brainstorming, brainstorming rules and decision making

17
23 VICKY L. PUTMAN PAUL B. PAULUS Brainstorming, Brainstorming Rules and Decision Making ABSTRACT This study investigated the effect of brainstorming experience on the ability of groups to subsequently select the best ideas for implementation. Participants generated ideas either alone or in interactive groups and with either the regular brainstorming rules or with additional rules designed to further increase the num- ber of ideas generated. All participants subsequently were asked to select their top five ideas in a group evaluation phase. Groups of individuals generating ideas in isolation (nominal groups) gener- ated more ideas and more original ideas and were more likely to select original ideas during the group decision phase than interactive group brainstormers. Addi- tional rules increased idea generation but not idea originality or idea selection. Key words: brainstorming, decision making, facilitators, group creativity. INTRODUCTION A popular method for generating ideas is the technique of group brainstorm- ing (Sutton & Hargadon, 1996) in which groups follow Osborn’s (1957) rules. These rules instruct the group members to build upon previous ideas, to not criti- cize any ideas, to generate a large number of ideas, and to feel free to contribute wild ideas. Although Osborn’s rules have been shown to improve group brain- storming performance compared to groups not using the rules (Parnes & Meadow, 1959), there is a great deal of evidence that interactive brainstorming in face-to- face groups leads to fewer ideas and fewer high quality ideas than nominal groups, which are composed of a comparable numbers of individual brainstormers (Diehl & Stroebe, 1987; Mullen, Johnson & Salas, 1991). These effects may be due to a variety of social and task factors such as evaluation apprehension, social loafing, production blocking, or downward comparisons (Diehl & Stroebe, 1991; Paulus, Dugosh, Dzindolet, Coskun & Putman, 2002). For example, group members may feel inhibited about expressing their ideas in groups because of social anxiety Volume 43 Number 1 First Quarter 2009 43-1-09.p65 2/27/2009, 7:10 PM 29 Black

Upload: vicky-l-putman

Post on 11-Jun-2016

218 views

Category:

Documents


3 download

TRANSCRIPT

Page 1: Brainstorming, Brainstorming Rules and Decision Making

23

V I C K Y L . P U T M A NP A U L B . P A U L U S

Brainstorming, BrainstormingRules and Decision Making

ABSTRACTThis study investigated the effect of brainstorming experience on the ability of

groups to subsequently select the best ideas for implementation. Participantsgenerated ideas either alone or in interactive groups and with either the regularbrainstorming rules or with additional rules designed to further increase the num-ber of ideas generated. All participants subsequently were asked to select theirtop five ideas in a group evaluation phase.

Groups of individuals generating ideas in isolation (nominal groups) gener-ated more ideas and more original ideas and were more likely to select originalideas during the group decision phase than interactive group brainstormers. Addi-tional rules increased idea generation but not idea originality or idea selection.

Key words: brainstorming, decision making, facilitators, group creativity.

INTRODUCTIONA popular method for generating ideas is the technique of group brainstorm-

ing (Sutton & Hargadon, 1996) in which groups follow Osborn’s (1957) rules.These rules instruct the group members to build upon previous ideas, to not criti-cize any ideas, to generate a large number of ideas, and to feel free to contributewild ideas. Although Osborn’s rules have been shown to improve group brain-storming performance compared to groups not using the rules (Parnes & Meadow,1959), there is a great deal of evidence that interactive brainstorming in face-to-face groups leads to fewer ideas and fewer high quality ideas than nominal groups,which are composed of a comparable numbers of individual brainstormers (Diehl& Stroebe, 1987; Mullen, Johnson & Salas, 1991). These effects may be due to avariety of social and task factors such as evaluation apprehension, social loafing,production blocking, or downward comparisons (Diehl & Stroebe, 1991; Paulus,Dugosh, Dzindolet, Coskun & Putman, 2002). For example, group members mayfeel inhibited about expressing their ideas in groups because of social anxiety

Volume 43 Number 1 First Quarter 2009

43-1-09.p65 2/27/2009, 7:10 PM29

Black

Page 2: Brainstorming, Brainstorming Rules and Decision Making

24

Brainstorming and Decision Making

(Camacho & Paulus, 1995) or they may have difficulty in effectively expressingtheir ideas when they have to wait their turn while others are expressing theirs(Nijstad, Stroebe & Lodewijkx, 2002). Social loafing (Karau & Williams, 1993)may occur in groups when individuals do not feel as accountable or identifiableto external evaluators for their performance in groups as they would if they per-formed as individuals. Several studies have found that performance in groupsmay move in the direction of the low performing members (Camacho & Paulus,1995; Paulus & Dzindolet, 1993). There is evidence that each of these factors playa role in the relatively poor performance of brainstorming groups, although somesuggest that production blocking or cognitive interference is the most importantfactor (Diehl & Stroebe, 1991; Nijstad, Stroebe & Diehl, 2003)

However, such negative effects of group brainstorming are not inevitable (Nijstad& Stroebe, 2006; Paulus & Brown, 2003; Paulus & Nakui, 2005). For example,studies in which groups exchange ideas on computers or by means of writinghave found that such groups can perform as well or better than nominal groups(Dennis & Williams, 2003; Dugosh, Paulus, Roland & Yang, 2000; Paulus & Yang,2000). Positive effects of group brainstorming may occur when other group mem-bers are a source of motivation (Paulus, Larey, Putman, Leggett & Roland, 1996)or a source of cognitive stimulation (Dugosh et al., 2000; Dugosh & Paulus, 2005).

Many manipulations appear to increase the number of ideas generated by bothnominal and interactive groups. Providing interactive groups or nominal groupswith high standards of reference for their performance has similar positive ben-efits for both types of groups (Paulus & Dzindolet, 1993). Trained facilitators whohelp manage the group interaction also often appear to increase the brainstorm-ing performance of groups (Kramer, Kuo & Dailey, 1997; Kramer, Fleming &Mannis, 2001; Offner, Kramer & Winter, 1996; Oxley, Dzindolet & Paulus, 1996).Facilitators have not been used with nominal groups, but we have found thatthe addition of “facilitator rules” to the brainstorming rules has similar positiveeffects on number of ideas generated for both interactive and nominal groups(Paulus, Nakui, Putman & Brown, 2006). These additional rules were developedfrom a search of the facilitator literature (Oxley et al., 1996). The additional rulesare as follows: stay focused on the task; do not tell stories; do not explain ideas;keep people talking, possibly by bringing up previous ideas; encourage others tocontribute. These additional rules may lead groups to be more efficient in theiridea sharing (less blocking) and may increase the motivation of all the brain-stormers in the group to persist in sharing ideas. One purpose of this study was toreplicate the utility of these additional rules for brainstorming. A second purposewas to determine the impact of prior solitary or group brainstorming and the useof regular or additional rules on a subsequent idea evaluation session.

