The Moderating Effect of Extraversion–Introversion Differences on Group Idea Generation Performance

Abstract

Previous research provides evidence that individual differences in the personality characteristic of extraversion/introversion can play a significant role in group idea generation. Cognitive stimulation has also been shown to have a significant, though inconsistent, influence on idea generation in computer-mediated groups. We conducted two controlled experiments using a web-based group simulator to empirically test the impacts of these factors on group idea generation. In the first experiment, extraverts outperformed introverts in computer-mediated groups. In the second experiment, we exposed participants in computer-mediated groups to four levels of idea stimulation ranging from none to extremely high. Extraverts generated more unique and diverse ideas than did introverts in moderate- and high-stimuli conditions only. In sum, our study provides empirical evidence regarding the significant effect of individual differences, in this case the personality characteristic of extraversion/introversion on idea generation in a computer-mediated group. We have also demonstrated that a relationship exists between the level of cognitive stimulation and idea generation, with moderate levels of idea stimulation optimal for extraverts in computer-mediated groups. Theoretical and practical implications of these findings are discussed.

Keywords

brainstorming computer-mediated communication performance creativity

Introduction

Over the past half century, substantial research attention across disciplines has focused on identifying and developing better techniques to enhance the effectiveness of idea generation (Girotra, Terwiesch, & Ulrich, 2010). A variety of idea-generation techniques have been introduced, including face-to-face brainstorming, nominal groups, and computer-mediated brainstorming (e.g., VanGundy, 1992). Empirical research to date has consistently found that both computer-mediated groups (CMGs) and nominal groups (NGs) generate more ideas than did face-to-face brainstorming groups (Gallupe et al., 1992). In spite of rich structural features (e.g., parallel input, group memory, anonymity) presumed to provide CMG a performance advantage over NG, performance differences between CMGs and NGs have not always been demonstrated (e.g., Pinsonneault, Barki, Gallupe, & Hoppen, 1999a, 1999b). For example, Valacich, Dennis, and Connolly (1994) determined that groups using a computer-based idea-generation system outperformed equivalent NGs in idea-generating task. Conversely, some researchers have found that nominal brainstorming groups generated ideas at least as good as, if not better than, electronic brainstorming groups (Barki & Pinsonneault, 2001). In sum, a quite thorough examination of a wide variety of factors affecting group idea generation has been done but the superiority of CMG over NG is yet to be established (Pinsonneault et al., 1999a, 1999b). Our review of the relevant research suggested that two factors, the individual personality characteristic of extraversion/introversion and the level of cognitive stimulation in CMG, may account for some of these inconsistent findings and warrant additional study (Nijstad, Stroebe, & Lodewijkx, 2002; Yellen, Winniford, & Sanford, 1995).

Personality trait researchers (e.g., Engle, 2002) have postulated that individuals differ in their attention levels and information-processing capacities. Of particular interest to our research, extraversion–introversion differences appear to have a significant influence on divergent thinking and creativity. For example, Kasof (1997) pointed out that extraverts have chronically wider attention breadth than do introverts. Martindale (1999) suggested that creative people tend to exhibit low levels of cortical arousal and frontal lobe activation during the process of creative thinking and extraverts demonstrated a lower level of arousal and activation than did introverts. Aguilar-Alonso (1996) further reported that extraverts produce more original, fluent, and flexible outcomes, which have been identified as main indicators of divergent thinking. Finally, Yellen et al. (1995) found that the personality characteristic of extraversion/introversion plays a significant role in affecting the level of member participation in electronically supported meetings. Consequently, we propose that the existing research into extraversion–introversion differences suggests it may play a significant role in group idea generation.

A substantial body of cognitive process research has also identified cognitive stimulation, perceived to be one of the main benefits of CMG, as a potential source of performance inconsistency (Dennis & Williams, 2003). Unlike NG, where participants work alone and independently, participants in CMG receive cognitive stimulation by reading the ideas of other group members. Several studies have suggested that exposure to others’ ideas can improve idea-generation performance (e.g., Coskun, Paulus, Brown, & Sherwood, 2000; Dugosh, Paulus, Roland, & Yang, 2000; Nijstad et al., 2002). However, findings on this topic are also somewhat mixed. Some researchers have argued that the larger the pool of ideas, the greater the likelihood of stimulating ideas (Osborn, 1957) because diverse stimulation increases the logical size of an idea pool (Valacich, Wheeler, Mennecke, & Wachter, 1995) as well as the breadth of idea production (Nijstad, Diehl, & Stroebe, 2003). For example, Dugosh et al. (2000) found that idea-generation performance was increased when a high number of ideas were presented. However, other researchers have indicated that a high volume of ideas can result in detrimental side effects (e.g., Nagasundaram & Dennis, 1993). For instance, Girotra et al. (2010) indicated that “building on others’ ideas (i.e., cognitive stimulation in interactive groups) is counterproductive . . . such buildup neither create more ideas, nor are the ideas that build on previous ideas better” (p. 591). Thus, the relationship between the level of idea stimulation and performance in CMGs remains somewhat unclear (e.g., Paulus, Larey, & Ortega, 1995).

