Means vs. Ends- Implications of Process and Outcome Focus for Team Adaptation and...

8/7/2019 Means vs. Ends- Implications of Process and Outcome Focus for Team Adaptation and Performance-Woolley

1/16

OrganizationScienceVol. 20, No. 3, MayJune 2009, pp. 500515

issn 1047-7039 eissn 1526-5455 09 2003 0500

informs

doi 10.1287/orsc.1080.0382

2009 INFORMS

Means vs. Ends: Implications of Process and OutcomeFocus for Team Adaptation and Performance

Anita Williams WoolleyTepper School of Business, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213, [email protected]

Knowledge work frequently involves both the redefinition of desired outcomes and the specification of task processes.The relative emphasis that teams place on these issues early in work can lead members to become outcome focusedor process focused, with consequences for performance. This paper develops and explores a theory of how outcomefocus and process focus develop in teams and the implications of each for team adaptation and performance. Outcome andprocess focus were both observed and experimentally manipulated in 90 teams working on an open-ended task. Measuresof the teams performance as well as level of action identification and ability to adapt work processes point to an advantagefor outcome-focused teams in dynamic environments. Implications for the design and management of knowledge workteams are discussed.

Key words: team adaptation; team performance; action identification; knowledge work

History : Published online in Articles in Advance September 17, 2008.

In some organizations, failure to adapt to changingcircumstances can lead to reduced profits or marketshare. In others, it can have catastrophic consequences.The 9/11 Commission concluded that teams in the intel-ligence community missed advance indicators of theSeptember 11 attacks because they had neglected toadjust their work processes to todays more fluid envi-ronment with multiple adversaries. Instead, they hadbeen following Cold War-era procedures designed tofight only one or two primary adversaries (NationalCommission on Terrorist Attacks Upon the United States2004). Examples of the failure to adapt work pro-cesses in response to changing environmental conditionsabound from other settings, as well. Senior managementteams tend to overlook significant technological innova-tions and changes in customer demands while continuingto focus on core competencies (Tushman and OReilly1996). Cockpit crews follow the same preflight proce-dures even when a new climate demands different mea-sures (Gersick and Hackman 1990).

Why do teams fail to adapt to changing circum-stances? I argue that the presence of a focus on pro-cess, rather than outcomes, is one factor that inhibitsthe ability of teams to adapt in a dynamic environment.Process-focused teams allow their concerns about fol-lowing a process (or means) to take precedence overand constrain outcomes, where process involves iden-tification of the specific tasks that need to be com-pleted, the resources available for doing so, and thecoordination of tasks and resources among members(LePine 2005, McGrath 1984). In contrast, outcome-focused teams allow outcomes (or ends) to take prece-dence over and constrain process, where outcomes refersto the intended final products or results of the teams

work. Precedence is an important part of the definitionof each, in that it refers not only to temporal precedence,such as a focus on processes prior to outcomes in orga-nizing work, but also to importance, in that processesare more central to decision making about work. Beingfocused on one element does not preclude considera-tion of the other; instead, it is the relative emphasis oneach that has significant implications for the focus thatevolves and the ultimate consequences for team adapta-tion and performance. Although many factors can leada team to become outcome or process focused, includ-ing the personal tendencies of individual team mem-bers (Barrick et al. 1998), the existence of strong scriptsor routines in the work context (Gersick and Hackman1990, Newell and Simon 1972), or the way a team islaunched and encouraged to organize work (Ericksenand Dyer 2004), the current study will focus on the influ-ence of the teams beginning on the development of anoutcome or process focus.

How and why teams develop a focus on either out-comes or process can be of particular consequence toteams engaged in knowledge work. Knowledge workfrequently involves multiple desired end states alongwith multiple paths for arriving at each possible endstate (Campbell 1988, Janz et al. 1997, McGrath 1984,Quinn 2005), fitting the definition of a highly complex,creative, open-ended task (McGrath 1984). Knowledgework teams frequently define (or at least redefine) boththe outcomes and processes of their work for them-selves (Hackman 1987, Quinn 2005, Staw and Boettger1990). Although prior work has examined the beneficialeffects of both clear goals and processes for team per-formance (Earley et al. 1987, Weingart 1992, Weldonand Weingart 1993), there has been no examination of

500


2/16

Woolley: Implications of Process and Outcome FocusOrganization Science 20(3), pp. 500515, 2009 INFORMS 501

the effects of the relative emphasis of these elementsover time in teams that are at liberty to structure boththeir own processes and outcomes. I explore the extentto which a teams relative emphasis on processes or out-comes endures and the implications for its ability toadapt as challenges arise during its work. In the next

section, I discuss relevant theory and develop a set ofhypotheses about outcome and process focus in teams.I then present a laboratory study to examine the develop-ment of outcome and process focus, their stability overa teams lifespan, and their effects on team adaptationand performance.

Theory and Hypothesis Development

Means vs. Ends: Evidence for a General

Process/Outcome DistinctionIn a variety of settings, a distinction is made betweenthinking about the ends we want to achieve (outcomes)versus the means we use to achieve them (process). Inresearch and theory about this distinction at the organiza-tional level, it has generally been concluded that which ofthese dimensions is given priority is consequential for theaction that ensues. For instance, government and militaryintelligence organizations have been criticized for beingtoo process-centric or driven by their methods of infor-mation collection. The restructuring of national intelli-gence agencies is based in part on the ideal of a moreobjectives-focused, target-centric approach to collectionand analysis (e.g., with activities organized around thegroup or geography of interest) on the assumption thatthis will make the overall enterprise more adaptive andagile (Clark 2004). Similarly, the culture of innovationin commercial organizations can be characterized bywhether outcomes or processes dominate; organic orga-nizational cultures are more results oriented and out-ward facing, whereas mechanistic organizational culturesare more process oriented, hierarchical, and inward fac-ing (Burns and Stalker 1994). Beyond the specific orga-nization for innovation, entrepreneurial scholars arguethat as most organizations grow from being small andentrepreneurial to being larger and more bureaucratic,many of their strategic decisions become less driven byexternal opportunities and more driven by internal pro-cesses and the resources already owned or controlled(Stevenson and Jarillo 1991), with implications for theirability to adapt and change.

In contrast, at the team level, little research has beendone explicitly on how team members thinking aboutprocesses and outcomes relate to one another. Researchon the relationship between goals and task planning sug-gests a complementary relationship between the consid-eration of processes and outcomes (Locke et al. 1997,Weingart 1992, Weldon and Weingart 1993). In suchstudies, however, the goals are established by the experi-menters at the outset of work as an independent variable,and the relationship of task planning to performance

is examined in the context of these desired outcomes,which the team does not have the ability to change.Research on goal orientation distinguishes between theimplicit pursuit of a learning or skill development goalversus a performance goal (Diener and Dweck 1978) andthe implications of each for an individuals responses to

challenge and feedback. At the team level, these indi-vidual tendencies influence a teams learning orienta-tion, which is associated with more openness to feedbackand improved performance in teams that are struggling(Bunderson and Sutcliffe 2003). Although teams witha greater learning orientation may certainly engage inmore extensive consideration of either processes or out-comes, this line of work does not examine the effectsassociated with whichever element takes precedence.

The meansends distinction has been more explicitlyexplored at a much lower level of analysis than teams orlarger social systems: in individual-level cognitive neu-roscience. Research in cognitive psychology has shown

that different parts of the brain specialize in thinkingabout how (process) as opposed to what (outcomes)(Ungerleider and Mishkin 1982). The brains proce-dural or how pathway (Goodale and Milner 1992)is particularly important for acquiring and performingskills involving sequenceswhether the sequences areserial or abstract, or sensorimotor or cognitive (Squireet al. 1993, Ullman 2004). Meanwhile, the declarativeor what system (Goodale 2000) underlies the learn-ing, representation, and use of knowledge about factsand events (Ullman and Pierpont 2005). Research onmental imagery and cognitive styles suggests that a pref-erence for reasoning in the procedural system (how)

is slightly negatively correlated with declarative system(what) reasoning, leading to stable individual cog-nitive styles that can be characterized on this basis(Blajenkova et al. 2006, Kozhevnikov et al. 2005). Takentogether, this work suggests that an individual who tendsto reason in a manner that engages one of these subsys-tems is unlikely to shift easily to the other subsystem,leading that person to remain more focused on eitherwhat or how for at least a certain period of time.