In most real-world settings groups that generate ideas also spend some timesifting through these ideas to select the best ones for further consideration(Mintzberg, Raisinghani & Theoret, 1976). Essentially, groups often must decidewhich of the ideas that have been generated are the best ones. In the group brain-storming literature quality has been defined along several dimensions. One of the

43-1-09.p65 2/27/2009, 7:10 PM30

Black

Page 3: Brainstorming, Brainstorming Rules and Decision Making

Journal of Creative Behavior

25

chief aims of brainstorming is to come up with original or novel ideas. Noveltycan be defined objectively as low frequency ideas from the total pool of ideas(Dennis, Valacich, Carte, Garfield, Haley & Aronson, 1997). Typically, originalityis determined by the subjective ratings of judges (Amabile, 1996; Barki &Pinsonneault, 2001; Diehl & Stroebe, 1987). Another important dimensionfor idea quality is feasibility. Feasible ideas are those that have some reasonablepotential for application given worldly constraints. Again, feasibility is typicallybased on subjective ratings of judges. The best ideas may be those that are bothoriginal and feasible (Amabile, 1996; Simonton, 1988).

There is a literature on the idea evaluation process itself (e.g., Runco, 2003).For example, Mumford and his colleagues have examined the factors that influ-ence this process and the outcomes of different procedures (Dailey & Mumford,2006; Mumford, Lonergan & Scott, 2002). They also point out that this evalua-tion process is multifaceted involving appraisal, forecasting, and refinement(Mumford, Blair & Marcy, 2006). Although the idea evaluation process is of inter-est as a separate area of study, our concern was with the evaluation of ideasgenerated by the brainstormers themselves.

Evaluation of individually generated ideas in a subsequent group discussionsession was part of the Nominal Group Technique (Hegedus & Rasmussen, 1986;Van de Ven & Delbecq, 1971), but there are only a few studies that have examinedthe evaluation/decision making process after face-to-face brainstorming. In a studyof the effects of brainstorming on an induction task, it was found that prior brain-storming did lead to improvement in subsequent induction performance (McGlynn,McGurk & Effland, 2004). Kramer, et al. (1997) examined decision making in aself-report format. Groups and nominals were given Osborn’s typical brainstorm-ing instructions and a short practice session, whereas groups in another condi-tion were given no instructions or training. The participants brainstormed andthen reached a decision about the issue. For all three conditions, self-rated deci-sion quality was similar, but the two trained conditions were more positive both intheir evaluation of the group decision as well as in their evaluation of the decisionprocess. Larey and Paulus (1999) instructed nominal and interactive group par-ticipants to brainstorm for one session and then to select the best ideas from aseparate pool of ideas that they did not generate. They found that nominals per-formed better on the divergent task of brainstorming, whereas the groups per-formed better on the convergent task of deciding upon the top ideas. Faure (2004)had individual and group brainstormers select the best three ideas as a groupafter brainstorming. These ideas were either ones they had generated or onesgenerated by another group. Even though the individual brainstormers gener-ated more ideas than group brainstormers, there were no differences in thequality of the subsequent ideas selected as a group. However, groups selectingfrom their own pool of ideas selected ideas higher in effectiveness than those whoselected from ideas of another group. Most recently, Rietzschel, Nijstad & Stroebe(2006) examined idea selection after both group and individual brainstorming.Groups generated and selected ideas in groups and individuals generated and

43-1-09.p65 2/27/2009, 7:10 PM31

Black

Page 4: Brainstorming, Brainstorming Rules and Decision Making

26

Brainstorming and Decision Making

selected ideas alone. Although individuals generated ideas that were on the aver-age more original than ideas generated by groups, the individuals did not selectmore original ideas than groups.

The past research does not provide a clear consensus about the effectivenessof group evaluation of ideas after brainstorming. This is not surprising since thereare a number of bases for predicting alternative outcomes. Group members canhelp eliminate bad alternatives, such as ideas that are not feasible or not likely tobe accepted (Larey & Paulus, 1999; Laughlin, VanderStoep & Hollingshead, 1991).The diverse perspectives of the group members may be useful in evaluating theutility of the various ideas (Jackson, 1992; Schulz-Hardt, Jochims & Frey, 2002).However, there is some support for predicting that groups will not be particularlygood at the evaluation process. The well-established bias that groups oftenhave for paying attention to shared or common information may lead groups tofocus on ideas that reflect similar perspectives instead of considering original orinfrequent ideas (Stasser, Vaughan & Stewart, 2000; Wittenbaum, Hubbell &Zuckerman, 1999)

There are also some methodological reasons why previous studies may nothave found differences in idea evaluation after individual and group brainstorm-ing. In the Rietzschel et al. (2006) study the brainstorming and decision makingwas either as individuals or groups, so there was no assessment of the relation-ship of the prior brainstorming experience on a common decision making pro-cess. In the Faure (2004) study the brainstorming procedure was designed tominimize differences in performance between individuals and groups by havinggroups generate ideas in a round robin fashion. As a result she obtained only a26% difference in ideas between those conditions in comparison with the typical100% difference. In contrast the present study investigated group decision mak-ing after typical oral individual and group brainstorming, unconstrained by a roundrobin format. The nominal groups should be able to take advantage of their greaterpool of ideas in the group decision phase and select more good ideas. However, itis also possible that the nominal groups will have difficulty shifting to an interac-tive format in the decision making session. Since there has not yet been any shar-ing of ideas for the nominals, this process will have to be included in the decisionphase as well. If the nominal group members are susceptible to the shared infor-mation bias, they may not fully share their more original ideas with the rest of thegroup. This may be particularly true because nominal groups will have generatedmore total ideas and therefore will have a potentially more difficult task thaninteractive groups when selecting the top ideas from the total idea pool.