In sum, previous research provides evidence that the personality characteristic of extraversion/introversion can play a significant role in groups, influencing an individual’s attention span, level of arousal, and participation in groups (Yellen et al., 1995). Cognitive stimulation has been shown to have a significant, though inconsistent, influence on idea generation in CMG. This body of research provided the impetus for us to conduct two controlled experiments. In the first experiment, we investigate whether personality differences (extravert vs. introvert) significantly influence idea-generation performance in CMG. In the second experiment, we investigate the presence of an interaction effect between differences in the individual characteristic of extraversion/introversion and the level of idea stimulation in CMGs. As done in previous studies (e.g., Nijstad et al., 2003), we measured idea-generation performance using the number of unique ideas and that of diverse ideas. We believe the results of the experiments offer additional insight into the group performance inconsistencies discussed above.

Theoretical Framework

Individual Differences and Idea-Generation Performance

Eysenck (1997) notes that extraversion/introversion has long been identified as a major dimension of personality. Significant differences between extraverts and introverts have been found in a variety of contexts (e.g., Eysenck, 1997; Kasof, 1997; Yellen et al., 1995). Of relevance to our study, researchers have identified extravert/introvert differences in attention span (Kasof, 1997), attention focus (e.g., Eysenck, 1982), and working memory (e.g., Elliott, 1972).

First, researchers have posited that extraverts tend to have a broader attention span and more actively seek external stimulation than do introverts, who are generally more narrowly focused and avoid external stimulation (Eysenck, 1982). The broader attention of extraverts has been directly related to their superior performance in creative idea generation compared with introverts. For example, Kasof (1997) hypothesized that “creative ability is related to chronically wide breadth of attention” (p. 304) and found that breadth of attention (i.e., the number and range of environmental stimuli attended to at any one time) correlates positively with creativity and performance of novel idea generation. Similarly, Dykes and McGhie (1976) suggested that less creative persons seem to exhibit more narrowly focused attention than do more creative people. Based on the literature, we expect extraverts to produce more ideas as well as more diverse ideas in CMG, where external stimulation is provided, than in the NG, where it is not.

Second, researchers (e.g., Zuckerman, 1991) have found that extraverts and introverts differ in attention focus, especially in high-arousal settings. For example, extraverts performed better than did introverts under high-arousal conditions when a large number of idea stimuli are presented. Research indicates that the cortex of introverts is chronically more aroused than that of extraverts (Eysenck, 1997). As a result, introverts tend to inhibit their attention and attend to a narrower range of cue utilization, whereas extraverts are more likely to expand their range of cue utilization. Under high-arousal conditions, “introverts reported a tendency to make mistakes and become confused, whereas extraverts perceived themselves as being able to integrate many stimuli effectively and to process a great deal of information” (Eysenck, 1982, p. 128). Extrapolating these findings to our research context may provide additional insight into why cognitive stimulation in CMG has been shown to have a significant, though inconsistent, influence on idea generation (Dennis & Williams, 2003). Extraverts may find that access to others’ ideas stimulates idea generation, whereas introverts may find that stimulus ideas interfere with their train of thought. In turn, shorter trains of thought have been associated with a loss of ideas (Nijstad et al., 2002).

Third, prior research also indicates that differences in working memory utilization in stimulating environments similar to that of CMG may be source of performance differences among group members. For example, extraverts have a larger working memory for storing, searching, retrieving, and reproducing information than do introverts (Saint-Aubin, Ouellette, & Poirier, 2005). Extraverts also have been shown to comprehend written words and sentences by selecting and assimilating stimuli more effectively than do introverts (Riding & McQuaid, 1987). Furthermore, a significant positive correlation between reading achievement and extraversion was found (Elliott, 1972). Given that stimulus ideas in CMG are generally presented as text, larger working memory and superior reading comprehension would seem likely to give extraverts another performance advantage over introverts in CMG.

Fourth, an individual’s breadth of attention—the number and range of stimuli attended to at any one time—also correlates positively with creativity and the generation of novel ideas (Brown & Paulus, 1996; Martindale, 1999). Likewise, previous studies (e.g., Dykes & McGhie, 1976) have suggested that less creative persons exhibit more narrowly focused attention than do more creative people. They have also found that extraverts are more flexible in dealing with semantically diverse stimulation and tend to have better memory and greater recall than are introverts (e.g., Lieberman & Rosenthal, 2001). Moreover, extraverts appear more able to search for ideas in associative memory with less interruption (Nijstad et al., 2003).

In sum, based on our review of the relevant literature, we expect extraverts to be more comfortable and cognitively capable of processing a wider range of cues in arousal conditions (e.g., CMG) than do introverts. Therefore, we hypothesize that extraverts will outperform introverts in group idea generation in CMG. Specifically,

Hypothesis 1a: Extraverts will generate a significantly greater number of unique ideas than introverts in CMG.