Processes vs. Outcomes and Action

Identification in TeamsThe implications of the what versus how systemsof the brain become applicable to team functioning asmembers are cued by their environment to attend toprocesses or outcomes and develop norms about theways they will deal with these issues in their worktogether. In complex or ambiguous social situations,members look to one another for cues regarding howto behave in the situation and follow each others lead(Bettenhausen and Murnighan 1985, Hackman 1992).In newly formed teams, initial member commentary candirect other members attention to processes or outcomes,stimulating related thoughts and commentary as mem-bers follow conversational norms of reciprocity (Burgoon


3/16

Woolley: Implications of Process and Outcome Focus502 Organization Science 20(3), pp. 500515, 2009 INFORMS

et al. 1993). In ongoing teams or those coming froma shared context, understandings about standard oper-ating procedures or expected outcomes may already bein place (e.g., Woolley 2009). Whether self-generatedor imported from their context, early decisions aboutprocesses or outcomes assert a structural influence over

subsequent decisions the team makes. For example, aprocess-focused research group might decide to pursueits work by first establishing a weekly meeting of adetermined length with the entire group, accomplish-ing whatever outcomes it can within those parameters.In contrast, an outcome-focused group might first deter-mine its objectives for a particular time frame and thendetermine whether meetings are necessary and, if theyare, when they should be held and who should attend.As subsequent decision points about a groups progressare encountered, I argue that the early pattern of relativeemphasis on outcomes or processes is replicated, withthe process-focused group constraining their outcomes to

their weekly meeting schedule and the outcomes-focusedgroup constraining their processes to meet their objec-tives. In this manner, a teams outcome or process focusis perpetuated over time.

Hypothesis 1. Early cues (e.g., conversations, in-structions, expectations) encourage a relative emphasison process or outcomes that is sustained over time in ateams work.

A focus on process or outcome is further perpet-uated in a team through the inherent influence ofeach orientation on the action identities team mem-bers adopt in their work. Work on action identification

(Vallacher and Wegner 1987) has shown that individ-uals can identify actions as low-level specific activi-ties (e.g., I am designing an advertisement to sell thisproduct) or in higher-level terms that encompass mul-tiple specific alternative activities for enactment (e.g.,We are generating as much profit from this productas possible). The level at which people identify theiractions is highly influenced by cues provided by thetask context (Vallacher et al. 1989). Process-focusedteam discussion can involve identifying specific tasksand subtasks, assigning tasks to members, and specify-ing how these activities will be coordinated across peo-ple and/or over time. Members come to identify their

actions at a low level, reasoning in terms of specifictasks and their own personal role rather than higher-level team goals. In contrast, outcome-focused team dis-cussion centers on identifying desired results of workand the internal or external criteria for success. Norma-tive pressures serve to create uniformity in membersperceptions and discussion (Sherif 1936) while reinforc-ing and maintaining the level of action identification inthe team, as members resist discussion that changes thelevel at which their actions are identified. For instance, amember of a process-focused product development team

who questions individual tasks by trying to discuss thehigher-level meaning of what the team is really trying toaccomplish (e.g., Will advertising increase our profit?)will be seen as slowing the team down, whereas a mem-ber of an outcome-focused team that tries to get lower-level and tactical (e.g., Should we hire Michael Jordan

as a spokesman?) will be seen as similarly inappropri-ate. The cues that members give to sanction one anotherfor changing their focus can be subtle, such as ignor-ing someones off-level comment, or can be more direct,to the extent that members experience such commentsas attempts to change their own personal work (e.g.,the team member from marketing might respond, If wearent going to advertise, why am I on the team?). Inthis manner, team norms maintain the level of actionidentification that is dominant in the team, which servesto further reinforce the teams task focus.

Hypothesis 2. Actions are identified at a significantlyhigher level in outcome-focused teams than process-focused teams.

Outcome Focus, Process Focus, and Their

Consequences for Adaptation and Performance

In considering the discussion that leads to an outcomeor process focus in a new team, it is important to keepin mind that many teams tend to forgo any task discus-sion whatsoever unless they are explicitly encouraged(or even mildly coerced) to engage in it (Hackman et al.1976, Hackman and Wageman 2005, Shure et al. 1962).Given this tendency, any discussioneither process oroutcome focusedis likely to be beneficial when com-

pared with no organized discussion, because any plan-ning activities will help teams develop a better sharedbasis for proceeding with work, regardless of the groupsfocus (Klimoski and Mohammed 1994, Weingart 1992).When left to their own devices, teams can becomeunfocused by either failing to have any organized dis-cussion or by allowing members to develop differentfoci, which inhibits engagement in shared planning andleads members to work at cross-purposes (Dougherty1992, Woolley et al. 2008). In other instances, teamsmay begin to discuss the issues necessary for organiz-ing work but then become distracted by the work itselfand instead fall back on in-process planning (Weingart

1992). In such cases, getting team members to discussand come to agreement on core issues is likely to yieldbenefits when compared with no intervention, becausegetting a team to discuss their work together in anycapacity can allow them to be relatively more orga-nized and productive than not doing so (Hackman et al.1976).

Hypothesis 3. Interventions prompting either out-come- or process-focused discussion improve team per-formance when compared with the results where therehas been no intervention.


4/16


Although interventions to get a team focused on eitheroutcomes or processes can be beneficial in stable envi-ronments when little change is needed, they are likelyto yield different effects in dynamic environments inwhich it is necessary for teams to respond to unex-pected events. The ability of individuals and teams to

adapt to changing circumstances is critical to perfor-mance in a variety of contexts (Kozlowski et al. 1999,LePine 2005, Pulakos et al. 2000). The types of adapta-tions a team needs to make can take a variety of forms.Pulakos et al. (2000) identified eight different dimen-sions of adaptive performance, among them the abilityto solve problems creatively and the ability to deal withchanging work situations. Both process focus and out-come focus can potentially enhance a teams ability toadapt, but in slightly different ways.

Clear processes can have significant benefits, partic-ularly in groups where membership changes (Carley1992, Levine and Choi 2004, Rao and Argote 1995).

When reduced errors and reliable output are highlydesirable, well established procedural routines can sig-nificantly enhance team performance (Gersick andHackman 1990). Clear roles and work processes canhave significant emotional benefits, reducing stress anduncertainty for team members (Dollard and Winefield1994, Rizzo et al. 1970, Smith 1957, Sperry 1998). Welldefined roles also can reduce cognitive burden, freeingup team members to focus only on role-relevant infor-mation (Brandon and Hollingshead 2004, Hollingshead2000, Lewis et al. 2007, Moreland and Argote 2003,Wegner 1987) and generally work more efficiently(Moreland et al. 1996). Process-focused teams are likely

to segment their task into parts associated with mem-ber roles, leading them to be well prepared to reflecton internal processes (Arrow and McGrath 1993) andto make the accommodations needed for membershipchange, especially if the change has been anticipated(Moreland and Argote 2003). By contrast, outcome-focused teams may not spend enough time specifyinghow work will be accomplished and defining group rolesand thus experience process loss as members try toorganize themselves while work is in progress (Steiner1972, Weingart 1992). Membership change will be moredifficult for outcome-focused teams, because the lossof a person will be associated with gaps in informa-

tion and responsibility that cannot be easily identifiedor replaced (cf. Choi and Thompson 2005, Levine andChoi 2004, Lewis et al. 2005). Therefore, I hypothesizethat:

Hypothesis 4. Process-focused teams outperformoutcome-focused teams under conditions of membershipchange on open-ended, creative tasks.

Whereas process-focused teams will be more effec-tive in adapting to membership change, outcome-focusedteams will be better at dealing with task change.

Although the low-level action identifications embracedby process-focused team members allow new teammembers to more easily identify their tasks, low-levelaction identifications are not as flexible as high-levelaction identifications for facilitating overall task change.Maximizing profit (a high-level action identification)

can be accomplished in multiple ways, whereas design-ing an advertisement (a relatively lower-level actionidentification) does not lend itself to as many alter-native approaches (Vallacher and Wegner 1989). In ateam that has defined its work by deconstructing itinto a series of specific and/or individual-level tasks, itis difficult for members to identify problems that cutacross members roles or come to agreement on howto address such issues (Moreland and Levine 1992).In contrast, the discussion of potential outcomes inoutcome-focused teams results in members sharing agreater amount of information about the teams over-all end goal (Kerr and Tindale 2004, Latham et al.

1994). The resulting higher level of action identifica-tion by members allows outcome-focused teams to iden-tify multiple approaches to accomplishing their desiredoutcomes; maximizing profit can happen many dif-ferent ways and need not involve advertising at all. Asthe team encounters difficulties (e.g., Ad prices havegone sky high! or Shipping costs are astronomical!)an outcome-focused team has a better basis for refram-ing and changing its approach than does a process-focused team. Thus, outcome-focused teams will exhibitgreater flexibility in adapting their performance strat-egy to changing task conditions. Furthermore, as

depicted in Figure 1, the increased adaptation behaviorobserved as a result of being outcome versus processfocused will be mediated by the teams level of actionidentification.

