Understanding and Measuring Facilitated Idea Convergence

Idea convergence is a phase of idea selection and describes a team process in which team members reduce the ideas to focus on the few they deem worthy of further attention and in which they clarify those ideas to increase shared understanding among team members. Idea convergence is particularly relevant in settings where a crowd generated abundances of ideas, e.g., innovation contests and where idea selection of the best ideas is thus far a big challenge for organizations. The reason is that cognitive resources and attention of team members is in high demand during idea convergence. Firstly, teams are challenged how to meaningfully facilitate their idea convergence. Secondly, teams are challenged with how to meaningfully process crowd-sourced ideas after idea generation. This results in poor convergence performance and teams often fail in selecting the best ideas. Human facilitation interventions and automated feedback help guide the attention of teams so that cognitive resources can be effectively applied to potentially good ideas. However, it turns out that the existing conceptualization of human facilitation interventions does not allow us to sufficiently explain their effects on convergence outcomes. What is required is to go beyond the typology of human facilitation interventions and to put more emphasis on eliciting the underlying theoretical components that drive convergence outcomes. Moreover, past research has examined several factors that can be understood as determinants of good ideas. These factors can be automatically and unobtrusively collected during idea generation but they have not yet been investigated as automated feedback for convergence teams. This project aims at aims at 1) eliciting effective human facilitation interventions and 2) designing effective automated feedback to guide attention and help manage cognitive resources of teams converging on crowd-generated ideas. For research, this project contributes to the measurement of determinants of good ideas with feedback/feedforward profiles. Moreover, by examining the needs for human facilitation interventions and by understanding their direct effect on convergence outcomes, results of this project add to the further development of theory. We will apply quantitative and qualitative methods for investigation. Implications of this research will be beneficial to multiple research domains including small group research, computer-supported collaborative work, or research on leadership. For practice, project results will provide teams with decision- support on which ideas they should consider for successful idea convergence. The project team consists of one post-doc (project lead Isabella Seeber), three experienced senior researchers (Gert-Jan de Vreede, Ronald Maier and Barbara Weber), three PhDs and a student assistant. Over the past three years, the project lead and the three senior researchers have continuously collaborated in idea convergence-related research activities resulting in several research publications. They are keen to further strengthen their scientific collaboration with this research grant proposal.

Product and service innovations start with a good idea. The idea generation and idea selection phases, which are also referred to as the front end of innovation is fuzzy and difficult to manage. People are usually very good at generating hundreds of ideas. However, selecting the best idea out of hundreds is a difficult task for raters. Past research showed that raters do not perform much better than a selection by chance. One widely accepted reason for this is that raters are easily overwhelmed by the amount of information and have limited cognitive resources to process information. The UMIC project aims at providing solutions to this idea selection challenge by identifying effective human and automated facilitation interventions. In the course of the project, we uncovered mechanisms that explain why human-performed facilitation techniques (i.e. thinkLets) lead to improved team processes and team outcomes during team-based idea selection. We also uncovered several selection practices that practitioners adopt for improved selection outcomes of ideas generated by the crowd. Moreover, we developed and evaluated multiple so-called digital nudges, which are user interface elements that aim at influencing human decision behavior. Our empirical findings show that the digital nudges (idea subset size, selection direction, and idea subset similarity) have positive and negative effects on selection processes and outcomes. To gain these insights, we also developed an idea allocation algorithm (SIA algorithm) and devised constructs to measure idea subset elaborateness and information acquisition behavior. We also gained a first understanding about conversational agents (e.g., social robots or chatbots) as facilitators. The findings of this research help to improve team facilitation and digitally support idea selection process for improved innovation.