Hypothesis 1b: Extraverts will generate a significantly greater number of diverse ideas than introverts in CMG.

Individual Differences, Number of Idea Stimuli, and Idea-Generation Performance

Though extraverts appear more likely than introverts to respond positively to cognitive stimulation in CMGs, research suggests there is a level at which cognitive stimulation becomes cognitive interference (Nijstad et al., 2002). According to the Yerkes-Dodson law (Yerkes & Dodson, 1908) there exists a general relationship between arousal and performance, with performance tending to increase as the arousal or noise level increases up to an optimal threshold, after which further increases in arousal tend to diminish performance. Research by Matthews, Deary, and Whiteman (2003) affirmed Yerkes-Dodson, noting that persons with high sensory thresholds (i.e., extraverts who are chronically low in arousal) perform better in stimulating environments than do persons with a low sensory threshold (i.e., introverts who are chronically high in arousal) because they are less likely to be over aroused. However, the superior performance of extraverts only holds as long as environmental conditions, such as the number of idea stimuli, are not too stressful. For example, Eysenck (1982) found no significant performance differences between extraverts and introverts in no arousal conditions or in extreme arousal conditions. Applying the Yerkes-Dodson law and related research findings to the CMG context, we expect extraverts will outperform introverts only when germane cognitive stimulation conditions (i.e., moderate- and high-stimuli conditions) exist. Stated conversely, we expect no performance differences between extraverts and introverts in the absence of stimuli or in extremely high-stimuli conditions. Thus, we hypothesize:

Hypothesis 2a: Extraverts will generate significantly greater numbers of unique ideas than do introverts under moderate- to high-stimuli conditions in CMG.

Hypothesis 2b: Extraverts will generate significantly more diverse ideas than do introverts under moderate- to high-stimuli conditions in CMG.

Study 1

Study 1 tests Hypotheses 1a and 1b. We designed the study to determine if extraverts perform better for idea generation in CMG.

Research Design

A total of 342 upper-level business students from a large state university in the United States initially volunteered to participate in the study. Participants were directed to a secure website and asked to complete Francis, Brown, and Philipchalk’s (1992) six-item Extraversion–Introversion measure (e.g., “Are you a talkative person?” “Can you easily get some life into a rather dull party?”; response scale: yes/no). To create as large a difference in personality as possible and ensure a sufficient number of introverts, only those participants who scored 6 were recruited as extraverts, whereas those scoring 0 to 2 were recruited as introverts. This method is consistent with prior studies (e.g., Topi, Valacich, & Rao, 2002). A total of 75 target participants (38 extraverts and 37 introverts) were selected for the study. As compensation, they received course credit corresponding to less than 1% of their overall grade. The average participant age was 25.1 years (SD = 7.49), and 52% of participants were men.

A 2 × 2 factorial design was used, crossing brainstorming groups (nominal¹ and computer-mediated) and individual differences (introverts and extraverts). Although the influence of brainstorming group (i.e., NG vs. CMG) was not hypothesized, we included it to see whether the significant interaction between brainstorming group type and individual differences is present, which provides us a richer understanding of its impact as a valuable moderator in group idea generation. Participants identified as extraverts or introverts were randomly assigned to one of four treatment conditions. A web-based group simulator was developed using the Visual Basic programming language. A group simulator is an electronic environment that “looks and acts like a groupware system, but instead of sharing ideas among participants, the simulator presents participants with comments that appear to be from other participants but which are, in fact, drawn from a database of preset ideas” (Garfield, Taylor, Dennis, & Satzinger, 2001, p. 327).

The group simulator was used to control the presentation of ideas and minimize error variance, which inevitably occurs with interacting groups (Garfield et al., 2001), and to yield a more accurate and controlled measure of individual-level idea generation. The simulator in this study closely mimicked the sequence of a real, interacting, group idea-generation session, such that the idea seeds (including the participant’s own ideas) appear sequentially, on a first-come-first-served basis.

As in previous studies (e.g., Hilmer & Dennis, 2001), the group simulator does not present the stimuli at identical time intervals. Rather, a mathematical mechanism created pooled, preset ideas from a database to simulate the downward, curvilinear relationship that typically emerges for the numbers of ideas generated over time by real, interacting groups in idea-generation sessions (Brown & Paulus, 1996). This relationship depicts many ideas in the early stage and fewer responses toward the later stages and then runs out of ideas in the end. This pattern of idea presentation was controlled by programming within the group simulator.

Pilot testing confirms that the group simulator accurately reproduced the sequence and interactions of a real, interacting, group idea-generation session. During the experimental sessions, which simulated groups with five members, each respondent responded to a postsession question: “How many people do you think you were working with on this task?” On average, participants reported working with 4.76 group members (SD = 1.03). Thus, it appears that participants believed they were working in a real, interacting group.

Procedures

Participants were asked to generate ideas on “How can we improve the university’s parking problem?” This task was chosen for its high relevance to the participants because it stimulates participants to draw on their personal knowledge and experience and because it has been used successfully in many studies (e.g., Garfield et al., 2001).