The additional rules manipulation enabled us to examine the independent con-tribution of number of ideas and type of brainstorming (nominal or interactive)both for the generation of ideas and for the decision making process. If the num-ber of ideas generated is responsible for an advantage in the decision makingphase, similar effects should be observed for the additional rules manipulationand the nominal/group interaction manipulation. On the other hand, if the oppor-tunity for interaction during brainstorming is critical for group decision making,

43-1-09.p65 2/27/2009, 7:10 PM32

Black

Page 5: Brainstorming, Brainstorming Rules and Decision Making

Journal of Creative Behavior

27

then the effects of group-type should not match the effects of the additional rulesmanipulation on decision making. If the familiarity and cohesiveness from priorinteraction is helpful in subsequent group decision making, then groupbrainstormers should be more successful than individual brainstormers in select-ing the best ideas (e.g., original ones). However, familiarity and cohesiveness couldalso inhibit a full sharing of unique or original ideas in interactive groups (Stasser& Birchmeier, 2003) and lead the groups to focus on more frequent or highlyfeasible ideas. Moreover, individual brainstormers may be more attentive to theshared ideas in a subsequent group decision session since they are sharing theseideas for the first time. This increased attention may lead to a better decision forindividual brainstormers than for group brainstormers.

This study was not designed to examine why nominal brainstorming and addi-tional rules improve brainstorming performance. However, this study did providean opportunity to evaluate one potential factor. Part of the problem in interactivebrainstorming is that the shared idea generation process makes it difficult toexpress ideas when they come to mind (Diehl & Stroebe, 1991). Furthermore, wehave noted that brainstormers often get “off-task” by needlessly elaborating onideas or by telling stories. This makes it even more difficult for group brainstormersto generate a high number of ideas (c.f., Dugosh, et al., 2000). Some of the addi-tional rules are designed to minimize this tendency. We will examine the extent towhich the rules make groups and/or individuals more efficient in the idea genera-tion process by analyzing the number of words used to express the ideas.

METHODPARTICIPANTS

One-hundred and twenty participants were recruited from the introduction topsychology classes. Their participation served to fulfill a research requirement.

DESIGNThe experimental design was a 2 (nominal vs. group) by 2 (Osborn vs. addi-

tional rules) factorial design. The experiment involved a brainstorming sessionthat was followed by a decision making session. Half of the participants gener-ated ideas as individuals and the other half generated ideas in groups of three.Half of the participants were given only Osborn’s rules for brainstorming and theother half of the participants were given the additional rules (see Appendix). Allparticipants made decisions in groups of three.

PROCEDURESAll participants were brought into a central room for informed consent proce-

dures and instructions. Participants were told we were interested in the generalissue of brainstorming and their ability to apply the rules provided in brainstorm-ing. The experimenter read the instructions and also told the participants to writedown each idea on the notepad provided, for later reference during the decision

43-1-09.p65 2/27/2009, 7:10 PM33

Black

Page 6: Brainstorming, Brainstorming Rules and Decision Making

28

Brainstorming and Decision Making

making session. After the instructions, half of the participants were taken to smallrooms to brainstorm alone and half were taken to a large room to brainstorm inan interactive group of three people. In session 1, all participants orallybrainstormed for 15 minutes into individual microphones about ways to improvethe university. The experimenter left the room when the session began.

In session 2, the participants who had previously brainstormed alone were nowplaced into a group of three in the large lab room for the decision making phase.In contrast, for the interactive group participants the decision making phaseoccurred in the same lab room in which they brainstormed. These participantssimply moved over to a different part of the lab room where the chairs werearranged for the decision making session. All participants were given instruc-tions for decision making, which indicated that they should discuss their ideasfrom the previous session and come to a decision about the top five ideas. Thegroups were told to rank-order what they decided were the best five ideas. Re-search by Hollingshead (1996) showed that groups make better decisions whenthey are told to rank-order their answers instead of simply trying to come up withthe overall best answer. Forcing the groups to compare the ideas should causethe group members to more carefully consider a greater amount of relevant infor-mation. The decision process was videotaped.

The groups were told that they had 15 minutes to make this decision, but thatthey would be given more time if needed. The experimenter left the room oncethe session began. The group was finished with their decision once they filled outthe top five list. A flip chart was provided for the groups to use to list their ideas fordiscussion. When one page was filled with ideas, it was attached to the wall. Theexperimenter recorded the length of time to make the decision. After they haddecided on the top five ideas, the individual participants were given a question-naire about their perceptions of the two sessions on 9-point scales. They wereasked to rate the number and quality of ideas generated, rate their enjoyment ofthe brainstorming tasks, rate the extent to which the rules helped brainstorming,and to rate their feelings (nervous, self-conscious, ate ease, calm, self-aware, andworried) on five point scales.

CODING AND RELIABILITYAll audiotapes from the brainstorming session were transcribed verbatim and

then coded for number of ideas. The lists of ideas written by each participantduring the brainstorming session also were checked for any ideas that were writ-ten but not spoken. These were counted since they could be shared during thedecision phase. For the interactive groups the number of ideas for each of thethree members was combined into a group total, deleting any repeated ideas. Forthe nominal brainstormers the three transcripts were compared and any repeatedideas were deleted, which resulted in an idea list that included ideas from all threenominals. A second person coded 20% of the transcripts to check for reliability.The overall Cronbach’s alpha was 0.99. The number of words used for each ideawas counted and used to develop a score of brainstorming efficiency for eachperson — the number of words used divided by the number of ideas generated.

43-1-09.p65 2/27/2009, 7:10 PM34

Black

Page 7: Brainstorming, Brainstorming Rules and Decision Making

Journal of Creative Behavior

29

All ideas from the brainstorming session were combined into a total list of 1,514ideas. This list was narrowed down to 604 non-repeated ideas for rating purposes,since many participants generated similar ideas. For example, the idea of “moreparking lots” and the idea of “more parking spaces” were both combined into theidea of “more parking” for the 604 idea list. Therefore, the ideas that were sepa-rate ideas for brainstorming purposes were not necessarily separate ideas forthe rating list. For example, one group generated two different but similar ideasabout textbooks (e.g., “don’t overcharge for books,” “cheaper books”). The groupreceived credit for two separate ideas in the brainstorming phase but for the deci-sion making phase both ideas were given the same label and both ideas receivedthe same rating on both dimensions. One rater created the shortened list of ideasand two other raters evaluated the combination decisions. Any disagreementswere discussed until there was complete agreement on the final list.

