Authors:
Sarah Harvey – University College London
Chia-Yu Kou – University College Dublin

Interviewers:
Beth Devine – INSEAD
Amer Madi – INSEAD

Article link: http://asq.sagepub.com/content/58/3/109

Question 1. What motivated or inspired you to pursue this research question? Was it due to the inconsistencies that you noticed between creativity research at the individual and organizational levels versus the group level? Or did you set out exploring a different question altogether?

Our own personal experiences primed our interest in this topic – we have both long been fascinated by how teams work effectively (and not so effectively!) on novel tasks, and how knowledge and ideas are shared and combined by collectives. In this study, we initially set out to find a way to more systematically describe the creative process at the group level. When we began the study, so much of the team creativity research was experimental, and assumed that the process looked and worked like the individual creative process. We wanted to investigate whether that was true. From our own experiences, we had a hunch that it was not. For example, Sarah used to work in strategy consulting, and her most creative experiences on a consulting team happened in the moment, when a team meeting was really smooth and flowing, not during the kind of debate and discussion of different perspectives that is often described in the team creativity literature. Our intuition was that the ‘heart’ of the group creative process described in the literature didn’t always match people’s experiences.

It was only when we started examining the literature more broadly to help us understand the data and emerging insights that we discovered many accounts of collective creativity that better captured our experiences. Those accounts were often contained in studies at the organizational or individual level, and they didn’t always refer explicitly to the group creative process, at least as we conceptualized it from our perspective within the team creativity literature. So we did not initially set out to explore a difference between individual, organizational, and group literatures, but rather discovered through the research process that those literatures could provide new insights into group creativity.

Question 2. The premise of this article is to challenge the present-day view of groups’ creative idea evaluation. As early-career academics, were you concerned about potential pushback on your novel theory?

We were certainly concerned about push back on the idea that evaluation early on in the creative process could be beneficial. As we acknowledge in the paper, in the wrong kind of group – such as one with poor interpersonal relations or a negative evaluation environment – that process could be problematic. We were therefore concerned that people may focus on those limitations and dangers of the ideas rather than those instances where the alternative process can work.

A critical turning point in the research, however, was discovering the right audience for the paper. As we dug into the literature, it also became clear that these ideas already resonated with many academics who had found similar processes in their data but studied them through a different lens. Also, there has been a real burgeoning of the creativity literature over the past few years, and our perception is that increasingly, creativity researchers are hungry for interesting ideas to advance that area of research. Once we felt like we could identify groups of researchers who would appreciate our ideas, it gave us the confidence to pursue the paper.

Question 3. Traditional theory of group creative process involves early brainstorming followed by convergent evaluation of ideas, which you refer to as a generation-centered sequence. Your study revealed a reverse process of evaluation-centered sequence, characterized by initial collective evaluation of ideas and later-stage divergent brainstorming. Your findings suggest that an evaluation-centered approach may enhance creativity compared to a generation-centered one by allowing space for more idea elaboration and integration. What advice would you give managers in organizations wanting to shift from generation focus to evaluation focus?

Our advice would be to start the creative process by discussing one or two prototype ideas in some detail. They could be ideas that a group has generated, but they need not be – they could be the latest products from competitors, an idea that their own organization has previously tried, or an idea imported from another industry or geographic location (e.g., what if we launched product X in this market?). As IDEO and others have shown us, prototypes do not need to be correct to be useful. Discussing the idea in detail means identifying what the team likes and does not like about it, why it would and wouldn’t work, and what goals it will satisfy. We expect that process to help teams uncover their hidden assumptions and implicit goals, in a way that is more effective than discussing those issues in a more abstract way. Once team members’ understand one another’s perspectives in more detail, then move on to generating ideas. We expect that process to be most effective in diverse teams whose members have not worked together for a long period of time, and who have not experienced significant interpersonal problems.

Question 4. In your findings, groups HR and PD followed the traditional generation-centered sequence, whereas groups CD and ES followed the evaluation-centered sequence. What do you think accounts for this difference? In other words, why do you think some groups follow a generation-centered sequence and others an evaluation-centered one? We noticed that groups HR and PD had higher proportions of medical experts than groups CD and ES. Do you think that was a factor?

This is a great question that we would love to explore in subsequent research. We do not really know from our data why a group followed one trajectory or another, but we can speculate that your observations about group composition are on the right track. In particular, we view the group composition as a possible trigger for the beginning of one process of the other. In the teams with a higher percentage of medical experts, there was a lower degree of diversity. Members of those teams may have had many ideas – and therefore started off with brainstorming – but perhaps they also understood one another’s ideas and agreed with the goals that the ideas implicitly satisfied without much discussion, because they shared a similar background. In contrast, when team members had very different backgrounds, they may have immediately questioned or challenged an idea and its implied goals, either because they didn’t understand it well enough or because they held a different perspective on what the group should do. That questioning and challenging would likely provoke a more in depth discussion and evaluation of the idea, and may also stimulate contrasting ideas in other team members as they tried to explain a key point. Thus, the more diverse composition may have triggered the early parallel process that observed, because differences were much closer to the surface in those groups.

Question 5. Your research setting was four U.S. healthcare policy groups charged with generating policy recommendations for the U.S. Department of Health and Human Services. Given that much inductive, qualitative research is done via special permission or even conditions of anonymity, it is notable that the data — meeting transcripts and audio recordings — were publicly available! How did you come to find this data? Are there other public sources of qualitative data you know of and are willing to share?

Sarah searched out this data after being inspired by communications scholars she saw presenting at a conference. Communications researchers often seem to find unique and interesting but publicly available datasets. We would suggest reading communications journals and talking to communications researchers as one good starting point. Also, government agencies often have requirements for making their communications transparent and available, so that may be another good starting point. Specialized groups

The data was really exciting for us, because we were interested in tracking how ideas developed, so we needed to access team interactions, but were less concerned about team members’ perceptions and feelings. However, this type of data can limit the type of research questions that can be asked – for example, we had no opportunity to access participants’ views on the creative process. We did not view that as a problem (and in many ways it was an advantage) for our research, but it would not suit every type of research question.