Critères Ergonomiques pour l’Évaluation d’Interfaces Utilisateurs (version 2.1) J. M. Christian Bastien and Dominique L. Scapin Technical report N° 156 May 1993 Ergonomic Criteria (version 2.1) iii Abstract This technical report presents first a brief summary of the research conducted towards the design of ergonomic criteria for the evaluation of human-computer interfaces (HCI), and then, the full description of the most recent set of criteria (version 2.1) both in English and French. The summary outlines the context in which the criteria were developed, the goal of the criteria approach, the experiments conducted, and the results obtained. The set of ergonomic criteria that resulted from this work consists of a list of 18 elementary criteria (including 9 main criteria). The criteria are presented along with their definitions, rationales, examples of guidelines, and comments setting out the distinctions between some of them. Key-words: Analytic evaluation, User-interface evaluation, Ergonomic criteria, Normative dimensions, Standards. Résumé Ce rapport technique présente d’une part un bref résumé des travaux menés dans le cadre de la conception de critères ergonomiques pour l’évaluation des interfaces utilisateur (HCI) et d’autre part, la dernière version (version 2.1) du jeu de critères en anglais et en français. Le résumé expose dans ses grandes lignes le contexte dans lequel s’inscrivent les critères, l’objectif de l’approche par critères, les expériences réalisées et les résultats obtenus. Le jeu de critères issu de ces travaux est constitué de 18 critères élémentaires (dont 9 critères principaux). Les critères sont présentés avec leurs définitions, justifications, exemples de recommendations et commentaires permettant d’expliciter les distinctions entre certains d’entre eux. Mots-clés: Évaluation analytique, Évaluation d’interfaces utilisateur, Critères ergonomiques, Dimensions normatives, Standards. Ergonomic Criteria (version 2.1) v CONTENTS Abstract.................................................................................................................................iii Résumé..................................................................................................................................iii I. Introduction........................................................................................................................ 1 II. The Ergonomic Criteria (English Version)......................................................................... 7 1. Guidance................................................................................................................... 9 1.1. Prompting ................................................................................................... 10 1.2. Grouping/Distinction of Items ..................................................................... 13 1.2.1. Grouping/Distinction by Location ................................................... 14 1.2.2. Grouping/Distinction by Format...................................................... 15 1.3. Immediate Feedback.................................................................................... 16 1.4. Legibility..................................................................................................... 18 2. Workload................................................................................................................ 20 2.1. Brevity........................................................................................................ 21 2.1.1. Concision........................................................................................ 22 2.1.2. Minimal Actions.............................................................................. 23 2.2. Information Density..................................................................................... 25 3. Explicit Control....................................................................................................... 26 3.1. Explicit User Action.................................................................................... 27 3.2. User Control............................................................................................... 28 4. Adaptability ............................................................................................................ 29 4.1. Flexibility.................................................................................................... 30 4.2. User Experience.......................................................................................... 32 5. Error Management.................................................................................................. 34 5.1. Error Protection.......................................................................................... 35 5.2. Quality of Error Messages........................................................................... 36 5.3. Error Correction ......................................................................................... 37 6. Consistency............................................................................................................. 38 7. Significance of Codes.............................................................................................. 39 8. Compatibility .......................................................................................................... 40 III. The Ergonomic Criteria (French Version)....................................................................... 43 1. Guidage .................................................................................................................. 45 1.1. Incitation..................................................................................................... 46 vi Ergonomic Criteria (version 2.1) 1.2. Groupement/Distinction entre Items ............................................................ 49 1.2.1. Groupement/Distinction par la Localisation..................................... 50 1.2.2. Groupement/Distinction par le Format............................................. 51 1.3. Feedback Immédiat ..................................................................................... 53 1.4. Lisibilité ...................................................................................................... 55 2. Charge de Travail.................................................................................................... 57 2.1. Brièveté ...................................................................................................... 58 2.1.1. Concision........................................................................................ 59 2.1.2. Actions Minimales........................................................................... 61 2.2. Densité Informationnelle.............................................................................. 63 3. Contrôle Explicite ................................................................................................... 64 3.1. Actions Explicites........................................................................................ 65 3.2. Contrôle Utilisateur..................................................................................... 66 4. Adaptabilité............................................................................................................. 