Group information elaboration and group decision making Van_Ginkel_Van_Knippenberg_2008_v0443.pdf

Please download to get full document.

View again

of 16
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report



Views: 11 | Pages: 16

Extension: PDF | Download: 0

Related documents
Available online at Organizational Behavior and Human Decision Processes 105 (2008) 82–97 Group information elaboration and group decision making: The role of shared task representations Wendy
  Group information elaboration and group decision making:The role of shared task representations Wendy P. van Ginkel  * , Daan van Knippenberg Rotterdam School of Management, Erasmus University Rotterdam, Burgermeester Oudlaan 50, 3062 PA Rotterdam, The Netherlands Received 16 December 2004Available online 30 October 2007Accepted by David Harrison Abstract Decisionmakinggroupsoftenexchangeandintegratedistributedinformationtoalesserextentthanisdesirableforhigh-qualitydeci-sions.Weproposethatgroupmembers’sharedtaskrepresentationsplayanimportantroleinthisrespect,becausegroupsareofteninsuf-ficientlyattunedtothetask’sinformationelaborationrequirements.Taskrepresentationsemphasizingelaborationofdecision-relevantinformationshouldthereforeenhancedecision-makingperformance.Thisshouldholdespeciallywhengroupmembersrealizethattheyshare these task representations, because this realization removes psychological barriers to introducing new insights. Testing thesehypotheses, we compared information elaboration and decision-making performance of control groups and groups receiving instruc-tionsemphasizinginformationelaborationintwoexperiments.Halfoftheexperimentalgroupswerealsomadetorealizethattheysharedthe elaboration instructions. As predicted, groups with task representations emphasizing information elaboration and the realizationthey shared these representations outperformed groups in the other conditions. This effect was mediated by information elaboration.   2007 Elsevier Inc. All rights reserved. Keywords:  Group decision making; Distributed information; Shared task representations Organizations tend to rely on small groups fornumerous purposes. One function for which smallgroups seem especially suitable is making decisions thatrequire a wide array of knowledge (Cohen & Levinthal,1990; Kozlowski & Bell, 2003; Tindale, Kameda, &Hinsz, 2001). However, while a major reason for relyingon groups for decision making purposes has to do withthe broader range of resources groups possess comparedto individuals (Hinsz, Vollrath, & Tindale, 1997),research has shown that groups are not always goodusers of their informational resources. Not only do deci-sion making groups with distributed information oftenfail to pool members’ unique knowledge (Stasser,1999; Wittenbaum & Stasser, 1996), even when uniqueinformation gets pooled groups often fail to recognizeits relevance and focus more on information known toall members before group discussion (Gigone & Hastie,1993; Winquist & Larson, 1998). This can result inlower-quality group decisions.While earlier research addressing this issue has madegreat progress in identifying factors that affect groups’use of distributed information (Stasser, 1999), it has notfocused on what we propose is one of the issues layingattherootsofgroups’suboptimaluseofdistributedinfor-mation: group members’ task representations. Buildingon a growing body of knowledge about socially sharedcognition in groups (cf. Cannon-Bowers & Salas, 2001;Klimoski&Mohammed,1994;Mathieu,Goodwin,Heff-ner,Salas,&Cannon-Bowers,2005;Tindaleetal.,2001),we propose that group members’ shared task representa-tions (cf. Tindale, Smith, Thomas, Filkins, & Sheffey,1996)play animportantroleingroups’useoftheirinfor-mationalresources.Apivotalproblemwithdecisionmak- 0749-5978/$ - see front matter    2007 Elsevier Inc. All rights reserved.doi:10.1016/j.obhdp.2007.08.005 * Corresponding author. E-mail address: (W.P. van Ginkel).  Available online at Organizational Behavior and Human Decision Processes 105 (2008) 82–97  inggroupsisthattheyseemtobelievetheirtaskismainlyabout achieving consensus, for instance through poolingpreferences and making compromises rather than that itis about discussing (distributed) information and basinga decision on this information. Thus, group membersoftenseeminsufficientlyattunedtothetask’sinformationintegrationrequirementsandseemtofocusmoreonpool-ing preferences in reaching a decision. Groups shouldtherefore make better use of distributed informationwhen group members’ task representations emphasizethe exchange, discussion, and integration of decision-rel-evant information. We propose that the effects of suchtask representations are not just a matter of individualgroup members’ understanding of the task, but shouldobtain especially when members realize that they sharethese task representations, because the realization thatotherssharesimilarideascanremovepsychologicalbarri-erstointroducingnewinsightsintogroupdiscussion.Wetested these ideas in two experiments that illustrate howan analysis of the use of distributed information in termsof shared task representations may enhance our under-standingofgroupdecisionmaking,andpotentiallygroupperformance more generally. Information elaboration