Feasibility and originality were each measured on a 1 to 5 Likert scale with 5being the highest on each dimension. An idea was rated as high in feasibility if theidea was considered to be practical and likely to be implemented at the univer-sity. An idea was rated as high in originality if it was thought to be creative and nottypically mentioned by most brainstormers. The quality ratings of the ideas interms of feasibility and originality were checked for reliability. As in previous stud-ies concerning idea quality (Dennis et al., 1997; Diehl & Stroebe, 1987, 1991),one highly trained rater coded all of the ideas and then another highly trainedrater coded 25% of the ideas. The agreement among the raters was assessed bycounting the number of times that the two raters were within one rating point ofeach other. This number was then expressed as the percentage of the total num-ber of ideas that were rated by both raters. The overall reliability for feasibility was93.17% and the overall reliability for originality was 85.71%. New ideas that werebrought up in the decision making session (often combinations of brainstormingideas) were rated with a reliability of 98.63% for feasibility and 93.15% for original-ity. Condition information was not available during the rating process so thatthis could not influence the ratings. The correlations between the measures offeasibility and originality was .29.

Raters coded the videotapes of the decision phase for time to decision, degreeof sharing of ideas, cohesion, conflict, and interruptions. The interrater reliabilitiesfor these measures ranged from .64 (cohesion) to .98 (time to decision).

MEASURES OF IDEA QUALITY AND DECISION MAKING PERFORMANCEFour measures of idea quality were used in this study. Within each group of

three members, the average feasibility of the ideas was calculated as well as theaverage originality. The frequency of the idea in the entire pool of brainstormingideas was used as a third measure of idea quality. The average frequency of theideas generated by each group was calculated. Because the best types of ideasmight be ones high in feasibility as well as high in originality, the fourth measureof idea quality in this study was a combined additive measure of feasibility andoriginality (Dennis et al., 1997; Diehl & Stroebe, 1987, 1991).

43-1-09.p65 2/27/2009, 7:10 PM35

Black

Page 8: Brainstorming, Brainstorming Rules and Decision Making

30

Brainstorming and Decision Making

To measure the quality of the decision made by the groups, the average feasi-bility of the group’s top five ideas was calculated as well as the average originalityand the average feasibility/originality combined score. Calculating the averagefrequency of the top five ideas was complicated by the fact that there were oftengroups that selected an idea as one of their top five but did not say the idea intheir brainstorming session or combined several brainstorming ideas into onetop five idea. Since the new ideas were not part of the brainstorming session,there were no frequency scores for these new ideas. Therefore, in calculating ourmeasure of frequency we gave a frequency of 1 both to ideas that were new to thedecision making session (not said by any of the 40 groups while brainstorming)and to top five ideas that were combinations of ideas from the brainstormingsession. The ideas that were common ideas (said by many groups but not thatparticular group) were given the frequency number of the idea in the entire poolof the experiment. Only seven of the 40 groups (five interactive groups and twonominal groups) had no new ideas or no combined ideas in their top five list.

RESULTSBRAINSTORMING PERFORMANCE

The number of ideas generated in session 1 was analyzed by means of a 2 x 2between-groups ANOVA. One factor was the group type (nominal or interactivegroup) and the other factor was the type of rules given (Osborn’s rules or addi-tional rules). Both main effects were significant. Participants who brainstormedas nominal groups generated more ideas (M = 45.75) than those who brainstormedin an interactive group (M = 29.95), F (1, 36) = 11.94, p < 0.01, eta2 = .25. Partici-pants given the additional rules generated more ideas (M = 44.00) than thosewho were given only the Osborn rules (M = 31.70), F (1, 36) = 7.23, p < 0.02,eta2 = .17. The interaction was not significant (see Table 1). Tukey post hoc com-parisons of the four conditions indicated that only the Osborn group conditionand the additional rules nominal condition were significantly different.

TABLE 1. Mean Number of Nonrepeated Ideas Generated and Mean Numberof Words Per Idea as a Function of Type of Rules and Group Type.

Type of rules

Osborn’s Additional

Group No. No. No. No.type ideas words/idea ideas words/idea

InteractiveM 23.40 123.47 36.50 62.75

SD 6.57 60.06 9.50 29.65

NominalsM 40.00 62.30 51.50 41.63

SD 17.81 17.65 19.65 19.19

43-1-09.p65 2/27/2009, 7:10 PM36

Black

Page 9: Brainstorming, Brainstorming Rules and Decision Making

Journal of Creative Behavior

31

The ratio of number of words per idea also was analyzed by a 2 x 2 ANOVA.Participants who were given the Osborn rules used more words to expresstheir ideas (M = 92.88) than did participants who were given the additional rules(M = 52.19), F (1, 36) = 12.82, p < 0.01, eta2 = .26. Participants in groups usedmore words per idea (M = 93.11) than did nominals (M = 51.96), F (1, 36) = 13.11,p < 0.01, eta2 = .27. (see Table 1). A regression analysis indicated that the wordusage mediated the effect of group condition on ideas. The regressions of num-ber of ideas on group condition and of number of words on group condition wereboth significant, B = .47, t(38) =3.24, p < .01 and B = .43, t (38) = 2.91, p < .01.The relationship between groups and number of ideas was no longer significantwhen both number of ideas and number of words were entered into the regres-sion model, B = .20, t(37) = 1.63, ns. The effect of number of words was stillsignificant, B = .61, t(37) = 4.85, p < .01. Sobel’s test indicated that this effect issignificant (Z = 2.49, p < . 05).

BRAINSTORMING QUALITY

For originality, the ANOVA revealed a main effect for group type. The ideasgenerated by the nominals had a higher mean level of originality (M = 2.83) thandid the ideas generated by the interactive groups (M = 2.63), F (1, 36) = 6.63, p <0.02, eta2 = .16. The ANOVA for frequency also revealed a group type main effect.The ideas generated by the groups were more frequent in the idea pool (M =9.06) than were the ideas generated by the nominals (M = 7.08), F (1,36) = 10.21,p < 0.005, eta2 = .22 (see Table 2). There were no significant effects of rule typeon the measures of idea quality.

TABLE 2. Mean Rating of Idea Originality and Mean Idea Frequency as aFunction of Type of Rules and Group Type.