Participants in the CMG conditions were instructed that they would work randomly with other group members who were remotely located, using a web-based group simulator that would allow them to observe others’ ideas. Participants assigned to the CMG conditions were asked to demonstrate sufficient proficiency interacting with the simulator by completing a small task before the main task. They also received additional instruction (i.e., a version of Osborn’s [1957] brainstorming rules) and were asked to follow it. The rules directed participants to generate as many ideas as possible, withhold criticism, include wild ideas, and build on the ideas of others. Each participant’s contributions and idea seeds from the database were anonymous.

Participants in the NG condition were told that they would work independently and were asked to generate as many ideas as possible. All participants were also told their contributions would help improve the campus parking problem. The group simulator was programmed to stop after 10 min. After that, the participants completed a brief questionnaire and were released.

Dependent Variables

Previous studies provide the basis for the selection of the two dependent variables, number of unique ideas and diverse ideas (e.g., Nijstad et al., 2003). The former measure indicates the efficiency with which people produce quality ideas; the latter measures their effectiveness in producing diverse ideas within given cognitive stimulations and time constraints.

To identify the number of unique ideas generated, one coder first analyzed all comments captured by the group simulator. A methodology similar to previous studies was used to avoid duplicate comments (Nijstad et al., 2003): If the participants’ ideas were unique and presented before the stimulus ideas, they were counted. If ideas are the same or very similar to the stimulus ideas and they were presented after the stimulus, they were not counted. Consistent with prior studies, a second coder then independently analyzed a random subset of transcripts from 19 participants to confirm the initial coder’s categorization. Cronbach’s interrater reliability value of .925 indicated the coding was highly consistent.

To determine the number of diverse ideas generated, two university parking experts separately categorized all participants’ ideas, resulting in 25 different semantic categories. The Cronbach’s interrater reliability value of .897 indicated the diversity coding was also highly consistent.

Results

Table 1 displays the means and standard deviations for the four groups. Table 2 provides summarizes ANOVA results and post hoc Tukey test analysis. As shown in Table 2, ANOVA indicated there were significant differences across groups for both the number of unique ideas (F = 4.974, p < .01) and for the number of diverse ideas (F = 3.455, p < .05) generated. The Tukey test was then applied to determine the source of the significant differences.

Table 1.

Means and Standard Deviations for Brainstorming Group-Individual Difference-Ideation Performance

	Brainstorming group
Dependent variable	CMG	NG
Number of unique ideas
Extravert
M	6.00	5.00
SD	2.16	2.98
Introvert
M	3.81	3.90
SD	2.05	2.13
Number of diverse ideas
Extravert
M	5.00	4.80
SD	1.65	2.94
Introvert
M	3.48	3.60
SD	1.72	2.06

Note: CMG = computer-mediated group; NG = nominal group.

Table 2.

Summary of ANOVA and Tukey Test Results

Dependent variable	SS	df	MS	F	Significance
Number of unique ideas	75.026	3	25.01	4.974	0.003
Number of diverse ideas	38.939	3	12.98	3.455	0.021
Dependent variable	IV (I)	IV (J)	MD (I-J)	SE	Significance
Number of unique ideas	Extravert-CMG	Introvert-NG	2.1	0.83	0.062
	Extravert-CMG	Extravert-NG	1	0.83	0.622
	Extravert-CMG	Introvert-CMG	2.19	0.60	0.003
Number of diverse ideas	Extravert-CMG	Introvert-NG	1.4	0.71	0.213
	Extravert-CMG	Extravert-NG	0.2	0.71	0.992
	Extravert-CMG	Introvert-CMG	1.52	0.52	0.025

Note: CMG = computer-mediated group; NG = nominal group.

The results of the tests of Hypotheses 1a and 1b are presented in Tables 1 and 2. As Hypothesis 1a predicted, extravert-CMG participants (M = 6.00, SD = 2.16) did produce a greater number of unique ideas than did introvert-CMG participants (M = 3.81, SD = 2.16). The difference between extravert-CMG and introvert-CMG participants was significant (mean difference = 2.19, p < .01), supporting Hypothesis 1a.

As Hypothesis 1b predicted, extravert-CMG participants also produced a greater number of diverse ideas (M = 5.00, SD = 1.65) than did introvert-CMG participants (M = 3.48, SD = 1.72). The difference between extravert-CMG and introvert-CMG participants was significant (mean difference = 1.52, p < .05), implying support for Hypothesis 1b.²

In sum, extravert-CMG participants produced significantly more unique and diverse ideas than did introvert-CMG participants.

Study 2

Study 2 was designed to test Hypothesis 2a and 2b and focused solely on participants in CMGs. We designed the study to determine if there is an interaction between individual differences in extraversion/introversion and varying levels of stimulation in CMG that produces significant difference in the number of unique and diverse ideas generated.