Hypothesis 5. Outcome-focused teams outperformprocess-focused teams under conditions of task changeon open-ended, creative tasks.

Hypothesis 6A. Outcome-focused teams exhibitgreater ability to adapt their activities in the course oftheir work than process-focused teams when working onopen-ended, creative tasks.

Hypothesis 6B. Teams level of action identificationmediates the relationship between outcome focus andadaptation.

Figure 1 Model of Hypothesized Effects of Outcome and

Process Focus on Action Identification, Adaptation,

and Team Performance

Outcome focus

Adaptation PerformanceAction

identificationProcess focus

+


5/16


Method

Participants

The experiment was conducted with 90 3-person teamscomposed of male and female undergraduates who wererandomly assigned to 1 of the 6 conditions of a 2(process- versus outcome-focused manipulation) 3

(midpoint material loss versus midpoint membershipchange versus no midpoint change) design with a sev-enth condition that received no manipulation or midpointchange (the control condition). All participants eitherwere paid or received course credit for their participationand were randomly assigned to one of two teams duringeach experimental session (within compensation type),which were in turn randomly assigned to conditions.1

Of the teams, 15 were all female, 18 teams were allmale, and the rest were mixed. There were no system-atic differences in performance between the two types ofcompensation (F1 89 = 124, p = ns) or on the basisof gender composition (F

2 87= 003, p = ns) in the

experiment overall. Therefore, all analyses will includeboth compensation types and all gender compositionstogether. To foster motivation among all participants, a$300 prize was offered to the team with the highest scoreat the end of the study. This prize was paid at the con-clusion of data collection.

Task

Participants were asked to work on a creative, open-ended task together with their team members dur-ing a one-hour laboratory session. Specifically, theywere asked to use a set of building blocks to build a

house, garage, and swimming pool, which were scoredaccording to a set of complex scoring criteria (seeAppendix A). The scoring of the task was intentionallycomplex and devised to force trade-offs; teams couldnot maximize all point sources simultaneously (muchlike a product development team usually cannot opti-mize on both price and performance) and thus neededto redefine the task for themselves and determine theirown intended outcomes as well as processes. Such a taskis very similar to the open-ended tasks typically givento knowledge workers, while not requiring the applica-tion of specialized expertise (a distinguishing feature ofknowledge work), allowing for the use of ordinary par-

ticipants in a laboratory.All teams except those in the control condition

received an intervention at the beginning of their workto focus their discussion on outcomes or processes.Halfway through their work, some teams experiencedeither a membership change or a loss of critical buildingmaterials. This allowed for the evaluation of the teamsresponses to a dynamic task environment in which mem-bers and resources can change unexpectedly.

Each team member was given an identical set of writ-ten task instructions, and the team was shown a video

detailing the criteria that would be used to compute ascore for the teams products. The final structures werescored on the basis of their size, quality (e.g., whetherthey would hold together when lifted, flipped over,and/or dropped), and the inclusion of features that quali-fied for bonus points (such as parking spaces included in

the garage). Each structure also featured some requiredelements; the roof of the house and garage as well asthe floor of the pool had to be made from blue blocks,and the foundations of the house and garage had to bebuilt from white blocks. All requirements were spelledout in detail to the teams both in the video and in thewritten instructions. Full task instructions can be foundin Appendix A.

Procedure

Each team worked alone in a private laboratory roomset up with a table, three chairs, and a video player. Thetable in front of each participant was labeled A, B, or C

as a means of identifying which participant to remove(for teams that experienced a membership change). Par-ticipants seating positions were decided randomly, andthe placement of these letters on the table was rotatedfor each experimental session. All teams were video-taped with the knowledge and consent of all participants,and conversations were transcribed for the purposes ofcoding and analysis.

In the process/no midpoint change condition, teamswere first shown the instructional video for the task.After the video, teams were instructed to use the next10 minutes to work on a preliminary task, whichserved as the process focus manipulation. This was a

worksheet that led the team through a discussion of rolesand assignments for the task, deciding who would focuson each of the structures as well as brainstorming, andassigning what other tasks needed to be done in thecourse of their work. Teams recorded their decisions onthe worksheet, which was returned to the experimenterbefore they began their 40-minute building period. Afterthe teams had worked for 20 minutes, they were givena 20-minute warning. At the 35-minute mark the teamwas given a 5-minute warning, and at the completion oftheir work the experimenter took the structures into anadjacent room to measure and photograph them whilethe participants individually completed a post-task sur-

vey without communicating with their teammates, afterwhich they were debriefed and released.

In the process/membership change condition, the firsthalf of the session proceeded exactly as it had in the pro-cess/no midpoint change condition. After the teams inthis condition had worked for 20 minutes, however, theexperimenter entered the room with a fourth participant.She explained that the participant seated in chair A wasneeded to work on something else in the other room, butthat this new person was being brought to take his/herplace. The team was informed that this new member was


6/16


familiar with the instructions for the task but that theyneeded to fill the new person in on what he/she wouldwork on. The new participant was a bona fide researchparticipant, not a confederate, and had viewed the sameinstructional videotape and received the same writteninstructions as the original team members prior to enter-

ing the room. The new member then entered the roomand sat in Member As chair, and the former Member Aleft the room with the experimenter. The remainder ofthe session was conducted in exactly the same manneras the process/no midpoint change condition.

In the process/material loss condition, the first half ofthe session proceeded exactly as it had in the other twoprocess conditions. After 20 minutes, the experimenterentered the room and explained that she needed to takesome of the blue blocks. Without additional explana-tion, she proceeded to remove a specific number of theteams blue building blocks, equivalent to one-third ofthe total they had available. After removal of the blocks,

the team was informed that they had 20 minutes remain-ing in their building period. The remainder of the sessionwas conducted in exactly the same manner as process/nomidpoint change condition.

In the outcome/no midpoint change condition, afterthe instructional video, the team was given a prelimi-nary task to work on for 10 minutes, which comprisedthe outcome focus manipulation. This was a worksheetthat led the team through a review of the relative pointvalues for the various structures they were being asked tobuild and encouraged them to determine their prioritiesand desired outcomes for the task by rank-ordering thethree structures and estimating the points they wanted

to earn. Teams recorded information in response to eachstep on the worksheet and were required to turn in acompleted worksheet before they began their 40-minutebuilding period. The remainder of the session was con-ducted using the same procedure as the process/no mid-point change condition.

In the outcome/membership change condition, the firsthalf of the session was conducted exactly as describedfor the outcome/no midpoint change condition, and thesecond half (from the midpoint through completion) wasconducted identically to the process/membership changecondition.

In the outcome/material loss condition, the first half

of the session was conducted exactly as described forthe other outcome/no midpoint change condition, andthe second half (from the midpoint through completion)was conducted identically to the process/material losscondition.

In the control condition, after the instructional video,the teams were given a worksheet directing them tointroduce themselves and spend 10 minutes reviewingand discussing the task. No additional directives weregiven regarding discussion of processes or outcomes.The remainder of the session was conducted using

the same procedure as in process/no midpoint changecondition.

All participants were debriefed in writing and verballyqueried about their observations regarding the experi-ment at the conclusion of each session. Although therewas much speculation regarding the reason for invoking

a membership change or removing building materials,there was no indication that any team suspected the roleof the opening exercise in enhancing or impairing theirability to deal with these obstacles.

Measures

Control Condition Focus Measures

Two measures were developed to provide indices of therelative attention that teams in the control condition paidto outcomes versus processes over the time they workedtogether. The control condition teams attention to theseissues is of particular interest in addressing Hypothe-

sis 1, because these teams were not directed to considertask processes or outcomes and, thus, their focus evolvedorganically.

Attention to Outcomes and Processes.2 Three re-search assistants who were blind to the experimentalhypotheses independently viewed the videotapes of eachteams initial 10-minute planning period and coded theamount of attention that teams gave to each of the fol-lowing 4 issues on a 7-point scale, from no attentionat all to extensive attention: (1) scoring and the pointvalues of different building features; (2) implications ofdifferent designs for the final score and what the team

wanted to build; (3) nuts and bolts of how to build thedifferent structures (e.g., specific techniques for connect-ing layers of blocks, how to satisfy the building codes,etc.); and (4) assignment of member roles and/or alloca-tion of time to each structure. Additional research assis-tants (2) watched tapes of the 5-minute period after thetemporal midpoint (minutes 2025) of work for eachteam and made the same ratings, because this was atime when all teams engaged in some additional discus-sion of their work, whereas discussion during all othertime periods varied considerably and was minimal formany teams. The first two items were summed to pro-vide an index of the amount of attention given to out-

comes, whereas the sum of the latter two items providedan index of attention to process. Raters were trained inthe use of the coding instrument as a group by view-ing tapes of pretest teams that were not part of thisdata set, then completed their evaluations of the teamsin the sample independent of one another. The scalesshowed acceptable levels of reliability for both outcomefocus (ICC1 = 065, p < 005; ICC2 = 089) and processfocus (ICC1 = 054, p < 005; ICC2 = 073). Ratingswere standardized and then averaged across raters foranalysis.