67 4.1. Flexibilité .................................................................................................... 68 4.2. Prise en Compte de l’Expérience de l’Utilisateur.......................................... 70 5. Gestion des Erreurs................................................................................................. 72 5.1. Protection Contre les Erreurs ...................................................................... 73 5.2. Qualité des Messages d’Erreur .................................................................... 74 5.3. Correction des Erreurs ................................................................................ 75 6. Homogénéité/Cohérence ......................................................................................... 76 7. Signifiance des Codes et Dénominations.................................................................. 77 8. Compatibilité........................................................................................................... 78 Ergonomic Criteria (version 2.1) 1 I. INTRODUCTION The design of ergonomic criteria is part of a wider research project1 aiming at the development of methods and tools that would incorporate human factors considerations into the process of designing and evaluating human-computer interfaces (HCI). While other parts of the research work concern task issues, expertise issues, interface modelling issues, etc., the design of criteria is viewed as a means of defining and operationalising dimensions of usability. With respect to the evaluation, the design of criteria represents a way of improving the completeness and explicitness of the diagnosis, of standardising the format of the evaluation, and of better documenting the evaluation. Criteria can also be helpful for teaching HCI issues. Other requirements for an evaluation method based on ergonomic criteria were that it should: - be based on analysis of the interface rather than user testing, owing to cost and time constraints; - be usable also by non-human factors specialists (e.g., interface designers) for whom direct availability of human factors results is important; - be explicit so as to permit measurements, and sufficiently standardised to be replicated. Normative dimensions such as criteria can be based on different design strategies. For instance, one can start from available knowledge in domains such as cognition, reasoning, memory, language, etc.; or start from currently published dimensions or heuristics with the goal of organising them into a common structure; or starting from experimental data and recommendations. The strategy chosen here was a bottom-up or inductive strategy (Scapin, 1990b). Available experimental data and recommendations were first translated into rules, then iteratively grouped into sets that were characterised by specific “criteria” which were supposed to best describe the rationale for using such recommendations (initially as a way of accessing a data base). After a few iterations, using also decisions based on group agreement, a set of criteria was established (Scapin, 1990b; 1990c). The validity and reliability of a tool for the evaluation of HCI are important issues. The validity of the criteria, i.e. the extent to which they allow one to evaluate the ergonomic quality of an interface comes from the strategy adopted in their design. As indicated above, the set of ergonomic criteria is based on a rather large set of individual recommendations that they characterise and synthesise. In this sense the criteria may be said to be valid. Although the set of criteria has already been found to allow an accurate description and classification of the usability problems found by experts in an evaluation task (Pollier, 1992), a more direct assessment of its reliability was needed. Reliability (or, in this case, evaluators' agreement) can be established by determining the consistency with which the criteria can be used: for instance, if evaluators use the criteria in the same way, i.e. if they uncover the same design flaws, or if they classify them under the same criteria, then the criteria may be said to provide reliable performances. Consistency of use can thus be assessed in different tasks (e.g., classification of design flaws, aid for the evaluation). The factors determining the reliability of the criteria are: the quality of the 1 The research project has been summarised in Scapin (1990a). A more detailed presentation of this project can be found in Scapin (in press). 2 Ergonomic Criteria (version 2.1) definitions; the explicitness of the rationale; the relevance of the examples provided with each criterion; and the distinctions given between criteria that are conceptually “close” to one another. A first assessment of the reliability of the set of criteria was performed in a classification task (Bastien and Scapin, 1992). The rationale of the experiment was that if the criteria were sufficiently well defined and exemplified, then evaluators would be able to identify them correctly, and assign them appropriately to interface problems: in other words the set of criteria would be reliable, and the agreement between would be high. Twenty-four subjects (12 human factors specialists and 12 non-specialists) were asked to identify which criterion was violated for each of thirty-six usability problems they were presented with. The results of this study showed no statistical difference between groups both in terms of performance times and in terms of correct identifications. The mean percentage of correct identifications was 63,7% and 56% for the human factors specialists and the non-specialists respectively. A detailed examination of the data and an analysis of confusion matrices allowed the identification of categories of well defined criteria and categories of criteria that were to benefit from improvements in their definition. Even though the experimental situation was very restrictive, the overall assignment performance was above average. The improvement of the set of ergonomic criteria proceeded through the inclusion of new examples, the addition of comments allowing a better distinction between criteria, and the refinement of some definitions. The effectiveness of ergonomic criteria as a tool or a guide for the evaluation of user interfaces was assessed in another experiment (Bastien and Scapin, 1993). Two groups of human factors specialists evaluated the interface of a musical database application. After an exploration- diagnosis phase in which all the evaluators’ actions and comments were recorded along with the corresponding states of the application, the participants re-evaluated the same interface states through the replay of their first exploration. This second evaluation was conducted with or without ergonomic criteria depending on the group. Preliminary results show that in the first phase, the number of usability problems uncovered and the proportions of usability problems found with respect to the size of the aggregates of evaluations were similar for both groups. In the second phase of the experiment, the use of criteria increased both the evaluation diagnosis and the proportions of problems with respect to the size of the aggregates. The criteria were thus uploads/Litterature/ bastien-e-scapin-1993-ergonomic-criteria-for-the-evaluation-of-human-computer-interfaces.pdf