Research Design

The 2 × 4 factorial design was used which crosses the extraverts and introverts with four levels of idea stimuli described in more detail in the next section: none (0), moderate (20), high (40), and extremely high (80). The same participants that had been identified as extraverts or introverts in Study 1 were randomly assigned to one of four treatment conditions. The web-based group simulator was again used to present ideas and measure performance. The task, setting, and procedures followed were exactly like those employed in Study 1. As in Study 1, the number of unique ideas generated and the number of diverse ideas generated served as the dependent variables.

Independent Variables

Prior studies have commonly employed the number of contextual cues (i.e., ideas) as a determinant of information overload (Grisé & Gallupe, 2000). We based our four categories of stimulus ideas and length of presentation on Grisé and Gallupe’s (2000) information overload model. Their research suggests 20 ideas are sufficient to induce moderate information overload when presented in a 10-min period, 40 ideas a high information load, and 80 ideas an extremely high information load. No (0) ideas served as a baseline.

To create the idea streams for the three load conditions, an independent coder reviewed a master list of 457 ideas created by a university in the United States. The coder identified 80 geography-neutral, high-quality ideas. Three senior parking experts also rated the quality of the ideas on the master list on a 7-point Likert-type scale. As in previous studies (e.g., Valacich et al., 1994), ideas with an average rating of 4 or higher on a 7-point Likert-type scale were selected as high-quality ideas (M = 4.69, SD = .93). A Cronbach’s value of .92 confirmed high overall interrater reliability.

The group simulator then presented the three randomized idea streams, with 20, 40, and 80 ideas each, to participants in a way that mimicked the downward, curvilinear pattern of idea generation that typically emerges for the numbers of ideas generated over time by real, interacting groups (Brown & Paulus, 1996). Participants in each treatment condition received the stimuli in the same order. A postsession question asked each participant to indicate the number of people they thought they were working with on this task. On average, participants reported working with 3.27 group members (SD = 1.10) in the moderate-stimuli condition, 4.38 group members (SD = 1.66) in the high-stimuli condition, and 8.45 group members (SD = 2.81) in the extremely high-stimuli condition. As in Study 1, it appears that participants believed that they were working in a real, interacting group.

Results

Table 3 displays the means and standard deviations for the number of unique and diverse ideas generated by participants in each condition. Table 4 summarizes the results of a series one-way ANOVA tests we conducted to compare the performance extraverts and introverts in each of the four stimuli conditions and found that extraverts performed significantly better in the moderate- and high-arousal treatments than did introverts.

Table 3.

Means and Standard Deviations for Level of Idea Stimuli-Individual Difference-Idea Generation Performance

	Level of idea stimuli
Dependent variable	No (0)	Moderate (20)	High (40)	Extremely high (80)
Number of unique ideas
Extravert
M	5.00	7.44	5.54	5.00
SD	2.98	2.19	1.51	2.27
Introvert
M	3.90	3.44	3.80	4.25
SD	2.13	2.07	2.09	2.19
Number of diverse ideas
Extravert
M	4.80	5.89	4.64	4.50
SD	2.93	1.45	1.12	2.20
Introvert
M	3.60	3.33	3.40	3.75
SD	2.06	1.87	1.50	1.98

Table 4.

Summary of One-Way ANOVA Test Results

Idea stimuli	Number of unique ideas (Means)	df	MS	F	Significance
No	Extravert (5.00), Introvert (3.90)	1	3.314	0.482	0.496
Moderate	Extravert (7.44), Introvert (3.44)	1	72	15.902	0.001
High	Extravert (5.54), Introvert (3.80)	1	15.958	4.865	0.04
Extremely high	Extravert (5.00), Introvert (4.25)	1	2.25	0.453	0.512
Idea stimuli	Number of diverse ideas (Means)	df	MS	F	Significance
No	Extravert (4.80), Introvert (3.60)	1	4.655	0.707	0.412
Moderate	Extravert (5.89), Introvert (3.33)	1	29.389	10.475	0.005
High	Extravert (4.64), Introvert (3.40)	1	8.007	4.618	0.045
Extremely high	Extravert (4.50), Introvert (3.75)	1	2.25	0.512	0.486

Tables 3 and 4 demonstrate that extraverts did produce more unique ideas (M = 7.44, SD = 2.19) than did introverts (M = 3.44, SD = 2.07) under the moderate-stimuli condition, and a difference that was significant (F = 15.902, p < .01). Extraverts produced more unique ideas (M = 5.54, SD = 1.51) than did introverts (M = 3.80, SD = 2.09) under high-stimuli conditions as well, a difference that was also statistically significant (F = 4.865, p < .05). No significant differences between extraverts and introverts were found under no stimuli (F = .482, p > .05) or extremely high-stimuli conditions (F = .453, p > .05).

Similar results were found for the generation of diverse ideas. Extraverts produced more diverse ideas (M = 5.89, SD = 1.45) than did introverts (M = 3.33, SD = 1.87) under the moderate stimuli condition, a difference that was significant (F = 10.475, p < .01). Similarly, extraverts produced more diverse ideas (M = 4.64, SD = 1.12) than did introverts (M = 3.40, SD = 1.50) under the high-stimuli condition as well, a difference that was again significant (F = 4.618, p < .05). No significant differences between extraverts and introverts were found under no stimuli (F = .707, p > .05), or extremely high-stimuli conditions (F = .512, p > .05).

