The Role of Time and Category Relatedness in Electronic Brainstorming

Abstract

This study tested creativity of small electronic brainstorming (EBS) groups as a function of the homogeneity and heterogeneity of assigned sub-topics (categories) of a broader problem over time. A total of 168 participants were exposed to categories of high or low degree of relatedness, and their performance was tracked over time in group and e-nominal paradigm. Findings revealed that the EBS groups became more creative and exhibited slower productivity loss compared with the e-nominal groups over time. The assignment of homogeneous categories enhanced the average originality of ideas, over time especially for the EBS groups. The findings have implications for theoretical as well as practical perspectives for creative processes in small virtual brainstorming groups.

Keywords

brainstorming group creativity idea generation virtual groups computer-mediated communication

An efficient way of sharing ideas is electronic brainstorming (EBS; Dzindolet, Paulus, & Glazer, 2012; Paulus, 2015). The EBS paradigm is a form of virtual communication in which ideas are generated using electronic media such as computer, smart phones, and teleconferencing. Several studies have found an increase in creativity of ideas over time in EBS paradigm (e.g., Beaty & Silvia, 2012; Kohn & Smith, 2011), but the results have been mixed. Although Beaty and Silvia (2012) reported a significant increase in creativity over time, Kohn and Smith (2011) did not find a statistically significant effect for the same by groups over time. Deuja, Kohn, Paulus, and Korde (2014) found an increase in originality of ideas over time in such groups, but only for those who were assigned with categories sequentially over time. We are aware of only one study that compared the performance of EBS with e-nominal groups (members generating ideas in isolation electronically), and this study found no difference in the pattern between those conditions (Kohn & Smith, 2011). Although some researchers (Brown & Paulus, 2002; Rietzschel, Nijstad, & Stroebe, 2007) proposed that more creative and less accessible ideas will tend to be generated toward the end of an idea generation session, very few studies (e.g., Kohn & Smith, 2011) have demonstrated such findings over time in the small group context. Thus, this issue merits further examination. With increased time, groups should be more likely to show enhanced performance in terms of number of ideas and novelty of ideas relative to e-nominals.

One factor that has been found to influence the creativity of the ideas is how the task is structured in terms of focus on categories of ideas (Baruah & Paulus, 2011; Coskun, Paulus, Brown, & Sherwood, 2000; Deuja et al., 2014). Research on the influence of task focus on performance has so far demonstrated that consideration of one topic or category at a time (Coskun et al., 2000), or priming individuals and dyads with a subcategory of a broader problem (Rietzschel et al., 2007) led to more and better ideas (Rietzschel, Nijstad, & Stroebe, 2014) and more in-depth exploration of ideas (Deuja et al., 2014) than directly assigning the group with the broad problem. However, so far, no study has examined the effects of category focus over time.

Theoretically, the cognitive models of brainstorming (Nijstad & Stroebe, 2006; Paulus & Brown, 2007) posit that when an idea is activated in memory, ideas semantically related to the activated idea will get activated first. The higher activation of relevant knowledge structures will lead to the generation of more related ideas within the assigned category. However, as the related or homogeneous ideas get depleted over time, more unique ideas are generated toward the end of the brainstorming session. Building on this perspective, we predict that if the groups are assigned with highly related categories, initially, they should rapidly generate ideas within the assigned categories as these domain-related topics would result in faster activation of related ideas. Thus, exploring deeper into a limited set of highly related (homogeneous) categories will result in more original ideas over time. In addition, mutual exchange and building on each other’s ideas in group context should enhance the originality even more over time.

We examined these ideas in an experiment in which participants generated ideas electronically either alone or in groups and were primed with either high or low related categories. Thus, it is expected that the performance of EBS interactive and e-nominal groups should show a different pattern over time in terms of quality and quantity of ideas generated. We hypothesize that the (H1) interactive EBS groups would generate more and better ideas over time compared with e-nominal groups, and that (H2) providing groups with highly related categories will increase the originality of ideas over time especially for EBS groups.

