Options
Regular patterns in the group process: how they are detected, what they tell us, and how they are related to performance
Auteur(s)
Nägele Stalder, Christof
Editeur(s)
Tschan Semmer, Franziska
Date de parution
2004
Résumé
Teams have the potential to perform much better than many individual does. However, often teams do not reach their full potential. It is often reported that teams use inefficient strategies to accomplish a task. It is also reported that teams are often very resistant to the (spontaneous) development of new and better strategies. It is important to understand the team process. In small group research, the interaction process is a hot topic. Almost in all perceptions on small groups, the process is implicitly or explicitly mentioned (Poole et al., 2004). But if one looks at how 'process' is defined and operationalized in empirical studies a great diversity becomes apparent. Three types of measures on group processes were distinguished: (i) evaluative measures on the group process (typically questionnaire based data, where team members are asked to evaluate the group process), (ii) frequencies of behaviors (coding and counting approaches), and (iii) measures reflecting temporal sequences in the group process. It is argued that evaluative measures and summaries of frequencies cannot represent all aspects of the group process and that an analysis of the temporal micro-behavioral organization of the group process offers additional information. This information can then be used for an effective coaching of co-acting teams. Data of 109 teams of three persons were analyzed. An air-traffic control (ATC) simulation was used. The teams had to observe an airspace and keep track of the changing positions of the airplanes. Every single act was coded such that its type and its meaning were captured together with its time stamp. Sequential analyses were run to gain gaining insight into the group process. Three methods are discussed and applied to the ATC-data: (i) lag sequential analysis, (ii) procedural network representation, (iii) data mining techniques. These methods are never (data mining) or seldom used in small group research. Results show that either of these methods can be used to analyze the group process on a micro-behavioral level. They can be used as exploratory tools and aids to visualize the group process. Several examples are shown. All these sequential patterns must be evaluated in the context of the concrete task. This is perfect for team coaching or the development or refinement of categories developed in a top down process (like for example the task adaptive behaviors in Tschan, Semmer, Nägele et al., 2000). There are also concrete recommendations to commanders of co-acting virtual teams: Use the information from the specialists, read what they are writing. Do not send task or strategy related messages if the message content is too old. Nevertheless, commanders should not stay too long in a mode of just reading messages. Measures form sequential analyses were used in regression models to predict the team performance. This was done separately for every of the seven work-shifts of at least fifteen minutes. Results show that using this information on the sequential patterns in regression models explains additional variance in performance per shift in the range from 16% to 35% (additionally to input factors, preceding performance and frequencies of behaviors). The information extracted with sequential analyses is relevant to performance. To improve the group process in co-acting teams detailed analyses of the temporal sequences on a micro-behavioral level is indispensable. It was shown that lag sequential analysis, procedural network representations (PRONET), and data mining techniques can be used. It is also demonstrated how this methods can be applied to the ATC-data.
Notes
Thèse de doctorat : Université de Neuchâtel, 2004 ; 1782
Identifiants
Type de publication
doctoral thesis
