Short-term and long-term relationships between reflection and performance in teams: evidence from a four-wave longitudinal study

Abstract

Previous theory and research suggest that team reflection is beneficial for team performance. We argue that results remain inconclusive because prior studies have not accounted for the dynamic nature of this relationship. This paper addresses this research gap by examining time-specific relationships among variables and the intra-team variability of changes across time. In a four-wave longitudinal field study with 97 teams (N = 453 team members) performing a business simulation task, short-term (i.e., autoregressive, cross-lagged effects) and long-term (i.e., latent trajectories) relationships between team reflection and performance were explored. We found evidence that reflection had negative autoregressive effects and that there were direct positive short-term relationships between reflection and performance. Reflection trajectories were seen to decline across time and performance trajectories to increase across time. In addition, teams with either low initial reflection or low initial performance showed higher increases in reflection across time, whilst higher increase in reflection was negatively related to change in performance. Findings are discussed with respect to how they extend the previous literature and what directions they suggest for future research.

Keywords:

• reflexivity
• reflection
• performance
• teams
• time
• change
• trajectories
• autoregressive latent trajectory models

Acknowledgement

The authors greatly appreciate the assistance of Dr Corinna Steenfatt on this project.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Though West (2000) describes this as a measure of reflexivity, the criticism has been made that items represent typical reflection activities, including reviewing and discussion team’s methods and strategies (Konradt et al., 2016). Items for adaptation were left out. Thus, we used this measure to represent team reflection.

2. To avoid complexity, models were estimated with manifest variable indicators (mean scale scores). To ensure that our longitudinal ALT models based on manifest indicators may rely on stable reliability with stability/instability of the construct (Marsh et al., 2009) we tested for the assumption of measurement invariance. In this study, we found evidence of longitudinal measurement invariance for reflexivity and performance, indicating that our decision to rely on manifest indicators did not affect the results.

3. We do not interpret data on reflection and performance at Time 1, because no hypotheses were proposed.