Method

A total of 168 undergraduate students participated in groups of three for a total of 56 groups. The age range was 18 to 37, and the average age was 23.3 years. The study was a 2 (high/low related categories) × 2 (EBS/e-nominal) × 3 (Time 1/Time 2/Time 3) mixed factorial design. Time was repeated measures, whereas group type and category relatedness were between-group factors. Participants were randomly assigned to either e-nominal or interactive EBS groups of three. Although all members were to engage in brainstorming on the same broad problem of “how to improve your university” in both EBS and e-nominal groups, each member received a single category related to the broad problem. Each of the three participants in each group was randomly assigned with one of three highly related or one of three low related categories. High and low related categories were derived from Baruah and Paulus (2011). In the EBS condition, all members were anonymously logged into Yahoo Messenger group conference tool (similar to Group Decision Support Systems; cf. Baruah & Paulus, 2011). These group members could view and share each other’s ideas on the screen. Participants in the e-nominal condition could type in their ideas but could only view their ideas and not others’ ideas. All members in the group knew that they were working toward solving a common problem, but each member was assigned one category related to the broader problem.

The dependent variables were number of ideas, rated for originality on a 5-point Likert-type scale (intraclass correlation = .88), and within-category fluency (total number of ideas generated within the assigned categories).

Past research has shown that the ideas decline rapidly after the first few minutes (5 to 6 min) of the brainstorming session (e.g., Kohn & Smith, 2011) resulting in a performance level that is rather low, typically less than one idea a minute. Hence, we divided the 15-min brainstorming session into three 5-min time quadrants: Time 1 (first 5 min), Time 2 (second 5 min), and Time 3 (third 5 min), respectively.

Results

Manipulation Check

As a check on the manipulation of category relatedness, participants were asked, “How different were the types of ideas you generated on your topic from those generated by others on their topic?” A one-way ANOVA revealed a significant difference between the responses of the groups assigned with highly related and low related categories, F(1, 33) = 7.52, p < .01, η_p² = .19. Groups assigned with low related categories reported the topics to be more different from others (M = 5.07) compared with groups assigned with highly related categories (M = 3.95).

Main Effects of Time and Category Relatedness

A 2 (groups) × 2 (relatedness) × 3 (time) mixed factorial ANOVA was performed for the total number of overall ideas, within-category ideas, and the average originality of ideas generated. The three-way ANOVA revealed a significant main effect of time on total number of ideas, F(2, 104) = 177.89, p < .005, η_p² = .77, and within-category fluency, F(1.68, 87.30) = 90.68, p < .001, η_p² = .64, of ideas generated. The post hoc Bonferroni test on quantity of ideas revealed that there was a significant difference between Time 1 and Time 2 (M = 6.97, p < .001), and Time 1 and Time 3 (M = 7.47, p < .001), but no significant difference was found between Time 2 and Time 3. The Bonferroni test on within-category fluency of ideas revealed that there was a significant difference between Time 1 and Time 2 (M = 7.57, p < 001), Time 1 and Time 3 (M = 9.29, p < .001), and Time 2 and Time 3 (M = 1.71, p < .05). Participants in all conditions generated more ideas and exhibited higher within-category fluency during Time 1 compared with Time 3 (see Table 1). However, there was no main effect of time on the average originality of ideas generated.

Table 1.

Relatedness of Assigned Topic on Quantity, Originality, and Within-Category Fluency of Ideas Generated Over Time.

Dependent measures			Time 1	Time 2	Time 3
Dependent measures	Relatedness	Groups	M (SE)	M (SE)	M (SE)
Quantity	High	e-nominal	24.86 (1.80)	15.29 (1.63)	10.79 (1.39)
		EBS	24.50 (1.80)	16.38 (1.63)	15.50 (1.40)
	Low	e-nominal	29.00 (1.80)	19.07 (1.63)	13.64 (1.39)
		EBS	24.29 (1.80)	15.86 (1.63)	15.64 (1.39)
Originality	High	e-nominal	2.99 (0.05)	3.04 (0.08)	2.84 (0.10)
		EBS	2.87 (0.05)	3.02 (0.08)	3.21 (0.10)
	Low	e-nominal	2.90 (0.05)	3.00 (0.08)	2.91 (0.10)
		EBS	2.99 (0.05)	2.97 (0.08)	3.07 (0.10)
Within-category fluency	High	e-nominal	16.71 (1.80)	8.86 (1.29)	5.50 (1.08)
		EBS	16.93 (1.80)	9.43 (1.29)	8.43 (1.08)
	Low	e-nominal	15.43 (1.80)	7.86 (1.29)	5.79 (1.08)
		EBS	13.29 (1.80)	5.93 (1.29)	5.50 (1.08)

Note. EBS = electronic brainstorming.