7/16


Self-Report Process and Outcome Focus. The surveyadministered to participants at the conclusion of theteams work included items measuring outcome and pro-cess focus. The scales provide an opportunity to evaluatehow members conversations throughout the work periodare reflected in their own understanding of the differ-

ent elements of their task. Specific scale items can befound in Appendix B. Team members responses showedacceptable levels of reliability for both process focus(ICC1 = 052, p < 005; ICC2 = 077) and outcomefocus (ICC1 = 050, p < 005; ICC2 = 072). This mea-sure was used to examine the consistency of outcome andprocess focus in teams over time by relating the teammembers post-work scores to the observers measures ofteam focus earlier in the work period and also served asa manipulation check of the outcome- and process-focusexercises in teams receiving those interventions.

Observational Measures

Action Identification. Independent raters (2) viewedthe video recordings of the teams working after their ini-tial, 10-minute discussion and during the first 20 minutesof their work together (prior to any midpoint manip-ulations) and evaluated the level at which the teamsdiscussed and identified actions. These raters were notinvolved in evaluating any of the other observationalmeasures in the study. Individual team member com-ments were evaluated on a scale of one to five for theextent to which they identified actions at a low level bydiscussing individual parts of the task relative to theirown role and resources versus discussing higher-level

team activities. Raters were trained together but thenmade their ratings independently. The raters evaluationswere standardized before being averaged together foreach team in order to prevent weighting any one evalu-ations contribution to the composite score due to vari-ances in the ratings. The internal-consistency reliabilityfor the observers rating of each teams level of actionidentification was ICC = 086, p < 005.

Process Adaptation. Two procedural suggestionswere implicit in the design of the taskthe correspon-dence of the number of assigned structures to the numberof team members (three) and the provision of interlock-ing blocks along with the inclusion of a quality rat-

ing as part of the teams performance measure. Thesetwo factors suggested a work plan in which each struc-ture is built separately and blocks are hooked togetherto make the structures as sturdy as possible. However,given the nature of the scoring criteria, plus the inclusionof large, flat roof pieces whose bottom connectors weredeficient, the optimal solution actually involved combin-ing the three structures into one to maximize materialusage (instead of building them separately) and layingblocks end-to-end rather than interlocking them to maxi-mize structure size (which was much more valuable than

structure quality). Both of these necessary adjustmentsin the teams approach required them to adapt their workprocesses as they discovered these issues and possibil-ities. Two raters reviewed videos and photographs ofthe teams work and coded teams for whether or notthey (a) combined structures and (b) used their materials

in ways other than interlocking them. A team receiveda 1 if they made one of these adaptations, a 2 if theymade both adaptations, and a 0 if they did neither.The internal-consistency reliability for the observersinitial rating of each teams flexibility was ICC =082, p < 005; all disagreements were discussed andresolved.

Problem Adaptation. Teams in any of the sevenexperimental conditions could make the process adap-tations described above, but another important form ofadaptation is that which occurs in response to problems(Pulakos et al. 2000). Thus, another measure of adap-tation was developed specifically for examining teamslosing a third of their critical building materials halfwaythrough their work. The materials that were removed(blue blocks) were required for all of the structures,and the realization of the difficulty associated withtheir removal was expected to prompt discussion amongteam members regarding the allocation of the remain-ing materials. The 5 minutes after material removal werecoded for both discussion and alteration of strategy inresponse to this manipulation. A team received a 1if they merely noted the difficulty but made no signifi-cant change and a 2 if they noticed the problem anddiscussed a change in strategy as a team. If they did

neither of these things, they received a 0. All con-versations were coded independently by two researchassistants who were blind to the experimental conditionand who were not involved in the coding of processadaptation. The internal-consistency reliability for theobservers initial rating of each teams flexibility wasICC = 087, p < 005; all disagreements were discussedand resolved.

Task Performance. In addition to earning points (cal-culated in U.S. dollars) based on the cumulative valuefor the size of their structures, their quality, and aestheticappeal, teams were also assessed penalties for violatingany of a number of complex building codes. The teamswere not told their score on the task prior to complet-ing any post-task survey measures and were not givenany information about their performance relative to otherteams until the study was concluded.

ResultsDescriptive statistics and intercorrelations for all obser-vational measures can be found in Table 1 for the anal-ysis of control condition teams and in Table 2 for teamsreceiving experimental manipulations.


8/16


Table 1 Means, Intercorrelations, and Reliability Statistics for

Control Condition Observational and Survey

Measures

1 2 3 4 5 6

1. Beg. outcome 089

focus

2. Midpoint 059

090outcome

focus

3. End outcome 031 058 082

focus

(survey)

4. Beg. process 060 063 031 090

focus

5. Midpoint 075 047 059 074 090

process

focus

6. End process 035 053 031 035 044 060

focus

(survey)

Minimum 109 130 250 122 131 330

Maximum 140 125 650 153 162 700Mean 0 0 483 0 0 5 25

SD 100 100 120 100 100 116

Notes. Values on diagonal are interrater reliability (ICC) for the

measures.Correlation is significant at the 0.10 level (one-tailed).Correlation is significant at the 0.05 level (one-tailed).Correlation is significant at the 0.01 level (one-tailed).

Manipulation Check

As a check of the focus manipulation, outcome and pro-cess focus were evaluated using a post-task survey mea-sure. Teams in the process-focused conditions scored

significantly higher than those in the outcome-focusedcondition on the post-task measure of process focus(M= 587, SD = 102 versus M = 506, SD = 110,t80 = 342, p = 00001, one-tailed, Cohens d = 077)whereas teams in the outcome-focused condition scoredhigher than the process-focused team on the outcome

Table 2 Means, Intercorrelations, and Reliability Statistics for

Observational Measures N= 90

1 2 3 4

1. Process adaptation 082

2. Problem adaptationa 028 087

3. Action identification 026 016 086

4. Final score 025 044 022

Minimum 0 0 110 2160

Maximum 200 2 169 42480

Mean 054 115 000 18076

SD 064 081 100 7257

Notes. Values on the diagonal are interrater reliability (ICC) for the

measures.aBased on the 40 teams that lost materials midway through their

work.Correlation is significant at the 0.10 level (two-tailed).Correlation is significant at the 0.05 level (two-tailed).Correlation is significant at the 0.01 level (two-tailed).

focus measure (M = 505, SD = 061 versus M = 480,SD = 066, t80 = 142, p = 008 one-tailed, Cohensd = 040). Because the discussion activity was con-ducted before teams began work on the task and the sur-vey was administered after work was completed, theseresults are also interpreted as preliminary support for

Hypothesis 1, predicting that the early attention thatteams give to process versus outcome issues is predic-tive of their ongoing emphasis on these elements as theirwork progresses.

Process and Outcome Focus Over Time

Figure 2 displays a cross-lag panel analysis to examineHypothesis 1, predicting that early emphasis on processor outcomes predicts a teams relative emphasis on eachover the course of its work. The teams in the controlcondition are uniquely suited to such an examination,because they received no external directives to consideroutcomes or processes at any point in their work. As

shown in Figure 2, attention to outcomes and process atthe beginning of work are significantly correlated withthe amount of attention given to the same issue at thetemporal midpoint, which are in turn correlated with thelevel of focus on each on the corresponding survey mea-sure at the conclusion of work. Focus on each elementis negatively correlated with focus on the other elementboth at the same time points and at later points in time,supporting the notion that teams do not generally movefluidly back and forth between these categories of focalissues. Taken together, these findings are interpreted asstrong support for the notion that teams develop and sus-

tain a process or outcome focus over the course of theirwork.In looking more closely at how the control condition

teams worked, it appears that process and outcome focusdevelop largely through the effect that team membersearly comments have on the attention and subsequentcommentary of other team members. For the purposesof comparison, these control condition teams were clas-sified as outcome- or process-focused if they had a con-sistently higher score on the outcome or process focusmeasures at all three time periods measured. By thisstandard, five teams were process-focused, three wereoutcome-focused, and two did not exhibit a clear out-

come or process focus according to these criteria. Inmost of the outcome- and process-focused teams, earlycomments primed similar comments from other teammembers on similar issues, and most of the remainingconversation followed suit with the early pattern. Thefollowing is an example of the very beginning of theconversation in a process-focused team:

C: What do you want to do first?B: Do you all want to work on the same thing together,

or delegate?A: Sure.