Overall, the results provide support for Hypothesis 2a and 2b. Extraverts generated a greater number of unique and diverse ideas under moderate- and high-stimuli conditions than did introverts.

Discussion

Novel idea generation using brainstorming techniques in small group settings is quite pervasive. Small group researchers (e.g., Barki & Pinsonneault, 2001) have been conducted to examine the idea-generation performance differences across brainstorming techniques and those between CMGs with different level of idea stimuli (e.g., contextual cues). Interestingly, no consistent results have been accumulated. Thus, researchers have requested studies to find the ways to reconcile the discrepancies. As an effort to react the request, this study adopts extraversion–introversion individual differences as a factor to address an interesting but controversial phenomenon of group idea generation and confirms its significant moderating effect.

Individual Differences and Idea-Generation Performance

The result of Study 1 verifies that individual differences indeed work as a significant moderating factor for the group idea-generation performance in CMG. In particular, we found that the extravert-CMG group performed significantly better than the introvert-CMG group. Although more validation should be warranted, the finding of this study indicates that CMG seems a more productive technique for extraverts.

No significant performance difference between extraverts in computer-mediated (extravert-CMG) and nominal groups (extravert-NG) is a surprising finding. There are two possible interpretations for the insignificant results. First, the insignificant results may be because NG participants used a computer (i.e., the web-based simulator) to record their ideas, rather than the paper-and-pencil traditionally used in NGs. Although previous studies (e.g., Valacich et al., 1994) found no significant performance difference between these methods, the use of a computer in the NG condition may have made it more convenient and faster to record ideas in the time allowed. One possible future research is to conduct a replicate study with traditional paper-and-pencil-based NG and to compare its performance with that of CMG and that of computer-based NG.

Another factor that may explain the lack of differences between extraverts in CMG and NG may have been the exclusive use of high-quality stimuli in the extravert-CMG condition in Study 1. The ideas generated in actual CMGs typically include a mix of low- and high-quality ideas, as well as task-irrelevant comments. In contrast, this study only provided participants with high-quality ideas. Perhaps we established a performance norm that encouraged study participants to perceive other group members as highly proficient. Thus, participants may have felt that at least some of their ideas did not meet the apparent high-quality norm, resulting in the production of a smaller number of ideas. It is recommended to conduct future studies that include low- and high-quality idea stimuli to examine the performance differences under differently mixed idea stimuli conditions.

Individual Differences, Number of Idea Stimuli, and Idea-Generation Performance

The results of Study 2 provide evidence that individual differences and cognitive stimulation in CMGs can interact and have a significant impact on group idea generation. We found that extraverts produced significantly more unique and diverse than did introverts in moderate-stimuli (20 ideas) to high-stimuli (40 ideas) conditions. However, there were no significant performance differences between them in no stimuli (0 ideas) or extreme high-stimuli (80 ideas) conditions. The results are in concert with the curvilinear pattern predicted by the Yerkes-Dodson (1908) model: The performance of extraverts increases as the level of stimuli increases up to an optimal threshold (20 ideas in this study) and then decreases. The results confirmed our expectation that extraverts produce more unique and diverse ideas than do introverts in CMG but only under moderate- and high-stimuli conditions. The results are consistent with those of previous studies (Bouchard, 1969; Dennis & Valacich, 1999; Diehl & Stroebe, 1987).

The curvilinear pattern and optimal threshold of idea stimuli for extraverts that we identified suggests additional examination of the interval levels of idea stimuli presented in CMG. Future studies that use more granular stimuli intervals (e.g., 0, 5, and 10) may provide more understanding of the impact of the different levels of idea stimuli on idea-generation performance of introverts and extraverts.

Correspondingly, we suggest that levels of idea stimuli be studied in conjunction with the time allotted for idea generation. Our study allowed 10 min in all conditions. Prior research indicates that the time provided for idea generation has varying effects on the number of ideas people produce (e.g., Valacich, Jung, & Looney, 2006). Coupled with the curvilinear trend in idea generation found in our study, we expect that the optimal number of idea stimuli will change across different periods of ideation time. By examining the relationship between time pressure and levels of idea stimuli, we may be able to identify different threshold values of each. Such findings would seem of particular interest to small group researchers and practitioners who wish to provide the optimal level of idea stimuli in CMG sessions of varying length.

In summary, this study confirms that individual differences can be a significant moderator of idea generation between CMGs and NGs. It also validates that individual differences successfully moderate the relationship between the level of idea stimulation and idea generation in CMGs.

Limitations

This study does have some limitations that must be acknowledged. First, it employed a laboratory experiment, with student participants and simulation, which is not a natural group setting. Although a laboratory experiment is generally accepted to test the theory with precision and to use the simulation for increasing precision by controlling for error variances, it is recommended to conduct future studies in real-group settings. Second, the small sample size is another limitation of this study. Although recruiting small samples was inevitable because of our strict rules to select the participants with different personality traits, future replicate studies with a larger sample to ensure a higher statistical power are recommended. Finally, this study focused only on the direct relationship between idea stimulation and performance. Other mediating factors, such as arousal levels and memory capacity, should be assessed in future studies.