There was a marginally significant effect of category relatedness in terms of within-category fluency, F(1, 52) = 3.10, p = .08, η_p² = .06, with highly related (homogeneous) categories yielding higher within-category fluency than low related categories (see Table 1). However, there was no effect of category relatedness on quantity and originality of ideas generated.

Interaction of Time and Group Type

The ANOVA revealed a significant interaction of time and group type in terms of quantity, F(2, 104) = 11.05, p < .005, η_p² = .175. The e-nominal groups generated more ideas than EBS groups during Time 1, but EBS groups generated more ideas than e-nominal during Time 3 (see Figure 1), consistent with H1.

Figure 1.

Quantity of ideas as a function of time and group type.

The ANOVA revealed a significant interaction of time and group type in terms of average originality of ideas generated, F(1.79, 93.19) = 5.14, p < .05, η_p² = .09. E-nominal groups generated ideas of higher average originality than EBS groups during Time 1, but EBS groups outperformed the e-nominal groups during Time 3 (see Figure 2).

Figure 2.

Average originality of ideas as a function of time and group type.

Time, Group Type, and Category Relatedness

We found a marginally significant three-way interaction of time, group type, and category relatedness, F(1, 52) = 3.47, p = .07, η_p² = .06, in terms of originality of ideas generated. The EBS groups generated ideas of highest average originality under the highly related category condition at Time 3 (see Figure 3).

Figure 3.

Average originality of ideas as a function of time, category relatedness, and group type.

Because we found a marginally significant three-way interaction, we conducted separate two-way ANOVAs for the EBS and e-nominal conditions to examine the simple main effects. In the e-nominal group condition, the quantity and within-category fluency decreased over time, but there was no effect of originality. Also, there was no significant interaction of time and category relatedness. However, in the case of EBS groups, quantity and within-category fluency decreased over time whereas originality of ideas increased over time. The originality of ideas was found to be highest during Time 3 under highly related category condition (a significant interaction of time and category relatedness), F(1, 26) = 3.99, p = .05, η_p² = .13, in the EBS group paradigm. H2 was supported.

Discussion

The results were consistent with our predictions. The assignment of highly related categories to groups enhanced the average originality of ideas over time especially for the EBS groups. The e-nominal groups exhibited a rapid reduction of productivity over time compared with the EBS groups, and the originality of ideas increased over time for EBS but not e-nominal groups.

A key finding is that the exposure to homogeneous categories enhanced the originality of ideas over time among EBS groups. Getting exposed to similar ideas from others might have helped groups to connect the ideas better (Dugosh, Paulus, Roland, & Yang, 2000). Our findings further revealed that the within-category fluency of ideas was not correlated with originality in the beginning of the session (Time 1), but was positively correlated toward the end of the session (Time 3), r = .35, p < .01. This suggests the creation of common and unoriginal ideas in the beginning of the session (e.g., Larey & Paulus, 1999) and novel connection through deeper exploration over time toward the end of the session. Thus, exploring deeper within the category, the interactive EBS groups might have come up with more unique associations, which was not possible in e-nominal groups due to the lack of interaction and synergy.

Past studies have typically measured performance in group brainstorming tasks (in electronic as well as face-to-face groups) by aggregating ideas of the overall brainstorming session, and thus, little information is available on the fluctuation of performance over time. Although there was no main effect of time on the originality of ideas generated in our study, the temporal analyses revealed that interactive groups generated ideas of significantly higher average originality at the end of the session. Thus, the current study demonstrates that an investigation of fluctuation of quality and quantity of ideas over time provides a more accurate perspective of the effectiveness of small EBS groups. Although the overall mean difference in performance between EBS and e-nominal groups may reflect superiority in e-nominal groups’ performance over interactive groups, a detailed temporal analysis reveals that this difference is not consistent over time. Thus, the search for creative synergy in groups needs to take the time factor into account. Demonstrations of such synergy have been rare (cf. Baruah & Paulus, 2009) possibly in part because research studies have not examined performance patterns over time. Future studies should examine fluctuation of creativity over a longer period of time to determine whether this pattern will continue and how various task features affect this. Future studies should also examine fluctuation of creativity over time for large EBS groups. It is possible that the positive effects of originality over time will be evident earlier in large groups compared with small EBS groups.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The preparation of this paper was supported by a collaborative grant to the second author from the National Science Foundation (INSPIRE BCS 1247971). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Author Biographies

Jonali Baruah is an assistant professor in the Department of Psychological Sciences at Tarleton State University, USA. She maintains a research interest that incorporates social, cognitive, and diversity aspects in the investigation of creativity and performance in small groups and individuals.