9/16


Figure 2 Relationships Among Ratings of Outcome and Process Focus at Beginning, Midpoint, and Conclusion of Work Within

Control Conditions Receiving No Interventions

Beginning Midpoint Conclusion

Beg. processfocus

Beg.outcome

focus

Midpointprocessfocus

Midpointoutcome

focus

0.44^

0.31^

0.63*

0.53*

0.58*

0.47*0.60*

0.75**

0.59*

0.59*

0.74**Process

focussurvey

Outcomefocussurvey

Notes. Correlation is significant at the 0.10 level (one-tailed).Correlation is significant at the 0.05 level (one-tailed).Correlation is significant at the 0.01 level (one-tailed).

C: One thing we could do first is to separate all thepieces I think for the pool we need to know whichone is the we need to do it all one solid color.

B: Oh yeah, right.A: Ill separate the whites out.C: Ill read the specifications.B: Im just wondering if its necessary we could just

start with the pool, so pull out all of the blue

ones pick out the ones it looks like.C: So we want these [blocks]?A: We might as well save the thin ones for the roofs and

use the other ones.B: Theres another one A: The walls should be one knob thick and three blocks

high.B: What color do we want the walls to be?

In this example, the team demonstrates a clear preoc-cupation with how they will conduct work before havingsaid anything about what they will build. They start dis-cussing how to divide the work among them, how to orga-nize materials, in what order to focus on each element,

and the techniques to use for building the structures.In contrast, outcome-focused teams began their work

by focusing on more outcome-related elements of thetask, such as understanding scoring guidelines or explor-ing the kinds of building materials available and theimplications for what they could create. The followingoutcome-focused team started by discussing strategiesfor earning the most points for their structure:

B: So the bonus is going to be given to the group thatcumulatively, of all the structures, has the highestvalue the pool has $50 per square foot.

A: 6 by 12, 12 times 50 is what Okay, so the area of2 of these, this much, is worth the bonus If weare talking about $20 per square foot, so, relatively,the bonus

B: We could do both [aesthetic bonus and the pool].A: Okay, what occurred to me is like blue seems to be

at a premium because the pool C: Yeah, the pool.

A: At the same time, the roof can go over multiplestructures.

C: [We should build a] tall skinny house.

In contrast to the team discussed previously, this teambegan its conversation by thinking about their outcomehow to maximize the points they could earn. Theymoved from talking about point values to discussingthe design of their structures, with suggestions such asbuilding a tall skinny house or combining structuresby putting the roof over multiple structures (and thusmaximizing their points). Comments about materials orbuilding processes were directly related to the outcome

they were creating.Group norms served to maintain a teams focus.

Sometimes these norms were expressed indirectly, as inthe following exchange in which Member A kept try-ing to get the team to think how to maximize outcomes(related statements in italics below) while Members Band C ignored him and were more focused on workprocess:

B: You know what we could do, one of us could do each(structure)

C: You want to split up?


10/16


A: Well, if we do that were not going to have all of thepieces to use are we?

B: I dont know.C: Well, that is an interesting thing.A: Well, if we want to use as many pieces as possible in

order to get a lot of points B: We could kind of divide up, and something could be

mostly red and another thing mostly blue.A: Well my question is do you lose a lot of square

footage though.

B: I dont know well.C: Well I think it makes sense that we could do one at a

time, but, lets just say what do we have 40 minutesto do the whole thing? So lets just say we have 30minutes or 10 minutes to do each one.

B: Okay, yeah, well thats what I was worried about,coming to 20 minutes or something gets left out.

At this point in the discussion, Member A gave up andwent along with the plans of B and C. In another group,one member was more overt in her discouragement when

another started to stray from the discussion of processesto raise issues related to team outcomes:

B: The thing is, if we want to get things started, howshould we start I mean, should we all start on thishouse, or should each person take a part?

C: We should coordinate the colors though B: Yeah, you want to divvy them up?A: Oh, you know what we should do? Like the supplies,

like put what is the same here, and the same here, sowe could each get to them.

C: Do you want to read back the rules?B: Okay, Ill read off the rulesfor the house, $100 A: You know you shouldnt get really preoccupied with

the score until the end. Thats something we can addon later.

C: Well, yeah, not the score but the rules.

In this team, members wanted to talk about therules, which governed the specifics of how they wouldbuild their structures, without complicating the discus-sion with scoring criteria, thus keeping the team dis-cussion to a fairly low level of action identification. Inthis instance, Member A acquiesced and the team devel-oped a strong process focus as indicated by all othermeasures.

Action Identification, Adaptation, and Performance

Hypothesis 2 predicted a significantly lower level ofaction identification among process-focused teams com-pared to outcome-focused teams, and vice versa. Themean ratings for teams on this and all observational andperformance measures for the six experimental condi-tions are displayed in Table 3. As predicted, outcome-focused teams did exhibit a significantly higher levelof action identification than process-focused teams:M= 052, SD = 088 versus M = 053, SD =062; t80 = 623, p < 00001 two-tailed, Cohensd = 139.

Table 3 Means of Observational and Performance Measures

by Condition

Action Process Problem Final

Condition identification adaptation adaptation score

Process/no midpoint

change (n = 10)

M 031a 020a 24,611bSD 069 042 5,928

Process/membership

change (n = 10)

M 092b 010a 19.386aSD 068 032 6,479

Process/material

loss (n = 20)

M 047ab 035a 070a 17,091aSD 096 052 067 3,196

Outcome/no midpoint

change (n = 10)

M 072c 092b 24,954bSD 032 067 4,117

Outcome/membershipchange (n = 10)

M 060c 100b 19,058aSD 080 067 6,593

Outcome/material

loss (n = 20)

M 041c 090b 160b 23,428bSD 064 042 070 7,308

Control condition

(n = 10)

M 020c 010a 17,982aSD 093 032 5,732

Notes. N= 90 groups. All means in each column that do not share

a subscript are significantly different at p < 005 (one-tailed).

Final team performance scores were analyzed usinga 2 3 ANOVA to examine the effects of the manip-ulations on task performance. The results are not sig-nificant for the process or outcome focus manipulationalone (F 1 85 = 118, n.s.) but do yield significantresults for the midpoint material loss versus member-ship change manipulation (F 2 85 = 340, p = 004,eta = 025) and the interaction of focus and midpointchange (F 2 85 = 390, p = 002, eta = 027), withoutcome-focused teams that lost materials outperform-ing all other teams. All additional data analyses wereconducted as planned comparisons examining the rela-

tionships between the cells of the experimental designthat are relevant to each prediction, as recommended byRosenthal and Rosnow (1991).

Hypothesis 3 predicted that teams receiving eithera process- or outcome-focused intervention in the nochange conditions would perform significantly betterthan control condition teams not receiving an inter-vention. A contrast comparing the results of controlcondition teams to those of outcome- and process-focused/no midpoint change conditions supports thishypothesis: t85 = 254, p = 001, Cohens d = 055.


11/16


Additionally, in the absence of midpoint difficulty ormembership change, there is not a significant differencein the performance of outcome- or process-focused teams(t85 = 015, p = n.s.). A closer look at specific com-ponents of the performance measures provides additionalinsight into the differences between the control condi-

tion teams and the process- and outcome-focused teams.The third floor of the house was among the most valu-able sources of points but the most difficult for teams toattain, because it required working fast to get first andsecond floors built to provide the base for higher levels.Only organized and productive teams managed to buildhigher floors before their time elapsed. Control condi-tion teams added significantly less space higher up intheir structures than process- or outcome-focused teams:M= 3680, SD = 6220, versus M= 8370, SD = 2568for control condition versus process-focused/no midpointchange teams and M= 7000, SD = 2856 for outcome-focused/no midpoint change teams; t85 = 198, p =005, Cohens d = 043. This suggests that teams in bothconditions receiving an intervention got to work morequickly and worked more efficiently than control condi-tion teams receiving no intervention, which allowed theteams with interventions to build larger structures withinthe time allowed.

Hypothesis 4 predicted significantly better perfor-mance by process-focused teams than outcome-focusedteams under conditions of membership change. Asshown in Table 3, the data in this study do not supportthis hypothesis (t18 = 012, n.s.). Hypothesis 5 pre-dicted a performance advantage for the outcome-focusedteams in the face of task problems (here introducedby material loss) and was supported. Outcome-focusedteams that lost materials at the midpoint earned a higherscore overall than process-focused teams (t38 = 355,p = 0005, Cohens d = 115).