Type of rules

Osborn’s Additional

Group type Originality Frequency Originality Frequency

InteractiveM 2.58 9.74 2.69 8.39

SD 0.25 2.42 0.23 2.09

NominalsM 2.77 7.11 2.88 7.06

SD 0.18 1.67 0.28 1.53

43-1-09.p65 2/27/2009, 7:10 PM37

Black

Page 10: Brainstorming, Brainstorming Rules and Decision Making

32

Brainstorming and Decision Making

DECISION PHASE; ANALYSES FOR TOP FIVE IDEASFor the top five ideas selected by the groups, the ANOVAs revealed no signifi-

cant differences among conditions for the feasibility measure. For the averageoriginality of the top five ideas, there was a group type main effect. Participantswho brainstormed as nominals selected top five ideas that were more original(M = 2.88) than did those participants who brainstormed as interactive groups(M = 2.55), F (1, 36) = 9.05, p < 0.01, eta2 = .20. For the frequency of the top fiveideas, there was a group type main effect. Interactive brainstorming groupsselected ideas that were more frequent in the idea pool (M = 10.19) than thosewho brainstormed as nominal groups (M = 7.46), F (1, 36) = 4.42, p < 0.05, eta2

= .11. To determine to what extent the mean quality of the selected ideas in thedecision phase differed from the average of the ideas of the brainstorming phasefor nominals or groups, separate analyses were done within each of these condi-tions. The only significant effect obtained was for frequency of ideas in groups.The ideas selected in the decision phase were more frequent (M = 11.51, SD =4.56) than those in the brainstorming phase (M = 9.06, SD = 2.31), F (1, 19) =5.15, p < .05, eta2 = .21.

The amount of time taken to make a decision was not correlated with the num-ber of ideas generated but was significantly correlated both with the originalityof the brainstorming ideas, r = 0.38, p < 0.05, and with the originality of the topfive ideas, r = 0.33, p < 0.05. No significant differences were obtained for dataobtained from the videotapes of the decision session.

QUESTIONNAIRE ANALYSESFor questions about feelings during the brainstorming session, the only signifi-

cant result was a main effect for rules. The participants who were given the Osbornrules reported being significantly calmer (M = 4.27) than did the participantswho were given the additional rules (M = 3.75), F (1, 116) = 7.92, p < 0.01. Forsatisfaction with the group’s final decision, there were also main effects for rules.Participants who were given the Osborn rules were more satisfied with the finaldecision (M = 8.68) than were those participants who were given the additionalrules (M = 8.20), F (1, 116) = 7.99, p < 0.006. Participants who were given theOsborn rules felt the process was more thorough (M = 7.78) than did thoseparticipants who were given the additional rules (M = 7.20), F (1, 116) = 5.66,p < 0.02. There were no differences among the conditions in ratings of behaviorduring the decision phase.

DISCUSSIONThis study examined both the idea generation and idea selection processes for

brainstorming groups as a function of the type of brainstorming conditions andthe rules employed. The pattern of number of ideas generated was consistentwith past research (e.g., Diehl & Stroebe, 1987; Paulus et al., 2006), with bothnominal status and additional rules leading to increased generation of ideas. Theideas generated in the nominal conditions also tended to be more originaland less frequent than those generated in the group conditions. One of the main

43-1-09.p65 2/27/2009, 7:10 PM38

Black

Page 11: Brainstorming, Brainstorming Rules and Decision Making

Journal of Creative Behavior

33

questions in this study was the extent to which participants in the various condi-tions would be able to sort through their idea pool and then select the best ideas.The idea selection reflected the higher quality of ideas generated in the nominalcondition, with those who had generated ideas as nominals subsequently select-ing more original and less frequent ideas than those who had generated ideas ingroups. So this study demonstrated that the advantage of nominal groups car-ries over to a subsequent idea selection phase. This outcome may seem obvious,but it should be noted that this did not occur in two prior studies of idea selectionafter brainstorming (Faure, 2004; Rietzschel, et al., 2006). In the Rietzschel et al.(2006) study individuals who brainstormed individually also generated more ideasand ideas that were more original and less feasible than those who brainstormedin groups. The individuals subsequently selected ideas as individuals and thegroups as a group. There was no difference between the conditions in the qualityof ideas selected. It is possible that the contrast between individual idea genera-tion and group decision making in the present study was critical in producing abenefit of nominal brainstorming in the decision making phase. Individualbrainstormers may be more attentive to the ideas being shared in the group brain-storming session since they are hearing them for the first time. This may increasetheir sensitivity to the quality (originality and frequency) of those ideas. It is alsopossible that performing as an individual prior to group decision making makesone less prone to some of the factors that inhibit effective decision making suchas the commonality bias (Stasser & Birchmeier, 2003). As mentioned earlier, thefailure of the Faure (2004) study to find a difference in group idea selection afterindividual or group brainstorming may be related to the smaller than usual differ-ence in performance between the individual and group brainstormers.

Our findings indicate that having a high number of ideas to sort through doesnot necessarily distract or overwhelm participants in their attempts to select thebest ideas. The interaction time and prior sharing of ideas that the groups hadtogether during brainstorming did not seem to put them at an advantage over thenominals who had not interacted with each other prior to the decision makingsession. However, spending more time in the decision making phase was relatedto more original ideas being selected in the top five. Thus taking more time canlead to a better decision making outcome (c.f., Hirokawa, 1990). Surprisingly,the mean originality of the ideas selected in the decision phase was not higherthan mean originality of the ideas in the brainstorming phase. This finding isconsistent with a similar finding by Rietzschel et al. (2006) that individuals orgroups are not particularly good at picking out high quality ideas. Essentially thegroups in our study performed at a chance level in the selection of the best ideasalong the dimension of originality. There was a bias away from chance in theselection of more frequent ideas in the decision phase. This suggests that ideasthat are more likely to have been have been generated by more than one indi-vidual in the group or are more likely to “resonate” with group members (Tindale,Meisenhelder, Dykema-Engblade & Hogg, 2001) and have a higher chance ofbeing selected in group decision making. This of course is consistent with the

43-1-09.p65 2/27/2009, 7:10 PM39

Black

Page 12: Brainstorming, Brainstorming Rules and Decision Making

34

Brainstorming and Decision Making

bias toward shared information found in the problem solving literature (Stasser &Birchmeier, 2003).