Paul B. Paulus is a distinguished professor in the Department of Psychology at the University of Texas at Arlington, USA. For the past 25 years, his research has focused on the factors that influence group and team creativity. He has served as chair of the Department of Psychology and dean of the College of Science.

References

Baruah

Paulus

P. B.

(2009). Enhancing creativity in groups: The search for synergy. In Neale

Mannix

Goncalo

(Eds.), Research on managing groups and teams (pp. 29-56). Oxford, UK: Elsevier Science Press. doi:10.1108/S1534-0856(2009)0000012005

Baruah

Paulus

P. B.

(2011). Category assignment and relatedness in the group ideation process. Journal of Experimental Social Psychology, 47, 1070-1077. doi:10.1016/j.jesp.2011.04.007

Beaty

R. E.

Silvia

P. J.

(2012). Why do ideas get more creative over time? An executive interpretation of the serial order effect in divergent think tasks. Psychology of Aesthetics Creativity, 6, 309-319. doi:10.1037/a0029171

Brown

V. R.

Paulus

P. B.

(2002). Making group brainstorming more effective: Recommendations from an associative memory perspective. Current Directions in Psychological Science, 11, 208-212. doi:10.1111/1467-8721.00202

Coskun

Paulus

P. B.

Brown

Sherwood

J. J.

(2000). Cognitive stimulation and problem presentation in idea generation groups. Group Dynamics: Theory, Research, and Practice, 4, 307-329. doi:10.1037/1089-2699.4.4.307

Deuja

Kohn

N. W.

Paulus

P. B.

Korde

(2014). Taking a broad perspective before brainstorming. Group Dynamics: Theory, Research, and Practice, 18, 222-236. doi:10.1037/gdn0000008

Dugosh

K. L.

Paulus

P. B.

Roland

E. J.

Yang

(2000). Cognitive stimulation in brainstorming. Journal of Personality and Social Psychology, 79, 722-735. doi:10.1037//0022-35I4.79.5.722

Dzindolet

T. M.

Paulus

P. B.

Glazer

(2012). Brainstorming in virtual teams. In Silva

C. N.

(Ed.), Online research methods in urban and planning studies: Design and outcome (pp. 138-156). Hershey, PA: IGI Global. doi:10.4018/978-1-4666-0074-4

Kohn

N. W.

Smith

S. M.

(2011). Collaborative fixation: Effects of others’ ideas on brainstorming. Applied Cognitive Psychology, 25, 359-371. doi:10.1002/acp.1699

10.

Larey

T. S.

Paulus

P. B.

(1999). Group preference and convergent tendencies in small groups: A content analysis of group brainstorming performance. Creativity Research Journal, 12, 175-184. doi:10.1207/s15326934crj1203_2

11.

Nijstad

B. A.

Stroebe

(2006). How the group affects the mind: A cognitive model of idea generation in groups. Personality and Social Psychology Review, 10, 186-213. doi:10.1207/s15327957pspr1003_1

12.

Paulus

P. B.

(2015). Electronic brainstorming research and its implications for e-planning. International Journal of E-Planning Research, 4, 42-53. doi:10.4018/ijepr.2015010103

13.

Paulus

P. B.

Brown

V. R.

(2007). Toward more creative and innovative group idea generation: A cognitive-social motivational perspective of brainstorming. Social & Personality Compass, 1, 248-265. doi:10.1111/j.1751-9004.2007.00006.x

14.

Rietzschel

E. F.

Nijstad

B. A.

Stroebe

(2007). Relative accessibility of domain specific knowledge and creativity: The effects of knowledge activation on the quantity and originality of generated ideas. Journal of Experimental Social Psychology, 43, 933-946. doi:10.1016/j.jesp.2006.10.014

15.

Rietzschel

E. F.

Nijstad

B. A.

Stroebe

(2014). Effects of problem scope and creativity instructions on idea generation and selection. Creative Research Journal, 6, 185-191. doi:10.1080/10400419.2014.901084