Hypothesis 6A predicted that outcome-focused teamswill exhibit significantly greater ability to make processadaptations in their work on the task than process-focused teams overall, which is supported. Outcome-focused teams exhibited significantly more processadaptation overall (t80 = 591, p < 0001, Cohensd = 132), and outcome-focused teams losing materi-als midway through work exhibited significantly bet-ter problem adaptation than process-focused teams(t38 = 293, p = 0009, Cohens d = 138). Further-more, mediational analysis confirmed Hypothesis 6B,predicting that team members action identificationmediates the effect of outcome focus on process adap-tation within teams experiencing midpoint difficulties.Among these teams, outcome focus had a significantdirect effect on adaptation ( = 025, t58 = 198,p = 005) as well as a direct effect on action identi-fication ( = 055, t58 = 495, p = 0001) whereoutcome-focused teams, as expected, exhibited signif-icantly higher-level action identification than process-focused teams. Action identification also had a direct

effect on adaptation ( = 034, t58 = 279, p = 0007),and, when the effects of the action identification medi-ator are taken into account, outcome focus of the groupno longer has a significant direct effect on adaptation( = 003, t = 019, p = ns). A Sobel test indicated thatthe difference between the direct and mediated effects

of outcome focus on adaptation are significant: z = 234,p < 005.

DiscussionWhereas existing research implies a complementaryrelationship between the work of clarifying outcomesand clarifying processes, this paper offers a different per-spective. Early team interactions can send knowledge-work teams down a path of emphasizing either taskoutcomes or work processes. Because of the cogni-tive separation of these elements of work as well asthe normative pressures manifested within teams, mem-

bers do not move fluidly back and forth but remainlargely focused on one or the other. The result is thedevelopment of a focus on either outcome or process,in which the focal element takes precedence over theother, with resulting implications for team adaptationand performance.

After observing the consistency of an outcome or pro-cess focus over the life cycle of a team, I tested a seriesof hypotheses regarding the effects of each on teamadaptation and performance. As predicted, outcome-focused teams exhibited higher levels of action iden-tification, a greater ability to identify problems, anda greater ability to adapt their work processes than

process-focused teams. The adaptability of outcome-focused teams appeared to allow them to do more withless, because outcome-focused teams losing materials atthe midpoint did not score significantly fewer points thanthe teams that had the benefit of all of their materials fortheir work. Furthermore, action identification mediatedthe effects of outcome focus on adaptation. Prior workon action identification suggests that higher-level actionidentities make changes easier for individuals who canconceive of multiple means for achieving the higher-level goal (Vallacher et al. 1989). That work is consistentwith the observations here, where a high level of actionidentification was associated with greater team process

adaptation.Contrary to predictions, process focus did not improve

a teams ability to deal with member change. Recentresearch by Lewis et al. (2007) suggests a possi-ble explanation for this unexpected result, demon-strating that membership change creates inefficienttransactive memory processes. While new members gen-erally adapt their specializations to maintain stabilityin the role structure of the team they join, the overallteam experiences a decreased ability to apply memberknowledge efficiently after experiencing a membership


12/16


change (Lewis et al. 2007). Thus, it is likely that themembership change that occurred in the current studyimpeded those teams ability to integrate what theyhad learned in the course of their work to improvetheir performance despite the clearly defined memberroles enjoyed by process-focused teams. In addition, the

lower level of action identification exhibited by process-focused teams impeded their ability to adapt in the faceof task change, putting them at a disadvantage when los-ing task materials as well.

Taken together, these findings add a new dimensionto what we continue to learn about the importanceof events at the beginning of a teams life (Ericksenand Dyer 2004, Ginnett 1993, Hackman and Wageman2005). The way a team conducts its initial interac-tion can establish important and lasting norms abouthow they will function as a team (Bettenhausen andMurnighan 1985, Ginnett 1993). The expansiveness ofthe decisions a team makes and the strategies they

devise will set them on a trajectory toward high perfor-mance or low performance (Ericksen and Dyer 2004).As teams move through their work and experience tran-sitions (Gersick 1988, Okhuysen 2001), whether theyare outcome-focused or process-focused will shape thenature of the changes they make in response. The currentstudy demonstrates that outcome-focused teams muchmore readily adapted their work process to changing taskdemands than process-focused teams. In related work,Woolley (2009) observed much greater resistance to amidpoint strategy intervention in field-based consultingteams when they were highly process-focused than whenthey were not. Thus, the work presented here expands

our understanding of the importance of early events byfurther fleshing out the implications of such activities fora teams level of outcome or process focus and on thediffering effects of those two foci.

A remaining question for future research is whetherteams will almost always be largely process-focused oroutcome-focused, or do some teams find a way to main-tain a healthy dual focus or (despite cognitive and nor-mative pressures) to shift more fluidly and effectivelybetween foci? Such teams, if they can be found and stud-ied, might be in the best position for peak performance.Based on this and related research, however, one canspeculate that truly dual-focused teams are unusual. For

example, Woolley (2009) surveyed student project teamsto assess outcome and process focus at the beginning ofa semester-long project and found that more than three-quarters demonstrated either a definitive outcome or pro-cess focus. In addition, similar to the findings herein,the majority of student project teams without an identi-fiable focus significantly underperformed the outcome-or process-focused teams, and interview data suggestedthey were unfocused rather than dual-focused. Thus,the history of distinctions maintained between meansand ends and the neuroscience evidence regarding the

cognitive separation of the two argues for a low likeli-hood that a cohesive team could move fluidly betweenthe two elements very early in their work or strad-dle both foci simultaneously. Indeed, focus on the twoelements was significantly negatively correlated in thecontent analysis of team-planning conversations in the

current study. Future research should focus on findingthose teams that manage to maintain a healthy, cohe-sive dual focus and understanding the conditions andcharacteristics that allow them to do so. In the mean-time, managers and researchers alike are well advised tounderstand the tendency of most teams to focus on oneor the other and its implications for team functioning indynamic environments.

Another remaining question for future research con-cerns the conditions encouraging an outcome or pro-cess focus. Although some differences were observed inthe levels of outcome and process focus that developednaturally in the control condition, we cannot pinpoint

exactly why some groups focused more on one ele-ment or the other in the absence of an early interventionprodding them to do so. Certainly, individual differencescould play a role. Future research might focus on exam-ining individuals propensities for high- or low-levelaction identification as a precursor to the developmentof an outcome or process focus, because these propen-sities have been demonstrated to be stable differencesacross individuals (Vallacher and Wegner 1989). Further-more, although the current study focused exclusively onteams working on open-ended tasks, teams working ona task that has a clearly defined, singular goal (savethe patients life or put out the fire) or where perfor-

mance is evaluated based on the quality of the process(e.g., legal proceedings, scientific research) may benefitfrom process focus in ways not explored here. Finally,factors in the context surrounding the task will certainlybe important. Issues such as how the project is situatedin the innovation cycle (Benner and Tushman 2003),how the team launch meeting is conducted (Ericksenand Dyer 2004), and other situational cues regardingthe right way to conduct work or the importance ofattaining certain outcomes will all influence the focus ateam adopts and the resulting implications for task per-formance. Future laboratory and field research can focuson determining the antecedent conditions that will influ-

ence these variables.Finally, there are clear implications of this research

for managers interested in structuring and standardiz-ing the processes of teams engaging in knowledge work.Managers attempting to compensate for team memberturnover by creating standard operating procedures forteams are put on notice; the present research suggeststhat creating process-focused teams does not compen-sate for team instability. Although popular team hand-books and existing research underscore the importanceof clear protocols and procedures for all teams, the


13/16


subtleties around when and how such tools are intro-duced are consequential. Managers are advised to payparticular attention to how they structure early teammeetings and the relative emphasis they place on workprocesses versus outcomes. Teams that begin by reaf-firming role and task assignments and project schedules

before discussing project objectives are setting them-selves up to be process-focused. In contrast, a team thatbegins its work by having an in-depth discussion of whatmembers want to accomplish stands a better chance ofreaping the benefits of an outcome focus, particularly ina dynamic environment. The distinction between a teamlaunch that leads a team down the path to process focusversus one that results in outcome focus is subtle butimportant to appreciate. A failure to do so can result ina team that seems to be functioning smoothly yieldinginexplicably poor results.

Acknowledgments

The author wishes to thank J. Richard Hackman, TeresaAmabile, Jeff Polzer, Amy Edmondson, and Hillary AngerElfenbein for their feedback and advice during every phaseof this research. Martine Haas, Connie Hadley, MichaelOLeary, and Ruth Wageman provided helpful commentson earlier versions of the manuscript. Invaluable researchassistance was provided by Ishani Aggarwal, Emily Atkins,Benjamin Bibler, Martin Gordon, Melissa Liebert, Sarah Link,Prerna Martin, Caitlin Poluska, Kimberly Pope, Devika Sarin,Jennifer Thompson, and Aaliyah Williams. The author wouldalso like to thank Andrea Hollingshead and two anonymousreviewers for their very helpful feedback and Cathy Senderlingfor editorial assistance. This research was supported by aNational Science Foundation Graduate Research Fellowship.