The differential pattern of results for the group type and rule type manipula-tions for brainstorming and decision making are theoretically instructive. Althoughthe additional rules condition enhanced the production of number of ideas, it didnot affect the quality of the ideas generated or selected in the decision phase.This result and that of past studies (Faure, 2004; Rietzschel et al., 2006) suggestthat simply generating a high number of ideas does not insure that an increasednumber of good ideas will be generated and therefore selected in the decisionphase. Why is increased idea generation related to increased originality for thenominal groups but not for the additional rules conditions? Possibly, with the lackof social distraction in the nominal conditions there is more careful processing ofthe ideas being generated (reflection about the ideas and memory for the ideas;Brown, Tumeo, Larey & Paulus, 1998). As a result there may be a higher numberof good ideas generated and a higher number available for the decision makingphase. This analysis is consistent with the functional perspective of Hirokawa(1990), which emphasizes the importance of thorough processing, especially whenthere is a high quantity of information.

The use of additional rules enhanced the number of ideas generated by nominalsas well as for groups. One possible basis for this effect, as well as the superiorperformance of the nominals, is the number of words used to express ideas. Highlevels of brainstorming performance for nominal groups and additional rulesgroups were associated with a low number of words per idea. These findings andthe related mediation analysis suggest that brainstorming efficacy is related toefficiency of idea generation. Interactive groups generated about half as manyideas as nominals possibly because they used almost twice as many words peridea as the nominal groups. Group discussions may be more off-task than nomi-nal idea generation because interactive groups may be prone to discuss issuesnot directly related to the brainstorming task. Alternatively, group members gen-erate their ideas in a more conversational style than nominals and this conversa-tional style may require more words. Interactive group members must makeaffirming statements to keep the conversation going. However, those statementscan take time away from generating ideas or in other ways distract the idea gen-eration process (Dugosh et al., 2000). In a similar fashion, participants given theadditional rules are more efficient in how they express their ideas, compared toparticipants who are only given Osborn’s rules. The additional rules may leadgroup members to be more task focused. One compelling feature of the resultsis that the additional rules interactive groups and the nominal Osborn groupswere similar both in number of ideas generated and in numbers of words used toexpress those ideas. This finding suggests that if the wordiness of interactive groupsand nominal groups is similar, then their performance will be similar. This effectoccurred in spite of the fact that groups have experienced some degree of pro-duction blocking and suggests that the associative benefits of group interactionmay counter the negative effects of production blocking (Brown & Paulus, 2002).

43-1-09.p65 2/27/2009, 7:10 PM40

Black

Page 13: Brainstorming, Brainstorming Rules and Decision Making

Journal of Creative Behavior

35

The findings for the quality of the brainstorming ideas also replicate previousresults (Barki & Pinsonneault, 2001; Casey, Gettys, Pliske & Mehle, 1984; Diehl &Stroebe, 1987, 1991). Groups not only generated fewer total ideas than didnominals, but they also generated fewer original ideas. Nominal groups gener-ated more original and low frequency ideas than interactive groups. Unlike mostof the previous idea quality studies, the current study also found differences inmean level of originality between nominals and groups.

The questionnaire results revealed that the additional rules participants hadsome negative reactions to the decision making phase. Possibly, the pressurefrom the implied standards set by the additional rules may be one reason thatthese participants were less satisfied with their decision, felt that their decisionwas less thorough, and felt less calm than did the participants who were onlygiven the Osborn rules. The negative feelings of the additional rules participantsare of course inconsistent with the fact that they performed better in the brain-storming phase and the same in the decision phase in comparison to the Osbornparticipants. The finding is reminiscent of previous findings that groups typicallydo not have a very good idea of how well they are performing (e.g., Paulus,Dzindolet, Poletes & Camacho, 1993).

Although this study has increased our knowledge about the idea evaluationprocess after brainstorming, we were not able to measure the processes criticalto the various effects obtained. For example, we do not know exactly what it isabout the additional rules that leads to increased performance. Our results sug-gest that increased efficiency in expressing ideas in the additional rules condition(fewer words per idea) may be an important factor. However, we did not obtainmeasures of how effectively participants followed the rules. Videotape analysesof the group decision process revealed no differences among the conditions inthe extent of sharing of ideas, how well the ideas were discussed, degree of cohe-sion or degree of conflict. Future studies will need to determine more exactly thebasis for the effects of additional rules and the effects of prior brainstorming onidea evaluation. Our task situation was limited in time for brainstorming and deci-sion making. More extensive time for both brainstorming and decision makingmight affect the type of results obtained and might yield a more extensive andricher type of evaluation process (Mumford et al., 2006).

In conclusion, it was found that individuals who generated ideas alone per-formed better in both the brainstorming and group decision phase in comparisonto those who brainstormed in groups. Additional brainstorming rules facilitategeneration of number of ideas but not the number of high quality ideas and thesubsequent decision making performance. The high number of ideas generatedby the nominal groups and the additional rules groups appears to be related tothe fact that these groups use fewer words to express their ideas. Future studieswill need to determine why the high idea generation of the nominal groups isassociated with increased originality but high idea generation produced byanother manipulation such as additional rules or individually generating ideasand selecting ideas (Rietzschel et al., 2006) does not have a similar benefit.

43-1-09.p65 2/27/2009, 7:10 PM41

Black

Page 14: Brainstorming, Brainstorming Rules and Decision Making

36

Brainstorming and Decision Making

REFERENCESAMABILE, T. M. (1996). Creativity in context. Boulder, CO: Westview Press.

BARKI, H. & PINSONNEAULT, A. (2001). Small group brainstorming and idea quality: Is electronicbrainstorming the most effective approach? Small Group Research, 32, 158-205.

BROWN, V. R. & PAULUS, P. B. (2002). Making group brainstorming more effective: Recommendationsfrom an associative memory perspective. Current Directions in Psychological Science, 11, 208-212.

BROWN, V., TUMEO, M., LAREY, T. S. & PAULUS, P. B. (1998). Modeling cognitive interactions duringgroup brainstorming. Small Group Research, 29, 495-526.

CAMACHO, L. M. & PAULUS, P. B. (1995). The role of social anxiousness in group brainstorming. Journalof Personality and Social Psychology, 68, 1071-1080.

CASEY, J. T., GETTYS, C. F., PLISKE, R. M. & MEHLE, T. (1984). A partition of small group predecisionperformance into informational and social components. Organizational Behavior and HumanPerformance, 34, 112-139.