and task representations In recognition of the fact that groups may reach high-qualitydecisionswhentheyareabletointegratethediver-sity of information and perspectives held by their mem-bers, research in group decision making has investedsubstantial effort in understanding groups’ use of distrib-uted information. This research has uncovered suchdiverse factors as team leadership (Larson, Foster-Fish-man, & Franz, 1998), knowledge about group memberexpertise(Stewart&Stasser,1995),andmemberfamiliar-ity with each other (Gruenfeld, Mannix, Williams, &Neale,1996)asdeterminantsofgroups’useofdistributedinformation (for a review, see Stasser, 1999). Somewhatsurprisingly, a factor that may lie at the basis of groups’use of their informational resources has received littleattentioninthisrespect:groupmembers’sharedtaskrep-resentations. We propose that the notion of shared taskrepresentationshasimportantimplicationsforourunder-standing of groups’ use of distributed information.Based for a large part on the work of  Tindale et al.(1996) shared task representations are defined as ‘‘anyconcept, norm, perspective, or process concerning theteam task that is held in common by team members’’.Recently, there has been an upsurge of evidence foreffects of shared cognition in groups. Several studieshave shown that when group members share task-rele-vant cognition (i.e., mental models), this can have bene-ficial effects on group functioning (cf. Marks, Zaccaro,& Mathieu, 2000; Mathieu, Goodwin, Heffner, Salas,& Cannon-Bowers, 2000; Mathieu et al., 2005). Whilemuch of the research in shared cognition seems to havebeen focused on tasks of a more executive nature (cf.McGrath, 1984), we propose that a shared understand-ing or shared representations of the task can also greatlyinfluence performance on tasks of a more problem-solv-ing or decision making nature (cf. McGrath, 1984). Thedegree to which group members adequately understandthe informational requirements of the decision makingtask should greatly affect their decision performance.The effective use of distributed information requiresthe exchange of distributed information, careful consid-eration of this information and its implications, and dis-cussion and integration of these implications. Thisprocess of exchange, consideration, and integrationhas been referred to as group  elaboration  of information(van Knippenberg, De Dreu, & Homan, 2004). Groupsseem to differ in the extent to which they recognize theimportance of elaboration of decision-relevant informa-tion for successful task performance.Research on jury decision making hints at the possi-bility that group members often insufficiently recognizethe need for information elaboration. Studies haveshown that juries can approach a decision task in at leasttwo ways. Juries using an evidence-driven style firstgather relevant evidence and then use the evidence incoming to a group judgment. Juries using a verdict-dri-ven style start with pooling their individual judgmentsand only after pooling group members’ judgments citeevidence in a piecemeal fashion organized by theverdicts (Hastie & Pennington, 1991; Kameda, 1991).Hastie and Pennington (1991) showed that in contrastto groups, individuals always proceeded in the evi-dence-driven manner when performing the same task,suggesting that the need to reach agreement with othersin addition to reaching an individual judgment maymotivate groups to use the verdict-driven strategy.Research by Wittenbaum, Stasser, and Merry (1996)also shows that group members may not recognize theneed for elaboration. They found that when group mem-bers believed they were about to participate in a groupdecision making session as compared to a group recallsession, they tended to focus more on information whichthey believed the other members would also have thanon distributed information. Our interpretation of thesefindings is that group members’ understanding of thedecision making task centered on the need to find com-mon ground more than the elaboration of information.Finally, also hinting at the role of task representations,Stasser and Stewart (1992) showed that groups thatbelieved there was an objectively correct solution tothe decision making problem made better use of theirdistributed information than groups that believed thatthere was no objectively correct solution. Although dif-ferent mechanisms may have been in operation (e.g., dif-ferential weighting of new information) and a secondstudy (Stewart & Stasser, 1998) was unable to W.P. van Ginkel, D. van Knippenberg / Organizational Behavior and Human Decision Processes 105 (2008) 82–97   83  completely replicate the results, we would propose thatthe perception that there was an objectively correct solu-tion gave rise to task representations suggesting a searchfor information to identify this solution.Although these studies at best only yield circumstan-tial evidence for the role of task representations, they areconsistent with the notion that misconceptions abouttask requirements may play a role in groups’ suboptimaluse of distributed information. In the present study, weput our proposition that the extent to which groupmembers have shared task representations that empha-size elaboration affects