Conclusion

Focusing on an introversion–extraversion individual difference, this study confirms significant performance differences between extravert-CMG and introvert-CMG regarding producing unique and diverse ideas. It also finds that extraverts perform better than introverts when the number of idea stimuli is moderate and high, not low and extreme. Although further research is necessary, this study provides valuable empirical evidence that addresses the inconsistencies of idea-generation performance across different brainstorming groups and between CMGs with different level of idea stimuli.

Footnotes

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

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Aguilar-Alonso

(1996). Personality and creativity. Personality and Individual Differences, 21, 959-969.

Barki

Pinsonneault

(2001). Small group brainstorming and idea quality: Is electronic brainstorming the most effective approach? Small Group Research, 32, 158-205. doi:10.1177/10464964010320020310.1177/104649640103200203

Bouchard

T. J.

(1969). Personality, problem-solving procedure, and performance in small groups. Journal of Applied Psychology, 53, 1-29. doi:10.1037/h002674710.1037/h0026747

Brown

Paulus

P. B.

(1996). A simple dynamic model of social factors in group brainstorming. Small Group Research, 27, 91-114. doi:10.1177/104649649627100510.1177/1046496496271005

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:10.1037/1089-2699.4.4.30710.1037/1089-2699.4.4.307

Dennis

A. R.

Valacich

J. S.

(1999). Electronic brainstorming: Illusions and patterns of productivity. Information Systems Research, 10, 375-377. doi:10.1287/isre.10.4.37510.1287/isre.10.4.375

Dennis

A. R.

Williams

M. L.

(2003). Electronic brainstorming: Theory, research and future directions. In Paulus

P. B.

Nijstad

(Eds.), Group creativity: Innovation through collaboration (pp. 160-178). New York, NY: Oxford University Press.

Diehl

Stroebe

(1987). Productivity loss in brainstorming groups: Toward the solution of riddle. Journal of Personality and Social Psychology, 53, 497-509. doi:10.1037//0022-3514.53.3.49710.1037//0022-3514.53.3.497

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:10.1037/0022-3514.79.5.72210.1037/0022-3514.79.5.722

10.

Dykes

McGhie

(1976). A comparative study of attentional strategies of schizophrenic and highly creative normal subjects. British Journal of Psychiatry, 128, 50-56. doi:10.1192/bjp.128.1.5010.1192/bjp.128.1.50

11.

Elliott

C. D.

(1972). Personality factors and scholastic attainment. British Journal of Educational Psychology, 42, 23-32.

12.

Engle

R. W.

(2002). Working memory capacity as executive attention. Current Directions in Psychological Science, 11, 19-23. doi:10.1111/1467-8721.0016010.1111/1467-8721.00160

13.

Eysenck

H. J.

(1997). Personality and experimental psychology: The unification of psychology and the possibility of a paradigm. Journal of Personality and Social Psychology, 73, 1224-1237. doi:10.1037//0022-3514.73.6.122410.1037//0022-3514.73.6.1224

14.

Eysenck

M. W.

(1982). Attention and arousal. New York, NY: Springer.

15.

Francis

L. J.

Brown

L. B.

Philipchalk

(1992). The development of an abbreviated form of the revised Eysenck personality questionnaire (EPQR-A): Its use among students in England, Canada, the U.S.A., and Australia. Personality and Individual Differences, 13, 443-449. doi:10.1016/S0191-8869(96)00159-610.1016/S0191-8869(96)00159-6

16.

Gallupe

R. B.

Dennis

A. R.

Cooper

W. H.

Valacich

J. S.

Bastianutti

L. M.

Nunamaker

J. F.

(1992). Electronic brainstorming and group size. Academy of Management Journal, 35, 350-369. doi:10.2307/25637710.2307/256377

17.

Garfield

M. J.

Taylor

N. J.

Dennis

Satzinger

J. W.

(2001). Research report: Modifying paradigms: Individual differences, creativity, techniques, and exposure to ideas in group idea generation. Information Systems Research, 12, 322-333. doi:10.1287/isre.12.3.322.971010.1287/isre.12.3.322.9710

18.

Girotra

Terwiesch

Ulrich

K. T.

(2010). Idea generation and the quality of the best idea. Management Science, 56, 591-605. doi:10.1287/mnsc.1090.114410.1287/mnsc.1090.1144

19.

Grisé

Gallupe

(2000). Information overload: Addressing the productivity paradox in face-to-face electronic meetings. Journal of Management Information Systems, 16, 157-185.

20.

Hilmer

K. M.

Dennis

A. R.

(2001). Stimulating thinking: Cultivating better decisions with groupware through categorization. Journal of Management Information Systems, 17, 93-114.

21.