DAILEY, L. & MUMFORD, M. D. (2006). Evaluative aspects of creative thought: Errors in appraising theimplications of new ideas. Creativity Research Journal, 18, 367-384.

DENNIS, A. R., VALACICH, J. S., CARTE, T. A., GARFIELD, M. J., HALEY, B. J. & ARONSON, J. E.(1997). Research report: The effectiveness of multiple dialogues in electronic brainstorming.Information Systems Research, 8, 203-211.

DENNIS, A. R. & WILLIAMS, M. L. (2003). Electronic brainstorming: Theory, research, and future directions.In P. B. Paulus & B. A. Nijstad (Eds.), Group creativity. (pp. 160-178) New York: Oxford.

DIEHL, M. & STROEBE, W. (1987). Productivity loss in brainstorming groups: Toward the solution of ariddle. Journal of Personality and Social Psychology, 53, 497-509.

DIEHL, M. & STROEBE, W. (1991). Productivity loss in idea-generating groups: Tracking down the blockingeffect. Journal of Personality and Social Psychology, 61, 392-403.

DUGOSH, K. L. & PAULUS, P. B. (2005). Cognitive and social comparison processes in brainstorming.Journal of Experimental Social Psychology, 41, 313-320.

DUGOSH, K. L., PAULUS, P. B., ROLAND, E. J. & YANG, H. C. (2000). Cognitive stimulation inbrainstorming. Journal of Personality and Social Psychology, 79, 722-735.

FAURE, C. (2004). Beyond brainstorming: Effects of different group procedures on selection of ideasand satisfaction with the process. Journal of Creative Behavior, 38, 13-34.

HEGEDUS, D. M. & RASMUSSEN, R. V. (1986). Task effectiveness and interaction process of modifiednominal group technique in solving an evaluation problem. Journal of Management, 12, 545-560.

HIROKAWA, R. Y. (1990). The role of communication and group decision-making efficacy: A task-contingency perspective. Small Group Research, 21, 190-204.

HOLLINGSHEAD, A. B. (1996). Information suppression and status persistence in group decision-making:The effects of communication media. Human Communication Research, 23, 193-219.

JACKSON, S. E. (1992). Team composition in organizational settings: Issues in managing an increasinglydiverse work force. In S. Worchel, W. Wood & J. A. Simpson (Eds.) Group process and productivity(pp. 138-173). Newbury Park: Sage.

KARAU, S. J. & WILLIAMS, K. D. (1993). Social loafing: A meta-analytic review and theoretical integration.Journal of Personality and Social Psychology, 65, 681-706.

KRAMER, M. W., KUO, C. L. & DAILEY, J. C. (1997). The impact of brainstorming techniques on subsequentgroup processes: Beyond generating ideas. Small Group Research, 28, 218-242.

KRAMER, T. J., FLEMING, G. P. & MANNIS, S. M. (2001). Improving face-to-face brainstorming throughmodeling and facilitation. Small Group Research, 32, 533-557.

LAREY, T. S. & PAULUS, P. B. (1999). Group preference and convergent tendencies in small groups: Acontent analysis of group brainstorming performance. Creativity Research Journal, 12, 175-184.

43-1-09.p65 2/27/2009, 7:10 PM42

Black

Page 15: Brainstorming, Brainstorming Rules and Decision Making

Journal of Creative Behavior

37

LAUGHLIN, P. R., VANDERSTOEP, S. W. & HOLLINGSHEAD, A. B. (1991). Collective versus individualinduction: Recognition of truth, rejection of error, and collective information processing. Journal ofPersonality & Social Psychology, 61, 50-67.

MCGLYNN, R. P., MCGURK, D. & EFFLAND, V. S. (2004). Brainstorming and task performance in groupsconstrained by evidence. Organizational Behavior and Human Decision Processes, 93, 75-87.

MINTZBERG, H., RAISINGHANI & THEORET, A. (1976). The structure of unstructured decision processes.Administrative Science Quarterly, 21, 246-275.

MULLEN, B., JOHNSON, C. & SALAS, E. (1991). Productivity loss in brainstorming groups: A meta-analytic integration. Basic and Applied Social Psychology, 12, 3-23.

MUMFORD, M. D., BLAIR, C. S. & MARCY, R. T. (2006). Alternative knowledge structures in creativethought: Schema, associations, and cases. In J. C. Kaufman and J. Baer (Eds.), Creativity andreason in cognitive development. (pp. 117-136). New York: Cambridge University Press.

MUMFORD, M. D., LONERGAN, D. C., SCOTT, G. (2002). Evaluating creative ideas: Processes, standards,and context. Inquiry: Critical thinking across the disciplines, 22, 21-30.

NIJSTAD, B. & STROEBE, W. (2006). How the group affects the mind: A cognitive model of idea generationin groups. Personality and Social Psychology Review, 10, 186-213.

NIJSTAD, B. A., STROEBE, W. & DIEHL, M. (2003). Cognitive stimulation and interference in idea-generating groups. In P. B. Paulus and B. A. Nijstad (Eds.), Group creativity: Innovation throughcollaboration (137-159). New York: Oxford University Press.

NIJSTAD, B. A., STROEBE, W. & LODEWIJKX, H. F. M. (2002). Cognitive stimulation and interference ingroups: Exposure effects in an idea generation task. Journal of Experimental Social Psychology,38, 535-544.

OFFNER, A. K., KRAMER, T. J. & WINTER, J. P. (1996). The effects of facilitation, recording, and pauseson group brainstorming. Small Group Research, 27, 283-298.

OSBORN, A. F. (1957). Applied imagination (1st ed.). New York: Scribner.

OXLEY, N. L., DZINDOLET, M. T. & PAULUS, P. B. (1996). The effects of facilitators on the performanceof brainstorming groups. Journal of Social Behavior and Personality, 11, 633-646.

PARNES, S. J. & MEADOW, A. (1959). Effect of “brainstorming” instructions on creative problem-solvingby trained and untrained subjects. Journal of Educational Psychology, 50, 171-176.

PAULUS, P. B. & BROWN, V. (2003). Enhancing ideational creativity in groups: Lessons learned fromresearch on brainstorming. In P. B. Paulus & B. A. Nijstad (Eds.), Group creativity. (pp. 110-136)New York: Oxford.

