A comprehensive model for analysing collaborative activity

Position paper for Workshop 5: 'Analyzing Collaborative Activity'

CSCW2002 Conference, Nov. 2002, New Orleans

J.H.Erik Andriessen

DelftUniversity of Technology

Faculty of Technology, Policy and Management

Section Psychology of Work and Organisation

Jaffalaan 5

2628 BX Delft, The Netherlands

tel. +31 15 2781742 / 2783720

e-mail:

Introduction

Analysing collaborative activities – and the role of groupware in support of these activities - can be approached from two sides. One is bottom up, i.e. through grounded interpretation of cases study material, the other is top down, i.e. through using a model based on previous research and on existing theory. We have developed such a model called the Dynamic Group Interaction model (DGIn-model) and honed it through application in several cases studies. The purpose of this model is NOT to limit the analysis of collaborative activities to certain quantifiable data, but to structure the analysis by providing ideas and insights that have proven their value and thereby preventing the re-invention of the wheel.

In this model elements of four grand theories, Action Theory, Activity Theory, Structuration Theory and Adaptive Structuration Theory and of several other, particularly group dynamic notions are brought together. In this way three levels of behaviour analysis are taken into account, i.e. individual goal directed behaviour and cognitive processes (Action Theory, Activity Theory), interpersonal and group processes (Activity Theory, Adaptive Structuration Theory, Group Dynamics) and macro-social processes (Structuration Theory). Each of the four grand theories is quite rich but also quite general. Other theories are needed to distinguish specific factors and processes, to explain certain phenomena and to identify relevant individual, group and organisational characteristics. Some of these other theories are also more suitable for deriving design guidelines. Together they point at the potential relevance of a large number of factors.

The various notions are brought together in a heuristic model concerning group processes. which is related to traditional input-process-output schema's. The field of social and organisational psychology has produced several sophisticated versions (McGrath, 1984; Hackman, 1987). In overviews of the findings in this field the basic model has been “filled” with many variables (e.g. Kraemer and Pinsonneault, 1990; McGrath and Hollingshead, 1994).

Many of these models are limited. They suggest a linear and one-directional causal relation: the characteristics of the input determine the processes which then determine the outcomes. It is clear, however, from experience and reflected in theories such as (Adaptive) Structuration Theory or group stage models, that human action and group interaction change the social structures, which means that they have feedback effects on group characteristics, task definition and tool use. These feedback effects can be explicitly planned, but unforeseen events will often result in the unplanned emergence of new settings and processes. For this reason the term context-of-useis adopted in this book, instead of the term input factors, which suggests to indicate the starting moment only.

Behind some of the models lies the contingency-idea, that the context-of-use factors must fit, i.e. must match to produce the best processes and output. The models are in this sense quite static in nature. However, several theories point to the fact that characteristics of tasks, tools and context are perceived and subjectively re-interpreted. Moreover, learning and adaptation processes take place through which input elements are constantly adapted to each other. Nevertheless one cannot count on an automatic adaptation process when introducing a new tool or a new setting. The larger the gap between existing and new settings the more energy has to be directed to the introduction processes.

Many models focus on the level of group task interaction and pay less attention to aspects of individual human computer interaction and task performance or to organisational functioning. As has been argued earlier, co-operative group work contains much individual task performance, and computer mediated group work implies individual tool use and issues of human computer interaction. Cognitive theories point to the fact that when designing and evaluating ICT-tools from a user perspective, one has to take into account the question of whether they fit the action mechanisms, the human motivation and the self-image of the potential users.

The model presented in the following section tries to take into account these criticisms. The models are a combination of the input-process-output model, with elements of Activity Theory, Adaptive Structuration Theory and several group dynamic theories combined.

The Co-operative Work model

Collaborative activities can be analysed without any reference to technology. The CSCW context however focuses attention on ICT supported collaboration. The purpose of introducing collaboration technology is to support the interaction of several people interacting in a certain way. Introduction of such systems however implies changes in ways of working, in the use of tools and in group interaction. TheDGIn model (see Figure 1) is based on these principles and on the integration of group dynamic and other theories.

Figure 1. The Dynamic Group Interaction Model (DGIn -model) (Andriessen, 2002)

The model has the following basic tenets:

  • The effectiveness of a group – in terms of three outcome categories - depends on the quality of five group processes: communication, which is a bases for cooperation, co-ordination, learning and social interaction . These five processes have to fit to each other.
  • The effectiveness of these processes in their turn depend on the extent to which six types of conditions (see figure) support them and on the interaction with the environment. These six aspects have to fit to each other.
  • The effectiveness of a (groupware) tool depends therefore on:
  1. the characteristics of the tool
  2. the fit to the other conditions
  3. the extent to which it in practice supports (or at least does not hinder) the group processes and the attainment of the outcomes
  • Groups develop and tools become adopted and adapted to, through interaction processes and feedback.

The DGIn model is constructed along three dimensions.

Dimension 1.Three levels of interaction: individual, group and organisation

Dimension 2.Context-of-use (input) – processes – output

Dimension 3.Feedback, appropriation and reflection

Both the interaction processes and the outcomes result in feedback processes, by which the original context-of-use factors are changed. Groups build cohesion and trust, shared knowledge and new task definitions through interaction. Tools will be appropriated and adapted, new ways of interaction are developed. This feedback takes place directly as a result of the interaction processes, or indirectly via the route of outcomes and their effect on organisational processes. They can emerge unplanned, develop slowly in daily practice, or they can be the result of explicit reflection of the group on its functioning and explicit change.

A case study

This model allows a systematic analysis of cases of collaborative interaction. From the model a checklist of 'attentionpoints' has been derived. This list contains items from each of the elements and relations in the model. When applying this to a certain case, the general list has to be adapted to the specific situation, and can than be used as a basis for drawing up interviews, observations and other data collections methods. For certain aspects calibrated questionnaires are available. This refers for instance to team functioning dimensions such as trust, cohesion or commitment, or to tool usability dimensions such as ease of use or learnability.

This approach is modular in the sense that one can choose to limit a study to certain aspects. The model has been used in various studies, recently in studies of Communities of Practice. It provides a rich picture of issues related to community functioning and tool use.

References

Andriessen, J.H.Erik (2002). Working with Groupware. Understanding and Evaluating Collaboration Technology. London: Springer Verlag.

McGrath, J. E. (1984). Groups: interaction and performance. New Jersey: Prentice-Hall.

Hackman, J. R. (1987). The design of work teams. In J.W.Lorsch (Ed.), Handbook of organizational behaviour (pp. p315-342). Englewood Cliffs, NJ: Prentice Hall.

Kraemer, K. L., & Pinsonneault, A. (1990). Technology and groups: Assessment of the empirical research. In J. Galegher, R. E. Kraut & C. Egido (Eds.), Intellectual Teamwork: Social and technological foundations of co-operative work. Hillsdale, NJ: Lawrence Erlbaum Publ

McGrath, A. J. E., & Hollingshead, A. B. (1994). Groups Interacting with Technology. Ideas, Evidence, Issues and an Agenda. London: Sage Publications.

1