Kasof

(1997). Creativity and breadth of attention. Creativity Research Journal, 10, 303-315. doi:10.1207/s15326934crj1004_210.1207/s15326934crj1004_2

22.

Lieberman

M. D.

Rosenthal

(2001). Why Introverts Can’t Always Tell Who Likes Them: Multitasking and Nonverbal Decoding. Journal of Personality and Social Psychology, 80, 294-310.

23.

Martindale

(1999). Biological bases of creativity. In Sternberg

R. J.

(Ed.), Handbook of creativity (pp. 137-152). Cambridge, UK: Cambridge University Press.

24.

Matthews

Deary

I. J.

Whiteman

M. C.

(2003). Personality traits (2nd ed.). Cambridge, UK: Cambridge University Press.

25.

Nagasundaram

Dennis

A. R.

(1993). When a group is not a group: The cognitive foundation of group idea generation. Small Group Research, 24, 463-489. doi:10.1177/104649649324400310.1177/1046496493244003

26.

Nijstad

B. A.

Diehl

Stroebe

(2003). Cognitive stimulation and interference in idea-generating groups. In Paulus

P. B.

Nijstad

B. A.

(Eds.), Group creativity: Innovation through collaboration (pp. 110-136). New York, NY: Oxford University Press. doi:10.1093/acprof:oso/9780195147308.003.000710.1093/acprof:oso/9780195147308.003.0007

27.

Nijstad

B. A.

Stroebe

Lodewijkx

H. F. M.

(2002). Cognitive stimulation and interference in groups: Exposure effects in an idea generation task. Journal of Experimental Social Psychology, 38, 535-544. doi:10.1016/S0022-1031(02)00500-010.1016/S0022-1031(02)00500-0

28.

Osborn

A. F.

(1957). Applied imagination (Rev. ed.). New York, NY: Scribner.

29.

Paulus

P. B.

Larey

T. S.

Ortega

A. H.

(1995). Performance and perceptions of brainstormers in an organizational setting. Basic and Applied Social Psychology, 17, 249-265. doi:10.1207/s15324834basp1701&2_1510.1207/s15324834basp1701&2_15

30.

Pinsonneault

Barki

Gallupe

R. B.

Hoppen

(1999a). Electronic brainstorming: The illusion of productivity. Information Systems Research, 10, 110-133. doi:10.1287/isre.10.2.11010.1287/isre.10.2.110

31.

Pinsonneault

Barki

Gallupe

R. B.

Hoppen

(1999b). The illusion of electronic brainstorming productivity: Theoretical and empirical issues. Information Systems Research, 10, 378-380. doi:10.1287/isre.10.4.37810.1287/isre.10.4.378

32.

Riding

R. J.

McQuaid

D. G.

(1987). Characteristics of failing readers at 16+. British Educational Research Journal, 13, 51-58. doi:10.1080/014119287013010510.1080/0141192870130105

33.

Saint-Aubin

Ouellette

Poirier

(2005). Semantic similarity and immediate serial recall: Is there an effect on all trials? Psychonomic Bulletin & Review, 12, 171-177. doi:10.3758/PBR.12.1.17110.3758/PBR.12.1.171

34.

Topi

Valacich

J. S.

Rao

M. T.

(2002). The effects of personality and media differences on the performance of dyads addressing a cognitive conflict task. Small Group Research, 33, 667-701. doi:10.1177/104649640223862010.1177/1046496402238620

35.

Valacich

J. S.

Dennis

A. R.

Connolly

(1994). Idea generation in computer-based groups: A new ending to an old story. Organizational Behavior and Human Decision Processes, 57, 448-467. doi:10.1006/obhd.1994.102410.1006/obhd.1994.1024

36.

Valacich

J. S.

Jung

J. H.

Looney

(2006). The effects of individual cognitive ability and idea stimulation on individual idea generation performance. Group Dynamics, 10, 1-15. doi:10.1037/1089-2699.10.1.110.1037/1089-2699.10.1.1

37.

Valacich

J. S.

Wheeler

B. C.

Mennecke

B. E.

Wachter

(1995). The effects of numerical and logical group size on computer-mediated idea generation. Organizational Behavior and Human Decision Process, 62, 318-329. doi:10.1006/obhd.1995.105310.1006/obhd.1995.1053

38.

VanGundy

A. B.

(1992). Idea power: Techniques and resources to unleash the creativity in your organization. New York, NY: AMACOM.

39.

Yellen

R. E.

Winniford

Sanford

C. C.

(1995). Extraversion and introversion in electronically-supported meetings. Information & Management, 28, 63-74. doi:10.1016/0378-7206(94)00023-C10.1016/0378-7206(94)00023-C

40.

Yerkes

R. M.

Dodson

J. D.

(1908). The relation of strength of stimulus to rapidity of habit-formation. Journal of Comparative Neurology and Psychology, 18, 459-482. doi:10.1002/cne.92018050310.1002/cne.920180503

41.

Zuckerman

(1991). Psychobiology of personality. Cambridge, UK: Cambridge University Press.