Inria / Raweb 2004
Project-Team: Eiffel

Search in Activity Report, year 2004:
HELP

INDEX

Project-Team : eiffel

Section: New Results


Axis 3: Methodological aspects

Research on this axis involves work on two methodological aspects: methodologies for the researcher in order to analyze cooperative design situations, and user-centered methodologies for the designer in order to design and evaluate new systems. Progress is to be made on these aspects for the researcher to be able to analyze collective design situations (whether they are mediated or not and in both synchronous and asynchronous modes), by focusing on both the functional aspects of the task and performances and on interaction aspects of cooperation and how it is supported through language and textual or graphical representations. Existing methodologies in this domain are generally ad hoc. Our objective is to develop methodological principles that are adapted to the aims of the analysis and that can be generalized to any collective design situation. We also need methods to analyse error production mechanisms. Finally, one needs methods to be created for the development of cooperative and interactive design systems. At present, our methodological research activities are focused on innovative virtual-reality based interactive systems.

Analysis methods of collaborative design

Participants: Françoise Darses, Françoise Détienne, Willemien Visser.

Cognitive psychologists and ergonomists have proposed various methods for the analysis of individual verbal protocols, but much less for dialogues in collective work settings. Many professional activities, however, are carried out by people working together through verbal interaction. From a perspective of cognitive ergonomics, we have developed principles for the analysis of collaborative design, amongst which the COMET method [3]. An extension of COMET has been elaborated for the analysis of distant and mediated collaboration [43].

Dialogue analysis has long been the concern of linguistics, especially pragmatic linguistics. In task-oriented design activities, dialogues are said to be cooperative since the partners share a common goal: they have to converge towards agreement concerning a solution. That is why they differ from several other types of dialogues, such as political debates, interviews, chatting, where the aim is not primarily to collaborate towards a common outcome.

In the framework of the MOSAIC project, we have compared analysis methods adopted and results obtained by researchers from cognitive ergonomics and linguistics, the two disciplines collaborating in this project. In addition to the analysis of collaborative activities through people's dialogues, we have also examined their generation and use of external representations, and more specifically their graphico-gestural activities. A graphico-gestural coding has been proposed. Various approaches to analysis have been elaborated, compared and discussed in the framework of the preparation of a book, to be published in 2005 at the PUN (co-edited by F. Détienne & V. Traverso).

Methods for the analysis of error production

Participants: Pierre Falzon, Hélène Faye.

In the 1980's, the automotive industry has supplanted standardized production by custom manufacturing. The organization is, however, still very affected by Taylorism. Operators on assembly lines are confronted with two contradictory logics that oppose flexibility and rigidity. In spite of the repetitive aspect of work, operators make errors that, if not immediately recovered, generate defects that have to be fixed. The purpose of this study [44] is to define the mechanisms of error production, both in relation to work organization and to the nature of work activity. Therefore, a series of interviews will be performed with team leaders in order to formulate a global hypothesis about the causes of error production. Subsequently, a method will be developed enabling to identify the representations that operators have built of their activity.

Methods for the design of new emerging technologies

Participants: Margarita Anastassova, Jean-Marie Burkhardt.

We have continued the research carried out within the framework of an industrial co-operation between the French Atomic Energy Commission (CEA) and Renault S.A.S. On the one hand, we have focused our attention on automobile mechanics' training in the context of recent design evolutions in order to assess the applicability of Augmented Reality (AR) as tutor assistance. The research was based on interviews and video observations. It has shown that tutors' difficulties were mainly related to the explanation of the functioning of vehicle electronics, and to the lack of information provided by vehicle designers [15]. The user interface specifications of the future AR system will be performed using this information. On the other hand, within this research, we have compared interviews and task analysis as two techniques for eliciting end-user needs for emerging technologies. Interviews proved to be more useful for revealing and improving the acceptability of the future emerging technology, while task analysis proved to be particularly useful for revealing utility and usability aspects.


previous
next