Table 3.

Comparison of CI applications across c-factor, synergy as well as process-oriented view.

	C-Factor	Synergy	Process-Oriented CI
Definition	CI is defined as the general ability of the team to perform well across a variety of cognitive tasks. In the context of multidisciplinary health care teams, the underlying premise is that if the team collectively scores high on a generic collective ability test, it is collectively intelligent and hence can ‘transfer’ it’s intelligence to other contexts as well.	CI is defined as whether the team outperforms the (best) individual team member. In the context of multidisciplinary health care teams, a team would be collectively intelligent if it jointly makes better decisions regarding treatment plans for patients compared to when one physician comes up with a treatment plan individually.	Intelligence is defined as an unfolding process of collective behaviors (content, rhythm, participation) that originate in individual level behavioral acts, that are appropriate for the tasks that are assigned to the team and in alignment with the environmental needs in which the team operates. In the context of multidisciplinary health care teams, the team would be intelligent if the content of the conversation, the way in which they discuss, as well as who participates is appropriate and effective to solve the task the team is working on and is also in line with changing environmental needs. One important change in the environment in multidisciplinary health care teams is that patient cases vary in terms of complexity. In low complexity patient case discussions, a more fast-paced, standardized process with fewer people contributing to the discussion is often considered as an intelligent way of organizing (by medical experts). However, in complex patient case discussions, a low pace, with input from varied medical experts, combined with actively questioning one another is generally considered as intelligent behavior. In sum, a relevant/salient changing environmental need (i.e., complexity/rareness of disease) requires different ways of organizing interactional structure and thus the team needs to be adaptive towards changing environmental needs. Collective intelligence is now considered high for medical teams that easily shift between discussion formats as they move from case to case, whereas less intelligent medical teams would be more stuck to a single way of discussing, regardless of the complexity of each specific case.
Operationalization	Collective intelligence is measured by giving teams various cognitive tasks (e.g., spatial reasoning, mathematical, linguistics tasks). Factor analytic approaches are used to identify one latent underlying ability factor reflecting the team’s intelligence. The higher the performance across cognitive tasks, the higher the team’s intelligence and hence the better the team is expected to come up with suitable treatment plans, now and in the future. One assumption is that the performance of the team can be measured correctly and objectively.	Each physician would be asked to come up with a treatment plan for patients individually. Subsequently, the team would be asked to collectively come up with a treatment plan, following group interaction. The health care team would be evaluated as intelligent if the team comes up with a ‘better’ or ‘more suitable’ treatment plan for the patient, compared to the physician making the best decision individually. One underlying assumption is that individuals can do the task, so a team is not necessary. Also, it is assumed that it is possible to objectively judge which treatment plan is “best.”	In the process-oriented CI approach the researcher analyzes transcripts of who says/does what at what point in time during the medical team meeting and investigates how medical expertise is shared across patient cases. First, the evaluation must be made whether the content is aligned with the needs of the patient and whether sufficient information and relevant medical expertise is communicated within the team. Second, it is evaluated whether the experts speaking up are also the ones that would be expected to contribute given the background/complexity of the patient case. Lastly, the researcher would look at the rhythm or pace of the decision-making process; do team members follow logical sequences of decision making? Or is the conversation totally scattered? Is the conversational pace efficient and clear for members to follow? Each of these features must be evaluated in context to reflect on the intelligence of the behaviors that take place within the team. Having interpreted the appropriateness of the interaction process for each patient case, the researcher then assesses to what extent the team was able to adopt fitting discussion procedures over the course of the entire meeting (so across all patient cases/tasks that the team had to formulate a solution for). The better the team adjusted its discussion format to the requirements of the specific case, the more intelligent it was. Then, if it is possible to collect such data over multiple meetings, it can be assessed whether collective intelligent teams are indeed able to display the required procedural flexibility in later meetings as well.