Learning for usability: An explorative study of qualities in use

ABSTRACT

Efforts for creating usable systems which fulfill the purpose of being efficient and effective tools in an enterprise have been focused on the software itself. The study proposed here turns to the user, and to what the user contributes with for that use. The study explores the concepts of usability and qualities of software in use, and their relationship to end-users learning to use the software, in a case study approach. The understanding developed during this study will be used in an intervention study, which aims at proposing a way for formal training to contribute to usability and quality in use.

Keywords

Usability, end-user training, quality in use.

© 1997 Copyright on this material is held by the authors.

INTRODUCTION

Usability occurs in the interaction between human, computer, and a task. Huge efforts are put into the design of user interfaces, assuming they make systems usable. In recent years this concept has been extended [e.g. 2] and its assumptions has rendered critique. [e.g. 7]. Other efforts focus on the supplier, and software quality. Through powerful concepts, such as process and product quality and user involvement, usability is incorporated in the production of software [4, 5].

Both these strands of efforts aim at the tangible product. The joint mindset is the object, although there are exceptions. When purchasing a piece of software users learn to use the software, grows to become useful and usable. The role of learning for the usability of a piece of software remains largely unexplored. As software becomes more integrated in business processes, it becomes increasingly influential how it is used. The qualities of software in use and their relationship to issues of learning and usability also remain unexplored.

It might seem obvious that with users more competent in their software, the usability of the software would be greater than with less competent users. In HCI there is a tendency of trying to eliminate the effects of competence by presupposing that the user knows how to use the software. Learning turns into discussion of memory aids, shortcuts, or ways to make graphical design unambiguous [12]. These solutions to the learning problem presuppose a simple relationship between the work task within which the software is used and the software. Presupposed is also that users are conscious of this relationship, and that the software solves the right problem. It is reasonable not to take these presuppositions at face value, but to find alternate approaches and frameworks, such as design for quality in use [6], or activity theory [10].

In an introductory study of the effects of end-user instruction on subjectively perceived usability [9], scored with SUMI complemented with interviews, it was observed that scores of subjectively perceived usability becomes less homogenous with experience of use, and that users who have a preconception of a software based on what others told them rather than experience, finds the software more usable as they gain experience but scoring lower. Although the situation of use was suppressed in the design of the questionnaires, this still seemed to influence the result. In the actual situation of use there is a need for treating the relationship between work task and software in combination with at least a competency complex.

THE RESEARCH QUESTION

The study will explore usability and qualities of software in use, in relationship to learning. This limits the questions to

• what qualities in use can be observed and how does quality in use evolve over time and with learning?
• what effects of learning can be observed on usability with traditional measures?

This calls for an empirical inquiry of situations of use, a contemporary phenomenon within its real life context. A case study approach would therefore be useful [13].

THE CONDITIONS OF THE STUDY

Other approaches than the traditional HCI approach, focusing on technology development, are plausible; with the users, organization and technology given, it would be possible to change the tasks which are performed in order to achieve better. The perspective I choose concerns the users part in the situation of use, and specifically what effect users have on the usability of software. A grounded theory analysis [11] of interview material collected during the introductory study [9] has been performed. Through that analysis it has been understood that the meaningfulness of training, the different strategies for getting help, and different strategies of use have influence on the effect of training.

Learning experiences are difficult to control, and therefore hard to design experiments for. Still a formal training project provides the opportunity to say at least something about the learning experience and the content of learning. It also provides a real world setting and an existing need for the understanding of learning, thus facilitating the finding of an industrial partner and a setting in which the results may be applied.

The software for which the training is designed is a common off the shelf system, COTS. We know from the beginning that these systems will be used for disparate purposes and in disparate organizations with disparate users. With that kind of variation it will be possible to collect and study a rich qualitative material. Looking at a tailored system, or a system developed for a specific customer, there would be a risk of getting too narrow data, with many of the qualities in use already settled in specifications, or by other means presupposed. It makes more sense to perform such a study after the explorative. In order for a COTS to be useful a user needs to learn how to use the software in an effective and efficient manner. Several studies on learning has been performed within which strategies and influential factors has been proposed [1, 3, 8]. In the final analysis of the collected material these will be considered.

Exploring qualities in use

Qualities in use and learning are contemporary phenomena best studied in their real life context. All in all 22 users participating in a training project will be informants, split up into three groups. Two methods of data collection will be used; one rich qualitative, similar to contextual inquiry, and one traditional usability testing.

For every group of informants data will be collected at three points in time. The first occurs before the formal training begins, the second after the last formal training session, and the third three months after the last training session.

To build a deeper understanding of the collected material the instructors, the instructional designer, material and design, the informant's managers, and CIO:s will be interviewed.

EXPECTED OUTCOME AND FUTURE WORK

It is expected that this study will generate an understanding of how the use of software change with learning, and how qualities in use and usability evolve. This understanding will be used in a second intervention study, as a basis for the design of training aiming at affecting some of the qualities in use.

REFERENCES

1. Busch, T. (1993). Overføring av læring [Transfer of training]. Nordisk Pedagogik, 14, pp 87-100. In Norwegian
2. XCarlshamre, P. (1994). A collaborative approach to usability engineering: Technical communication and systems developers in usability-oriented systems development. (Thesis no 455), Dept. Computer and Information Science, Linköping University, Linköping, Sweden.
3. Carroll, J., M. (1990). The Nurnberg funnel: Designing minimalist instruction for practical computer skill. The MIT Press, Cambridge, MA.
4. Dahlbom, B., Mathiassen, L. (1994). Computers in context: The philosophy and practice of systems design. NCC Blackwell, Cambridge, MA.
5. Ehn, P. (1992). Scandinavian design: On participation and skill. In Adler, P., S. and Winograd, T., A. (eds). Usability: Turning technology into tools. New York, Oxford University Press, Inc.
6. Ehn, P., Löwgren, J. Design for quality-in-use: Human-computer interaction meets systems development. Forthcoming in Helander, M. (ed.) Handbook of human-computer interaction, 2nd ed.
7. Grudin, J. (1993). Interface: an evolving concept. Communications of the ACM, 36(4):110-120.
8. Holmberg, L., M. (1996). Design of a learning environment for users of an accounting information system: Theoretical foundations and an empirical study (Report No 1996:01). Göteborg University, Dept. Education and Educational Research, Göteborg, Sweden.
9. Holmlid, S. (1995). The effect of end-user instruction on usability: usability as a key quality of instruction. Master thesis, Dept. Computer and Information Science, Linköping University, Linköping, Sweden.
10. Nardi, B., A. (1996). Context and consciousness: Activity theory and human-computer interaction. The MIT Press, Cambridge, MA.
11. Strauss, A., L. (1987). Qualitative analysis for social scientists. Cambridge University Press, Cambridge.
12. Trumbly, J., E., Arnett, K., P., Martin, M., P. (1993). Performance effect on matching computer interface characteristics and user skill level. International Journal of Man-Machine Studies, 38:713-724.
13. Yin, R., K. (1984). Case study research design and methods. Sage, Newbury Park, CA, revised edition.