PAULUS, P. B., DUGOSH, K. L., DZINDOLET, M. T., COSKUN, H. & PUTMAN, V. L. (2002). Social andcognitive influences in group brainstorming: Predicting production gains and losses. European SocialPsychology Review, 12, 299-325.

PAULUS, P. B. & DZINDOLET, M. T. (1993). Social influence processes in group brainstorming. Journalof Personality and Social Psychology, 64, 575-586.

PAULUS, P. B., DZINDOLET, M. T., POLETES, G. & CAMACHO, L. M. (1993). Perception of performancein group brainstorming: The illusion of group productivity. Personality and Social PsychologyBulletin, 19, 78-89.

PAULUS, P. B., LAREY, T. S., PUTMAN, V. L., LEGGETT, K. L. & ROLAND, E. J. (1996). Social influenceprocesses in computer brainstorming. Basic and Applied Social Psychology, 18, 3-14.

PAULUS, P. B. & NAKUI, T. (2005). Facilitation of group brainstorming. In S. Schuman (Ed.), The IAFhandbook of group facilitation (pp. 103-114). San Francisco: Jossey-Bass.

PAULUS, P. B., NAKUI, T., PUTMAN, V. L. & BROWN, V. R. (2006). Effects of task instructions and briefbreaks on brainstorming. Group dynamics: Theory, research, and practice, 10, 206-219.

PAULUS, P.B. & YANG, H.C. (2000). Idea generation in groups: A basis for creativity in organizations.Organizational Behavior and Human Decision Processes, 82, 76-87.

RIETZSCHEL, E., NIJSTAD, B. A. & STROEBE, W. (2006). Productivity is not enough: A comparison ofinteractive and nominal brainstorming groups on idea generation and selection. Journal ofExperimental Social Psychology, 42, 244-251.

43-1-09.p65 2/27/2009, 7:10 PM43

Black

Page 16: Brainstorming, Brainstorming Rules and Decision Making

38

Brainstorming and Decision Making

RUNCO, M. A. (2003). Idea evaluation, divergent thinking, and creativity: Perspectives on creativityresearch. In M. A. Runco (Ed.) Critical creative processes.(pp. 69-94) Cresskill, NJ: Hampton Press.

SCHULZ-HARDT, S., JOCHIMS, M. & FREY, D. (2002). Productive conflict in group decision-making:Genuine and contrived dissent as strategies to counteract biased information seeking. OrganizationalBehavior and Human Decision Processes, 88, 563-586.

SIMONTON, D. K. (1988). Scientific genius: A psychology of science. New York: Cambridge UniversityPress.

STASSER, G. & BIRCHMEIER, Z. (2003). Group creativity and collective choice. In P. B. Paulus & B. A.Nijstad (Eds.), Group creativity. (pp. 85-109) New York: Oxford.

STASSER, G., VAUGHAN, S. I. & STEWART, D. D. (2000). Pooling unshared information: The benefits ofknowing how access to information is distributed among group members. Organizational andHuman Decision Processes, 82, 102-116.

SUTTON, R. I. & HARGADON, A. (1996). Brainstorming groups in context: Effectiveness in a productdesign firm. Administrative Science Quarterly, 41, 685-718.

TINDALE, R. S., MEISENHELDER, H. M., DYKEMA-ENGBLADE, A. A. & HOGG, M. A. (2001). Sharedcognition in small groups. In M. A. Hogg and S. Tindale (Eds.), Blackwell Handbook of SocialPsychology: Group Processes. Malden, MA: Blackwell Publishers.

VAN DE VEN, A. & DELBECQ, A. L. (1971). Nominal versus interacting group processes for committeedecision-making effectiveness. Academy of Management Journal, 14, 203-212.

WITTENBAUM, G. M., HUBBELL, A. P. & ZUCKERMAN, C. (1999). Mutual enhancement: Toward anunderstanding of the collective preference for shared information. Journal of Personality & SocialPsychology, 77, 967-978.

This paper is based in part on a dissertation by the first author under the supervision of the secondauthor. Thanks are due to Toshihiko Nakui for his assistance with the data analysis. For additionalinformation about this research, contact the authors at [email protected] or the College of Science, Box19047, University of Texas at Arlington, Arlington, TX 76019.

43-1-09.p65 2/27/2009, 7:10 PM44

Black

Page 17: Brainstorming, Brainstorming Rules and Decision Making

Journal of Creative Behavior

39

APPENDIX

OSBORN’S RULES AND ADDITIONAL RULES FOR INTERACTIVE GROUPSBrainstorming is a form of group interaction which is used to facilitate the flow

of ideas. It is widely used in a large number of U.S. corporations, and is generallyused when new, unique, original, and creative ideas are desired. It is not used tosolve everyday problems. The procedure is relatively straightforward and easy tocomprehend. The following rules are for brainstorming in groups. We want you toapply these rules as best as you can while working as a group.

1.) Criticism is ruled out. Adverse judgment of ideas must be withheld. No oneshould criticize anyone else’s ideas. Say everything you think of.

2.) Freewheeling is welcome. The wilder the idea the better. It is easier to tamedown than to think up. Don’t be afraid to say anything that comes to mind.The farther out the idea the better. This will stimulate more and better ideas.

3.) Quantity is wanted. The greater the number of ideas the more likelihood ofwinners. Come up with as many as you can.

4.) Combination and improvement are sought. You should try to suggest howideas of others can be joined or changed into still better ideas. Don’t be afraidto combine and improve on them.

5.) Stay focused on the task. Concentrate on the problem at hand and avoidengaging in irrelevant thought processes or discussions. When it is neces-sary to interrupt a group member, say something like, “Remember that weneed to stay focused on our task.”

a.) Do not tell stories. We are only interested in your ideas. Do not allowyour group members to tell stories about their experiences.

b.) Do not explain ideas. Do not allow your group members to expand onwhy they think something is good or bad. Let them say an idea and theninterrupt them.

6.) Keep the brainstorming going. During a lapse of time when no one is talk-ing, someone in the group should say something like, “Let’s see what otherideas we can come up with for (restate the problem).”

7.) Return to previous categories. When the group members are not talkingvery much, go back to categories of ideas that have already been mentionedand try to build on these previous ideas. For example, say “Does anyonehave any more ideas related to (restate an idea already suggested)?”

43-1-09.p65 2/27/2009, 7:10 PM45

Black