Appendix A. Task Instructions for Laboratory Teams

House

Square footage (each knob is equal to one square foot): $10 per square foot on the ground floor $20 per square foot on the second floor $50 per square foot on subsequent floorsQuality: assessed using the lift, flip, and drop test No value added for buildings that cannot be lifted without

any pieces falling off $1,000 bonus for buildings that can be lifted $1,500 bonus for buildings that can be lifted and flipped

over $2,000 bonus for buildings that can be lifted, flipped over,

and droppedAesthetics: the attractiveness of your building will be

assessed according to the following criteria: No value added for buildings that are a total eyesore $1,000 bonus for good use of color $1,500 bonus for good use of color and symmetry $2,000 bonus for a work of artAdditional Rules:

The walls separating the individual stories of the housemust be least two (2) blocks high.

Each story of the house must be completely enclosed bya floor and a ceiling for it to count toward the point total. Theceiling of one story can also be considered the floor of thenext story.

The walls around the outside of the house must be twoknobs thick.

There must be a cement foundation (a layer of white Lego

bricks) between the house and the ground, though it does notneed to be directly touching the ground. This should be thesame dimension as the first story of the house and can, but isnot required to, serve as the floor of the first story as well.

There must be a solid layer of blue blocks between the topof the house and the sky. This should be the same dimensionas the top story of the house and can, but is not required to,serve as the roof of the house.

Garage

Square footage (each knob is equal to one square foot):$20 per square foot

Quality: assessed using the lift, flip, and drop test No value added for buildings that cannot be lifted without

any pieces falling off $1,000 bonus for buildings that can be lifted $1,500 bonus for buildings that can be lifted and flipped

over $2,000 bonus for buildings that can be lifted, flipped over,

and droppedAesthetics: the attractiveness of your building will be

assessed according to the following criteria: No value added for buildings that are a total eyesore $1,000 bonus for good use of color $1,500 bonus for good use of color and symmetry $2,000 bonus for a work of artCar spaces: $2,000 for each individual parking space in the

garage

Parking spaces must be completely enclosed and at least20 square feet (e.g., exposed knobs) in size and have a dooropening that is at least 4 knobs wide and 2 blocks high toqualify for this bonus.

Additional Rules: The walls of the garage can be either one knob or two

knobs thick. There must be a cement foundation (a layer of white Lego

bricks) between the garage and the ground, though it does notneed to be directly touching the ground. This can, but is notrequired to, serve as the floor of the garage as well.

There must be a solid layer of blue blocks between the topof the garage and the sky. This should be the same dimensionas the top of the garage and can, but is not required to, serveas the roof of the garage.

The garage must be completely enclosed on the top, bot-tom, and sides, except for specified doors.

Pool

Square footage (each knob is equal to one square foot):$50 per square foot

Aesthetics: the attractiveness of your pool will be assessedaccording to the following criteria:

No value added for pools that are a total eyesore $1,000 bonus for good use of color $1,500 bonus for good use of color and symmetry $2,000 bonus for a work of art


14/16


Additional Rules:

There must be a diving board included. The floor of the pool must be blue. The walls around the edge of the pool should be only

one knob thick and three blocks high and should be one solidcolor (of the designers choice) all the way around.

Appendix B. Survey Items for Outcome and ProcessFocus ScalesInstructions: In reflecting on your discussions with your team,please indicate on a scale of 1 to 7 (from Very uncertain toVery certain) the extent to which you discussed and devel-oped clarity or certainty about each of these issues.

Process Focus

What each of the subtasks are that need to be completedHow the team should divide its time among the various

subtasksWhat you personally are responsible for doing on the projectWhen each of the subtasks will be completed

Outcome Focus

What constitutes a successful performance on this taskWhat criteria will be used for evaluating the final productThe relative importance of the different parts of the task to the

final scoreWhat the final output of your teams work will look like

Endnotes1Data collection was conducted in two time periods, with50 teams from all but the no midpoint change condition par-ticipating in the first data collection phase and 40 additionalteams in the no midpoint change and material loss con-ditions participating in the second data collection phase. The$300 prize was offered during each data collection phase tothe highest-scoring team from among all conditions run duringthat phase. Because it was unknown a priori which conditionwould offer the most advantages to the teams, the offering ofa single prize during each data collection phase was deemedto be a fair approach to this incentive.2A second measure of attention to processes and outcomeswas calculated based on analysis of transcripts of group con-versations at the beginning and midpoint. Individual state-ments were coded for their relevance to group processes oroutcomes by two independent raters. The proportion of state-ments related to each was significantly correlated with themore global rating of attention reported here, and thus only

the latter is reported.

ReferencesArrow, H., J. E. McGrath. 1993. Membership matters: How member

change and continuity affect small group structure, process, andperformance. Small Group Res. 24 334361.

Barrick, M. R., M. J. Neubert, M. K. Mount, G. L. Steward. 1998.Relating member ability and personality to work-team processesand team effectiveness. J. Appl. Psych. 83(3) 377391.

Benner, M. J., M. L. Tushman. 2003. Exploitation, exploration, andprocess management: The productivity dilemma revisited. Acad.

Management Rev. 28 238256.

Bettenhausen, K., J. K. Murnighan. 1985. The emergence of normsin competitive decision-making groups. Admin. Sci. Quart. 30350373.

Blajenkova, O., M. Kozhevnikov, M. A. Motes. 2006. Object-spatialimagery: A new self-report imagery questionnaire. Appl. Cogni-tive Psych. 20 239263.

Brandon, D. P., A. B. Hollingshead. 2004. Transactive memory sys-

tems in organizations: Matching tasks, expertise, and people.Org. Sci. 15 633644.

Bunderson, J. S., K. M. Sutcliffe. 2003. Management team learningorientation and business unit performance. J. Appl. Psych. 88(3)552560.

Burgoon, J. K., L. Dillman, L. A. Stem. 1993. Adaptation in dyadicinteraction: Defining and operationalizing patterns of reciprocityand compensation. Comm. Theory 3(4) 295316.

Burns, T. R., G. M. Stalker. 1994. The Management of Innovation.Oxford University Press, Oxford, UK.

Campbell, D. J. 1988. Task complexity: A review and analysis. Acad.Management Rev. 13(1) 4052.

Carley, K. 1992. Organizational learning and personnel turnover. Org.Sci. 3(1) 2046.

Choi, H.-S., L. Thompson. 2005. Old wine in a new bottle: Impactof membership change on group creativity. Org. Behav. HumanDecision Processes 98(2) 121132.

Clark, R. M. 2004. Intelligence Analysis: A Target-Centric Approach.CQ Press, Washington, D.C.

Diener, C. I., C. S. Dweck. 1978. An analysis of learned helpless-ness: Continuous changes in performance, strategy, and achieve-ment cognitions following failure. J. Personality Soc. Psych. 36451462.

Dollard, M. F., A. H. Winefield. 1994. Organizational response torecommendations based on a study of stress among correctionalofficers. Internat. J. Stress Management 1(1) 81101.

Dougherty, D. 1992. Interpretive barriers to successful product inno-vation in large firms. Organ. Sci. 3(2) 179202.

Earley, C. P., P. Wojnaroski, W. Prest. 1987. Task planning andenergy expended: Exploration of how goals influence perfor-mance. J. Appl. Psych. 72(12) 107114.

Ericksen, J., L. Dyer. 2004. Right from the start: Exploring the effectsof early team events on subsequent project team developmentand performance. Admin. Sci. Quart. 49(3) 438471.

Gersick, C. J. 1988. Time and transition in work teams: Toward a newmodel of group development. Acad. Management J. 31 941.

Gersick, C. J., J. R. Hackman. 1990. Habitual routines in task-performing groups. Organ. Behav. Human Decision Processes

47 6597.

Ginnett, R. C. 1993. Crews as groups: Their formation and theirleadership. E. Wiener, B. Kanki, R. Helmreich, eds. CockpitResource Management. Academic Press, New York, 7198.

Goodale, M. A. 2000. Perception and action in the human visualsystem. M. S. Gazzaniga, ed. The New Cognitive Neurosciences.MIT Press, Cambridge, MA, 365378.

Goodale, M. A., A. D. Milner. 1992. Separate visual pathways forperception and action. Trends Neuroscience 15 2025.

Hackman, J. R. 1987. The design of work teams. J. W. Lorsch, ed.Handbook of Organizational Behavior. Prentice Hall, Engle-wood Cliffs, NJ, 315342.