groups’ exchange, consideration,and integration of information to a more direct test bycomparing the performance of decision making groupsthat are stimulated to form task representations empha-sizing elaboration with the performance of controlgroups. In addition to the way group members perceiveor understand their task (i.e., the content of their taskrepresentations), we propose there is a second aspectof group members’ task representations that is of impor-tance for groups’ information elaboration—the extentto which team members realize that they share these taskrepresentations. The role of team realization of sharedness of taskrepresentations Group members may be unaware of other groupmembers’ task representations, even when these areidentical. That is, group members may share similar rep-resentations of their task without knowing they do.Arguably, shared cognition entails more than groupmembers thinking along highly similar lines. Implicitin the study of shared cognition is the notion that sharedcognition not only involves group members thinking thesame about something, but also includes a realization of these converging cognitions (cf. Kerr & Tindale, 2004;Klimoski & Mohammed, 1994; Rentsch & Hall, 1994;Tindale & Kameda, 2000). Indeed, it may be argued thatit is this team realization that moves shared cognitionbeyond a mere aggregation of individual cognition andrenders shared cognition a true group-level construct(Klein & Kozlowski, 2000). Yet, while realization isoften assumed when discussing the effects of shared cog-nition, the effects of realization are usually not put to thetest. In the present study, we define shared task repre-sentations as not only entailing actual sharedness of mental representations of the task, but also a realizationof this sharedness. To be able to substantiate our claimthat team realization can be important for group func-tioning, we address the role of team realization empiri-cally, and test the prediction that the effects of taskrepresentations emphasizing elaboration of decision-rel-evant information are stronger when group membersrealize more that they share these task representations.Realization is expected to make a difference because itshould remove psychological barriers to introducing anddiscussing distributed information. One reason forgroups’ biased information use lies in group members’tendency tofocus on information shared by all members,orconsistentwithothermembers’preferences,outoffearof being rejected (Gruenfeld et al., 1996; Nemeth, 1986;Wittenbaum, Hubbell,&Zuckerman,1999;Wittenbaum& Park, 2001). Furthermore, there are indications thatgroup members who mention arguments that conflictwiththecommongrouppreferencemaygetnegativereac-tions from other group members and are liked less byother members than members who mention argumentsthatareconsistentwithanoptionpreferredbymostmem-bers(Edmonson,Bohmer,&Pisano,2001;Nemeth,1986;Wittenbaum et al., 1999; Wittenbaum & Park, 2001).Thus, fear of rejection seems to inhibit informationexchange. Psychological safety (i.e., the perception thatit is safe to speak one’s mind in group interaction;Edmondson, 2003) may attenuate fear of rejection. Psy-chological safety may foster the expectation that peoplewill not sanction the discussion of new information thatmayleadthegroupawayfromanagreement(e.g.,discus-sion of information pro A, while most members alreadybelieve option B to be the best option) and subsequentlystimulate discussion of more ‘‘risky’’ information thatcould interfere with an emerging group agreement (cf.Edmondson, 1999; Kramer, Brewer, & Hanna, 1996).Thus, higher psychological safety in teams can fosterinformation elaboration.Team realization of task representations may fosterpsychological safety, because it provides group memberswith relevant knowledge to base expectations of othermembers’ behavior on. When group members realizethat they share task representations emphasizing theimportance of information exchange and discussion, thisprovides them with knowledge that other members rec-ognize the importance of sharing unique informationand viewpoints for the task at hand. When group mem-bers know that other members also value informationexchange and discussion, they likely will expect thatother members will not react negatively to group mem-bers coming forward with unique information. Thisknowledge may create a feeling of psychological safetythat stimulates group members to voice information. Hypothesis 1.  Groups engage in more informationelaboration and reach higher-quality decisions whengroup members have shared task representationsemphasizing information elaboration than when theyhold such shared representations to a lesser extent. Hypothesis 2.  The effects of task representations empha-sizing elaboration on information elaboration and deci-sion-making performance are stronger when groupmembers realize they share these representations. 