15/16


Hackman, J. R. 1992. Group influences on individuals in organiza-tions. M. D. Dunnette, L. H. Hough, eds. Handbook of Industrialand Organizational Psychology. Consulting Psychologists Press,Palo Alto, CA, 199267.

Hackman, J. R., R. Wageman. 2005. A theory of team coaching. Acad.Management Rev. 30 269287.

Hackman, J. R., K. R. Brousseau, J. A. Weiss. 1976. The interaction

of task design and group performance strategies in determininggroup effectiveness. Organ. Behav. Human Decision Processes16 350365.

Hollingshead, A. B. 2000. Perceptions of expertise and transac-tive memory in work relationships. Group Processes IntergroupRelations 3(3) 257267.

Janz, B. D., J. A. Colquitt, R. A. Noe. 1997. Knowledge workerteam effectiveness: The role of autonomy, interdependence, teamdevelopment, and contextual support variables. Personnel Psych.50(4) 877904.

Kerr, N. L., S. R. Tindale. 2004. Group performance and decisionmaking. Annual Rev. Psych. 55 623655.

Klimoski, R., S. Mohammed. 1994. Team mental model: Construct

or metaphor? J. Management 20(2) 403437.Kozhevnikov, M., S. M. Kosslyn, J. Shephard. 2005. Spatial versus

object visualizers: A new characterization of visual cognitivestyle. Memory Cognition 33 710726.

Kozlowski, S. W. J., S. M. Gully, E. R. Nason, E. M. Smith.1999. Developing adaptive teams: A theory of compilationand performance across levels and time. D. R. Ilgen, E. D.Pulakos, eds. The Changing Nature of Performance: Implica-tions for Staffing, Motivation, and Development. Jossey-Bass,San Francisco, 240292.

Latham, G. P., D. C. Winter, E. A. Locke. 1994. Cognitive and moti-vational effects of participation: A mediator study. J. Organ.Behav. 15(1) 4963.

LePine, J. A. 2005. Adaptation of teams in response to unforeseen

change: Effects of goal difficulty and team composition in termsof cognitive ability and goal orientation. J. Appl. Psych. 9011531167.

Levine, J. M., H. S. Choi. 2004. Impact of personnel turnover onteam peformance and cognition. E. Salas, S. M. Fiore, eds. TeamCognition: Understanding the Factors That Drive Process and

Performance. American Psychological Association, Washington,D.C., 153176.

Lewis, K., D. Lange, L. Gillis. 2005. Transactive memory systems,learning, and learning transfer. Organ. Sci. 16 581598.

Lewis, K., M. Belliveau, B. Herndon, J. Keller. 2007. Group cog-nition, membership change, and performance: Investigating thebenefits and detriments of collective knowledge. Organ. Behav.Human Decision Processes 103 159178.

Locke, E. A., C. C. Durham, J. M. L. Poon, E. Weldon. 1997. Goalsetting, planning, and performance on work tasks for individ-uals and groups. S. L. Friedman, E. K. Scholnick, eds. TheDevelopmental Psychology of Planning: Why, How, and When

Do We Plan? Lawrence Erlbaum Associates, Mahwah, NJ,239262.

McGrath, J. E. 1984. Groups: Interaction and Performance. Prentice-Hall, Englewood Cliffs, NJ.

Moreland, R. L., L. Argote. 2003. Transactive memory in dynamicorganizations. R. S. Peterson, E. A. Mannix, eds. Leadingand Managing People in the Dynamic Organization. Erlbaum,Mahwah, NJ, 135162.

Moreland, R. L., J. M. Levine. 1992. Problem identification bygroups. S. Worchel, W. Wood, J. A. Simpson, eds. Group Pro-cess and Productivity. Sage, Newbury Park, CA, 1747.

Moreland, R. L., L. Argote, R. Krishnan. 1996. Socially shared cogni-tion at work: Transactive memory and group performance. J. L.Nye, A. M. Brower, eds. Whats Social about Social Cognition?Research on Socially Shared Cognition in Small Groups. Sage

Publications, Thousand Oaks, CA, 5786.National Commission on Terrorist Attacks Upon the United States.

2004. The 9/11 Commission Report. W.W. Norton & Company,Inc., New York.

Newell, A., H. A. Simon. 1972. Human Problem Solving. Prentice-Hall, Englewood Cliffs, NJ.

Okhuysen, G. A. 2001. Structuring change: Familiarity and formalinterventions in problem-solving groups. Acad. Management J.40 794808.

Pulakos, E. D., S. Arad, M. A. Donovan, K. E. Plamondon. 2000.Adaptability in the workplace: Development of a taxonomy ofadaptive performance. J. Appl. Psych. 85(4) 612624.

Quinn, R. W. 2005. Flow in knowledge work: High performance expe-rience in the design of national security technology. Admin. Sci.

Quart. 50 610641.

Rao, R., L. Argote. 1995. Collective learning and forgetting: Theeffects of turnover and group structure. Annual Meeting of theAmerican Psychological Association, Chicago.

Rizzo, J. R., R. J. House, S. I. Lirtzman. 1970. Role conflict andambiguity in complex organizations. Admin. Sci. Quart. 15150163.

Rosenthal, R., R. L. Rosnow. 1991. Essentials of BehavioralResearch: Methods and Data Analysis. McGraw-Hill PublishingCompany, New York.

Sherif, M. 1936. The Psychology of Social Norms. Harper Collins,New York.

Shure, G. H., M. S. Rogers, E. M. Larsen, J. Tassone. 1962. Groupplanning and task effectiveness. Sociometry 25 263282.

Smith, E. E. 1957. The effects of clear and unclear role expecta-tions on group productivity and defensiveness. J. Abnormal Soc.Psych. 55 213217.

Sperry, L. 1998. Organizations that foster inappropriate aggression.Psych. Ann. 28(5) 279284.

Squire, L. R., B. J. Knowlton, G. Musen. 1993. The structure andorganization of memory. Annual Rev. Psych. 44 435495.

Staw, B. M., R. D. Boettger. 1990. Task revision: A neglected formof work performance. Acad. Management J. 33(3) 534559.

Steiner, I. 1972. Group Process and Productivity. Academic Press,New York.

Stevenson, H. H., J. C. Jarillo. 1991. A new entrepreneurial paradigm.A. Etzioni, P. R. Lawrence, eds. Socio-economics: Toward a

New Synthesis. M. E. Sharpe, Inc., Armonk, NY, 185210.

Tushman, M. L., C. A. OReilly III. 1996. Ambidextrous orga-nizations: Managing evolutionary and revolutionary change.California Management Rev. 38(4) 829.

Ullman, M. T. 2004. Contributions of memory circuits to language:The declarative/procedural model. Cognition 92 231270.

Ullman, M. T., E. I. Pierpont. 2005. Specific language impairmentis not specific to language: The procedural deficit hypothesis.Cortex 41 399433.

Ungerleider, L. G., M. Mishkin. 1982. Two cortical visual systems.D. J. Ingle, M. A. Goodale, R. J. W. Mansfield, eds. Analysis ofVisual Behavior. MIT Press, Cambridge, MA, 549586.


16/16


Vallacher, R. R., D. M. Wegner. 1987. What do people think theyredoing? Action identification and human behavior. Psych. Rev.94(1) 315.

Vallacher, R. R., D. M. Wegner. 1989. Levels of personal agency:Individual variation in action identification. J. Personality Soc.Psych. 57(4) 660671.

Vallacher, R. R., D. M. Wegner, M. P. Somoza. 1989. Thats easy for

you to say: Action identification and speech fluency. J. Person-ality Soc. Psych. 56(2) 199208.

Wegner, D. M. 1987. Transactive memory: A contemporary analysisof the group mind. B. Mullen, G. R. Goethals, eds. Theories ofGroup Behavior. Springer-Verlag, New York, 185208.

Weingart, L. R. 1992. Impact of group goals, task component com-plexity, effort, and planning on group performance. J. Appl.Psych. 77(5) 682693.

Weldon, E., L. R. Weingart. 1993. Group goals and group perfor-mance. British J. Soc. Psych. 32 307334.

Woolley, A. W. 2009. Putting first things first: Outcome andprocess focus in task performing teams. J. Organ. Behav. 30

427452.

Woolley, A. W., M. E. Gerbasi, C. F. Chabris, S. M. Kosslyn, J. R.Hackman. 2008. Bringing in the experts: How team composi-tion and work strategy jointly shape analytic effectiveness. SmallGroup Res. 39(3) 352371.

Means vs. Ends- Implications of Process and Outcome Focus for Team Adaptation and...

Documents

Transcript of Means vs. Ends- Implications of Process and Outcome Focus for Team Adaptation and...