84  W.P. van Ginkel, D. van Knippenberg / Organizational Behavior and Human Decision Processes 105 (2008) 82–97   An experimental test To establish the causality implied in our analysis, weput our hypotheses to an experimental test. Because toour knowledge this is the first study in which shared taskrepresentations and team realization are manipulated(and measured), replication of our key findings wouldsubstantially bolster the confidence in our findings.Therefore, we tested our hypotheses in a small prelimin-ary study as well as in a more full-blown experiment. Inboth experiments we compared control groups to exper-imental groups that received task instructions designedto engender task representations emphasizing elabora-tion of decision-relevant information. In addition, half of the experimental groups where made to realize thatthey shared these task instructions (and associated rep-resentations). We relied on audio–video data to codegroup information elaboration, because audio–videodata tend to be more reliable and provide a richer sourceof information than self-report data (Weingart, 1997).To substantiate our analysis in terms of the processesimplied, we also formulated an explicit hypothesis aboutthe processes translating the experimental manipula-tions in decision-making performance. Hypothesis 3.  Group information elaboration mediatesthe relation between shared task representations andgroup decision-making performance.In both experiments we tested Hypotheses 1–3. Inaddition, in the main study we also assessed the roleof psychological safety proposed to underlie the effectof team realization on information processing andperformance. Preliminary study Method Design and participants Just as the main study, the preliminary study had aone-factor design with three levels (elaboration-realiza-tion, elaboration, control). A total of 112 Dutch under-graduate students (74 men, 38 women) were assigned to28 four-person groups. Groups were randomly assignedto conditions. The majority of participants were busi-ness administration students (86.6%). Participantsreceived 10  €  (±12 US$) for participation. Experimental task  The task was a cooperative decision making taskinspired by the Towers Market task (Weingart, Bennet,& Brett, 1993). While the srcinal task was designed as anegotiation task, the adaptation was such that the cur-rent task was a purely cooperative decision making task.So, contrary to the srcinal task, each group memberrepresented all interests. The task concerns the organiza-tion of a small market center that contains a bakery, aflorist, and a greengrocery. Participants are told thatthey function as an independent advisory committeethat is to aid the three stores in making three interrelateddecisions about the temperature for the market center,the division of maintenance costs between three stores,and the organization of marketing campaigns. To dothis all participants were given information on the pref-erences of the three stores and on the relative impor-tance of the three issues to the stores. In addition theywere told to take the interests of all three stores intoaccount. For each issue, groups could choose from alimited number of options. Based on all available infor-mation a hierarchy in the quality of the decision optionsexisted (i.e., some combinations of decision optionsserved the interests of all stores better than others). Allmembers received some information on all three stores.The information items were partially based on anotheradaptation of the task (Beersma & De Dreu, 2002)and partially designed specifically for this study.Following prior research in group decision making,part of the decision-relevant information was given toall group members and part of the information wasgiven to only one of the group members (cf. Stasser &Stewart, 1992). Items of information necessary for mak-ing good decisions were divided between the groupmembers, so that for all members to be able to learnabout these items groups had to share the information.For every decision issue there were always three itemsof information that were crucial for reaching an optimaldecision. Each of these was uniquely assigned to one of the group members. For each decision issue group mem-bers received this crucial, unshared information on a dif-ferent store (e.g., group member 1 received unsharedinformation about the bakery on item 1, about the flo-rist on item 2, etc., while group member 2 received cru-cial information about the florist on item 1 and aboutthe greengrocery on item 2, etc.). To make the task morecomplex, some irrelevant information about the threestores was also included (this information was alwaysgiven to all members). Groups were told that the infor-mation the members received might differ. Experimental manipulations Individual task representations were manipulatedthrough written instructions. Group members in boththe elaboration-realization condition (i.e., the conditionwhere members shared task representations for informa-tion elaboration  and   were made to realize this) and theelaboration condition (i.e., the condition in which mem-bers did share task representations for information elab-oration, but were  not  made aware of sharing them)received information that explained what kind of taskthey were going to work on and what would beimportant for this task. As a rational for receiving this W.P. van Ginkel, D. van Knippenberg / Organizational Behavior and Human Decision Processes 105 (2008) 82–97   85
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks