`1995,27 (1),12-24
`
`Networked simulations: New paradigms
`for team performance research
`
`JEANNE L. WEAVER and CUNT A. BOWERS
`University ofCentral Florida, Orlando, Florida
`and
`EDUARDO SALAS and JANIS A. CANNON-BOWERS
`Naval Air Waifare Center Training Systems Division, Orlando, Florida
`
`The prevalence of the use of teams in a variety of occupations and environments has increased the
`importance of investigating the processes involved in their performance. However, in the past, there
`have been few methodologies available for the investigation of team performance. The present man(cid:173)
`uscript attempts to contribute to this area of research by describing the rationale underlying the use
`of computer-based simulations in research on team performance. This is followed by a review of the
`networked simulations that are currently being used in team-performance research. This review em(cid:173)
`phasizes the capabilities provided by the networks and the types of research concerns for which they
`are effective. Finally, the application of this technology to the broader study of group performance
`is discussed.
`
`Because teamwork is prevalent in a number of occu(cid:173)
`pations (e.g., fire-fighting, aircrews, and medicine), the
`ability of teams to work effectively has become a vitally
`important issue. In fact, Watson (1990) states that the ef(cid:173)
`fective use of teams is "America's best hope" for compe(cid:173)
`tition in the worldwide marketplace. It has been noted
`that technological developments and global competition
`have placed added emphasis upon understanding the
`processes and performance of teams because many tasks
`are often beyond the mental and physical resources of
`one individual (Salas, Dickinson, Converse, & Tannen(cid:173)
`baum, 1992). Cannon-Bowers, Oser, and Flanagan (1992)
`cite three reasons underlying the increased importance of
`using teams in industry. The first is that there are critical
`tasks that cannot be accomplished by one individual
`alone. The second is the belief that groups will together
`perform better than single individuals. Furthermore,
`certain critical tasks often benefit from the redundancy
`offered by the use ofteams (e.g., nuclear power plant op(cid:173)
`erators). The third is that group structures have devel(cid:173)
`oped in response to the humanistic movement in indus(cid:173)
`try; that is, it is argued that the use of groups and work
`teams increases the source of significance and respon(cid:173)
`sibility of individuals in relation to their occupations.
`Cannon-Bowers et al. (1992) concluded, in their re(cid:173)
`view of the literature on the use of work teams in indus(cid:173)
`try, that "work groups are important and offer enough
`
`The views expressed in this paper are those ofthe authors and do not
`reflect the official position of the Department of the Navy, the De(cid:173)
`partment of Defense, or the U.S. Government. Please address corre(cid:173)
`spondence to Clint A. Bowers, Team Performance Laboratory, Psy(cid:173)
`chology Department, University of Central Florida, Orlando, FL
`32816.
`
`potential to warrant creative, innovative theoretical and
`methodological approaches to the study of their design
`and effectiveness" (Cannon-Bowers et al., 1992, p. 370).
`Despite the critical role that teams play in industry, how(cid:173)
`ever, science has made woefully little progress in under(cid:173)
`standing the factors that contribute to effective team per(cid:173)
`formance. In fact, reviewers in this area have severely
`criticized the available knowledge regarding team per(cid:173)
`formance (i.e., Dyer, 1984; Modrick, 1986). In large
`part, the absence of a sufficient data base in team per(cid:173)
`formance can be directly attributed to the lack of an ap(cid:173)
`propriate methodology for the study of teams.
`Research on team process and performance imposes
`a unique challenge to researchers. Because a team has
`been defined as "a distinguishable set of two or more in(cid:173)
`dividuals who interact dynamically, interdependently
`and adaptively to achieve specified, shared and valued
`objectives" (Morgan, Glickman, Woodard, Blaiwes, &
`Salas, 1986, p. 3), a "team task" must provide a situation
`in which multiple operators are required to interact in an
`interdependent manner. Yet, there have historically been
`relatively few laboratory paradigms that can be used as
`effective teamwork testbeds. Thus, researchers have
`been remanded to relatively contrived tasks that have
`questionable external validity (i.e., tower building). How(cid:173)
`ever, the advent oflow-cost, configurable computer net(cid:173)
`works might provide a technology that allows for the de(cid:173)
`velopment of much more realistic laboratory analogs of
`team tasks. Research paradigms using these tools have
`begun to appear, but are limited almost exclusively to the
`group decision-making literature. However, it is likely
`that the networked simulation approach will be equally
`useful for the study of other types of teams and issues in
`group process and performance.
`
`Copyright 1995 Psychonomic Society, Inc.
`
`12
`
`Petitioner Riot Games, Inc. - Ex. 1035, p. 12
`
`
`
`NETWORKED SIMULATIONS FOR TEAM RESEARCH
`
`13
`
`The present manuscript attempts to contribute to this
`area of research by describing the rationale underlying
`the use of computer-based simulations in research on
`team performance as related to team-performance the(cid:173)
`ory and the networked simulations that are currently
`being used in team-performance research. This review
`emphasizes the capabilities provided by the networks
`and the types of research concerns for which they are ef(cid:173)
`fective. Finally, the application of this technology to the
`broader study of group performance is discussed.
`
`LOW-FIDELITY SIMULATION AS A
`TESTBED FOR TEAM PERFORMANCE
`
`Tasks used in previous investigations of team perfor(cid:173)
`mance range from artificial and contrived laboratory
`ones to complex and expensive high-fidelity simulations
`(Bowers, Salas, Prince, & Brannick, 1992). The former
`have been criticized for their artificiality and the latter
`for their lack of experimental control. Furthermore, such
`simplistic laboratory tasks as tower building fail to cap(cid:173)
`ture the essence ofteam performance in that there is lit(cid:173)
`tle need for interdependence and interaction among team
`members. Bowers et al. argue that an understanding of
`naturalistic team performance will be forthcoming only
`by investigating teamwork behaviors among interdepen(cid:173)
`dent operators performing different types of taskwork.
`Bowers and his colleagues further state that the in(cid:173)
`creased requirement for coordination will probably im(cid:173)
`prove the generalizability ofteam research to real-world
`environments.
`In general, team researchers have delineated areas in
`need of further research and have called for the devel(cid:173)
`opment of better methodologies with which to meet this
`need (Dyer, 1984). In fact, it has been argued that "the
`lack of empirical studies ofteam training is secondary to
`the absence of methodologies to capture the dynamic
`behaviors inherent in team activity, assess the nature and
`levels of complex team performance, or determine the
`relationships among the relevant set of variables" (Bow(cid:173)
`ers, Morgan, & Salas, 1989, p. 10). Thus, while team re(cid:173)
`searchers are aware of the areas that need research, they
`are likewise aware that effective research can result only
`when sound methodologies are discovered and made
`available. In large part, the lack of useful paradigms for
`team-performance research can be attributed to limita(cid:173)
`tions in technology.
`In the past, the study of coordinated behavior pro(cid:173)
`vided a formidable challenge for researchers because it
`was difficult to create the task or measure the resulting
`performance. However, researchers interested in investi(cid:173)
`gating team performance have begun to employ low(cid:173)
`fidelity networked simulations to gain an increased un(cid:173)
`derstanding of the various factors that might impact
`team performance, such as structure (Bowers, Urban, &
`Morgan, 1992; Kleinman & Serfaty, 1989), team train(cid:173)
`ing load (Morgan, Coates, Kirby, & Alluisi, 1984), and
`communication (Bowers, Kline, & Morgan, 1992). Low(cid:173)
`fidelity simulations can be likened to computer games
`
`that are then networked in order to provide a task usable
`by more than one individual. That is, a networked simu(cid:173)
`lation can provide a task suitable for use by a team of in(cid:173)
`dividuals. More importantly, a task of this type provides
`a useful, low-cost method which answers the need ofre(cid:173)
`searchers for an interdependent and interactive approach
`with which to investigate team processes and perfor(cid:173)
`mance. Although the pioneer use of low-fidelity simu(cid:173)
`lation was undertaken within the Ohio State studies (John(cid:173)
`ston & Briggs, 1968; Kidd, 1961; Naylor & Briggs, 1965),
`relatively few contemporary team/group researchers
`have adopted the methodology. Other researchers have
`also noted that the "rich, colorful, and challenging envi(cid:173)
`ronments offered by computer games provide powerful
`tools with which the foundations of a new approach
`might be studied and tested" (Hart & Battiste, 1992,
`p.1291).
`
`TEAM EFFECTIVENESS MODEL
`
`It has been noted that research is best directed in rela(cid:173)
`tion to a particular theoretical paradigm. That is, it has
`been argued in the past that there is "nothing more prac(cid:173)
`tical than a good theory" (Marrow, 1969). This section
`will describe one of the most recent and inclusive mod(cid:173)
`els of team performance which might serve as a useful
`guide to team research. This model is based on the team
`performance literature and describes a number of rele(cid:173)
`vant factors for investigations of team performance in a
`variety of domains. The purpose ofdescribing the model
`here is to present an inclusive conceptualization for the
`study of team performance in order to illustrate the vast
`number of factors that require investigation for the de(cid:173)
`velopment of a thorough understanding of team pro(cid:173)
`cesses and performance.
`The team effectiveness model (Salas et aI., 1992;
`Tannenbaum, Beard, & Salas, 1992) represents an inte(cid:173)
`gration of a number of models developed in an attempt
`to explain team (group) process and outcomes (see Salas
`et al., 1992, for a review of these models). Figure 1 de(cid:173)
`picts the model. The team effectiveness model (TEM)
`builds upon the classic input-throughput-output model.
`Team inputs are individual and team characteristics, task
`characteristics, and work structure; examples of these
`variables are task structure, team norms, attitudes, and
`team cohesion. Throughputs are the processes by which
`the team communicates, coordinates, and makes use of
`its resources to produce outputs over a period of time;
`the variables include problem solving, communication,
`and coordination. Outputs include the quantity and qual(cid:173)
`ity ofwork or products produced by the team and changes
`in the team and its members; the changes might be new
`norms, attitudes, and communication patterns. Tannen(cid:173)
`baum et al. (1992) argue that these model components
`must be considered within the context of the organiza(cid:173)
`tional and situational environment.
`The TEM provides a useful method for the conceptu(cid:173)
`alization of team processes and performance and guid(cid:173)
`ance for team research. Although research on the com-
`
`Petitioner Riot Games, Inc. - Ex. 1035, p. 13
`
`
`
`14
`
`WEAVER, BOWERS, SALAS, AND CANNON-BOWERS
`
`ORGANIZATIONAL CHARACTERISTICS
`
`REWARD SYSTEMS
`ENVIRONMENTAL UNCERTAINTY
`
`!INPUT
`
`SUPERVISORY CONTROL
`RESOURCES AVAILABLE
`
`tTHROUGHPUT
`
`OUTPUT
`
`TASK
`CHARACTERISTICS
`
`task complexity
`
`task organization
`
`task type
`
`WORK
`CHARACTERISTICS
`
`work structure
`f---+ team norms
`communication
`structure
`
`-
`
`teamwork
`
`INDIVIDUAL
`CHARACTERISTICS
`
`TEAM
`CHARACTERISTICS
`
`task KSAs
`
`motivation
`
`altitudes
`
`r-----t power distribution
`
`member homogeneity
`cohesiveness
`
`I--
`
`I -
`
`f - -
`
`TRAINING
`
`task analysis
`
`raining design
`
`leaming principles
`
`TEAM
`PROCESSES
`
`coordination
`
`...-. communication
`
`TEAM
`PERFORMANCE
`quality
`quantity
`time
`errors
`
`----.
`
`I
`I
`
`FEEDBACK
`
`Figure 1. The team effectiveness model (rEM). (From "Toward an Understanding of Team Performance and Training;' by E. Salas. T. D.
`Dickinson. S. A. Converse, and S. I. Tannenbaum, 1992, in R. W. Swezey and E. Salas (Eds.) Teams: Their training andperformance, 1992,
`pp. 3-30, New York: Ablex. Copyright 1992 by Ablex Publishing Corporation. Reprinted by permission.)
`
`ponents defined within the model has been conducted to
`some extent, there is a need for researchers to systemat(cid:173)
`ically test the components of the model in order to deter(cid:173)
`mine their relative importance. The section that follows
`will review research conducted utilizing low-fidelity net(cid:173)
`worked simulation technology and detailed explanations
`of the simulations with illustrations of their appear(cid:173)
`ances. Each ofthese sections will conclude with discus(cid:173)
`sion relating the variables studied to the TEM.
`
`NElWORKED SIMULATIONS
`IN TEAM PERFORMANCE RESEARCH
`
`Low-Fidelity Aviation Research Methodology
`One area in which low-fidelity simulation has been
`applied is in aviation research. The methodology de(cid:173)
`scribed by Bowers and his colleagues (1992) utilizes a
`commercially available simulation presented on a per(cid:173)
`sonal computer and two monitors (connected via a video
`splitter) which functions as a "poorman's" network. Fig(cid:173)
`ure 2 depicts this configuration. This approach allows
`for the creation of task interdependence between team
`members by permitting the task to be divided so that each
`
`team member has both individual and overlapping tasks
`to perform. The operator serving as pilot inputs by uti(cid:173)
`lizing the joystick, while the operator serving as copilot
`inputs by utilizing the keyboard. The "pilot" controls al(cid:173)
`titude and heading, while the "copilot" is responsible
`for weapon selection and aircraft stabilization.
`Bowers and his colleagues delineate several advan(cid:173)
`tages to using such low-fidelity simulations for the in(cid:173)
`vestigation oftearn performance. First, the methodology
`is available at a relatively low cost. Second, it possesses
`the characteristics needed for use in team research, such
`as 2 or more subjects and the requirement for coordina(cid:173)
`tion and task interdependency. Finally, low-fidelity sim(cid:173)
`ulation provides the requisite experimental control of
`independent variables. Although there is a need to fur(cid:173)
`ther investigate the utility of this methodology, the re(cid:173)
`sults ofpast aircrew psychology investigations have con(cid:173)
`verged to suggest its reliability and validity (Bowers
`et al., 1992). That is, past studies that have adopted this
`methodology have obtained similar results, thus dem(cid:173)
`onstrating its consistency for investigating behaviors
`related to aircrew coordination (e.g., communication,
`assertiveness).
`
`Petitioner Riot Games, Inc. - Ex. 1035, p. 14
`
`
`
`NETWORKED SIMULATIONS FOR TEAM RESEARCH
`
`15
`
`,------~
`r-
`I
`I
`I
`I
`I
`1
`1
`I
`I
`I
`I
`
`1\
`
`1'\
`-,
`-,
`-,
`
`Monitor
`2
`
`I GJ
`f1
`
`-,111 r- ______
`-- _~:.-II
`---- -
`1\:
`1\:
`-, I
`
`I
`
`I
`I
`
`I
`I
`
`I
`I
`I
`I
`I
`I
`I
`I
`
`fbi
`
`B.
`Monitor 10-
`1
`
`D.
`
`I
`
`A.
`
`-
`
`Monitor
`3
`
`'"
`
`-,
`
`~
`
`[j-----
`
`COPILOT STATION
`
`PILOT STATION
`
`~
`
`c.
`
`__ Video
`
`- - - - Audio
`
`A. Video splitter D. Keybo.rd
`B. Audio mixer
`E. Joystick
`F. He.dsets
`C. Camcorder
`
`Figure 2. Schematic illustration of the low-fidelity research methodology (Bowers, Salas,
`Prince, & Brannick, 1992).
`
`A number ofresearchers (Smith & Salas, 1991; Stout,
`Cannon-Bowers, Salas, & Morgan, 1990) have made ef(cid:173)
`fective use of this methodology in aviation psychology
`research. For example, Stout et al. (1990) made use of
`the low-fidelity simulation methodology for their inves(cid:173)
`tigation ofthe relationship between aircrew coordination
`behaviors and performance. That is, these researchers
`demonstrated the utility of the methodology for investi(cid:173)
`gating coordination behaviors and their impact upon the
`performance of aircrews. This research is particularly
`critical given past reviews which have described the im(cid:173)
`pact ofineffective aircrew performance (Cooper, White,
`& Lauber, 1979). Failure to communicate and coordi(cid:173)
`nate effectively has been shown to lead to disastrous
`consequences. Driskell and Salas (1992) argue that re(cid:173)
`search conducted within the laboratory provides a unique
`opportunity to derive general principles of team perfor(cid:173)
`mance that can be applied to real-world situations in
`order to maximize team performance within operational
`settings. Consequently, low-fidelity flight simulations
`might provide a tool with which to gain an understand(cid:173)
`ing of aircrew coordination in order to permit optimal
`performance in aviation settings.
`This discussion illustrates the need for investigation
`of the dynamic nature ofteam process and performance.
`The low-fidelity network paradigm provides this capa(cid:173)
`bility by requiring team members to share functions.
`That is, this methodology appears amenable to the in(cid:173)
`vestigation ofthroughput factors, particularly such team
`processes as coordination and communication as por(cid:173)
`trayed by the TEM. The methodology also lends itselfto
`the investigation of "individual characteristics" such as
`those described by the TEM (e.g., attitudes, assertive(cid:173)
`ness). However, one shortcoming of this methodology is
`the extent to which such input factors as "task charac(cid:173)
`teristics" and "work structure" can be altered. For exam-
`
`pIe, it might prove difficult to provide a level of work(cid:173)
`load high enough to test its relationship to output fac(cid:173)
`tors, such as performance and team and individual
`changes, without bringing the task to an end (e.g., flights
`crashing). Finally, simulations ofthis type typically limit
`the number of team members to two. Therefore, "team
`characteristics" such as team size might be less amen(cid:173)
`able to investigation by this method. This methodology
`appears to be most effective for the derivation ofgeneral
`principles ofteam performance and for aviation-related
`research.
`
`Team Performance Assessment Battery
`The Team Performance Assessment Battery (TPAB)
`was developed as a tool to investigate team decision
`making (Bowers, Urban, & Morgan, 1992). However,
`because TPAB is somewhat generic, it appears to have
`utility for investigations ofteam performance outside of
`the tactical environment. The TPAB was developed on
`the basis ofresearch from two other methodologies, syn(cid:173)
`thetic work (Alluisi, 1967, 1969; Morgan & Alluisi,
`1972) and resource management (Kleinman & Serfaty,
`1989). The history ofthe synthetic work methodology is
`grounded in the work of Alluisi and his colleagues on
`the Multiple Task Performance Battery (MTPB; Alluisi,
`1967). The synthetic work methodology has a number
`of advantages (Alluisi, 1969). The primary ones are
`(I) relatively low cost, (2) the ability to measure many
`variables concurrently over extended periods of time,
`(3) capability for individual and team performance mea(cid:173)
`surement, (4) high face validity, (5) high degree of ex(cid:173)
`perimental control, and (6) simplicity of measurement.
`The purpose of synthetic work is to present multiple
`tasks to operators in a manner that requires time-sharing
`and results in realistic workload levels (Alluisi, 1969).
`For example, TPAB utilizes three watchkeeping tasks-
`
`Petitioner Riot Games, Inc. - Ex. 1035, p. 15
`
`
`
`16
`
`WEAVER, BOWERS, SALAS, AND CANNON-BOWERS
`
`warning-lights monitoring, blinking-lights monitoring,
`and probability monitoring-to provide the constant
`monitoring loads that are associated with many team
`tasks (Bowers et aI., 1992). The monitoring of both
`warning lights and blinking lights requires operators to
`respond to "critical conditions," or, in other words, de(cid:173)
`viations from their normal states. Reaction times of op(cid:173)
`erators to correct critical conditions are recorded by the
`simulation. Response times are also recorded for the
`probability monitoring task. This task requires operators
`to detect the presence ofa bias of pointer settings along
`two linear scales. Operator responses to "critical condi(cid:173)
`tions" for all three tasks are made via mouse interface.
`In addition to the presentation ofthe monitoring tasks
`borrowed from the synthetic work methodology, the
`TPAB presents a resource management task that is a
`modification of the distributed resource allocation and
`management (DREAM) task developed by Kleinman
`and his colleagues (Kleinman & Serfaty, 1989). Opera(cid:173)
`tors are required to utilize information from their com(cid:173)
`puter displays in order to coordinate resources and ac(cid:173)
`tions to prosecute incoming targets (Bowers et aI., 1992).
`Figure 3 depicts the display viewed by TPAB operators.
`
`The simulated radar scope displays incoming targets
`that must be prosecuted. Team members are required to
`coordinate in order to allocate two types ofrenewable re(cid:173)
`sources. Target and resource information are presented
`in a table containing current time, expected target pene(cid:173)
`tration time, target identification number and type, tar(cid:173)
`get status, score, and resources to be returned. The team
`is required to coordinate the allocation of their resources
`in order to prosecute as many targets as possible.
`The resource allocation task is presented simultane(cid:173)
`ously with the individual monitoring tasks. This ap(cid:173)
`proach is consistent with the synthetic work methodol(cid:173)
`ogy, and it has been noted that this approach enhances
`generalizability because operators are required to time(cid:173)
`share individual and team tasks (Alluisi, 1969). Further(cid:173)
`more, it has been noted that the created synthetic job
`places reasonable cognitive demands on operators while
`simultaneously providing effective performance mea(cid:173)
`sures (Bowers et aI., 1989).
`Bowers et al. (1992) provided an extension of the
`work of Kleinman and his colleagues (Kleinman & Ser(cid:173)
`faty, 1989; Kleinman, Serfaty, & Luh, 1984; Kohn,
`Kleinman, & Serfaty, 1987). That is, Bowers and his col-
`
`## /Type Alt Hdg Spd RR Stat TR TCPL Score
`2:20
`0
`34
`2
`0
`52
`0
`2:10
`1
`3
`1:05
`0
`16
`1
`75
`0
`1:50
`2
`
`2
`3
`
`o•
`
`I CLOCK I
`
`00:45
`
`ENEMYSTR I
`
`KILLED
`
`GEJ
`G:liJ
`
`ABCDE
`
`123458
`
`CANCEL
`
`•o I TRANS
`
`Figure 3. Team Performance Assessment Battery (TPAB) operator screen display (Bowers, Urban, & Morgan, 1992).
`
`Petitioner Riot Games, Inc. - Ex. 1035, p. 16
`
`
`
`NETWORKED SIMULATIONS FOR TEAM RESEARCH
`
`17
`
`leagues utilized the TPAB to investigate the relationship
`between team structure, or overlap of responsibility, and
`workload. However, this research goes beyond research
`previously conducted in that the synthetic work tasks
`add an additional and important element-that is, the
`degree that structure and workload relationships are
`impacted by the addition of individual tasks. In addi(cid:173)
`tion, Bowers et al. (1992) used 5- (rather than 2-) person
`teams.
`Results indicated that teams with no overlap of re(cid:173)
`sponsibility outperformed teams with partial overlap re(cid:173)
`gardless of workload conditions. These results contra(cid:173)
`dict earlier findings (Kleinman & Serfaty, 1989) that
`partial overlap of responsibility results in better perfor(cid:173)
`mance under high workload conditions than does total or
`no overlap. Kleinman and Serfaty also found that total
`overlap was associated with best performance under low
`and moderate workload conditions. Bowers and his col(cid:173)
`leagues (1992) hypothesized that tradeoff effects on the
`individual tasks might account for this contrary finding,
`because teams with overlapping responsibility outper(cid:173)
`formed no-overlap teams on the synthetic work or mon(cid:173)
`itoring tasks. Thus, it appeared that the teams with over(cid:173)
`lapping responsibility on the team task focused more on
`their individual tasks.
`These findings are indicative of the utility of TPAB
`for investigating team performance since most occupa(cid:173)
`tions require team members to perform individual tasks
`as well. Thus, TPAB provides a unique opportunity to in(cid:173)
`vestigate these circumstances. In terms of the TEM, the
`synthetic work methodology of TPAB provides an op(cid:173)
`portunity for the investigation of "task characteristics"
`such as workload and time pressure. The TPAB is also
`amenable to the investigation of "organization and situ(cid:173)
`ational characteristics" such as uncertainty. That is,
`TPAB provides a methodology for the investigation of
`work structure and task characteristics in a context that
`also necessitates performance of individual tasks. This
`should facilitate the generalizability of this task for the
`study ofteam processes and performance. Consequently,
`the TPAB should be useful for the study of groups in a
`variety of environments. Because TPAB simultaneously
`presents both resource allocation and synthetic work
`tasks, it is potentially very useful for study of the com(cid:173)
`ponents discussed within Tannenbaum's TEM.
`
`Tactical Naval Decision Making System
`The Tactical Naval Decision Making System (TAN(cid:173)
`DEM) has just been recently developed and provides a
`low-fidelity simulation of a command, control, and
`communication environment similar to that ofthe TPAB
`(Dwyer, Hall, Volpe, Cannon-Bowers, & Salas, 1992;
`Weaver, Morgan, Hall, & Compton, 1993). However, the
`TANDEM system was developed to provide a closer ap(cid:173)
`proximation of an actual combat information center for
`the investigation of team decision making than does the
`TPAB.
`
`The TANDEM task requires team members to query
`and share information in order to arrive at a decision
`about a target's identities and intentions. The TANDEM
`is a highly configurable PC-based simulation that was
`developed to allow for the investigation of factors such
`as task interdependence, time pressure, task load, and
`ambiguity, using from 1 to 3 operators. Operators per(cid:173)
`forming the TANDEM task are required to make deci(cid:173)
`sions regarding unknown contacts by consulting and
`integrating pieces ofinformation regarding contact char(cid:173)
`acteristics. That is, operators are required to make deci(cid:173)
`sions regarding the type, threat, and intent of contacts on
`the basis ofa total of 15 information pieces which some(cid:173)
`times can be ambiguous or conflicting. Each of the de(cid:173)
`terminations (i.e., type, threat, and intent) is made by
`integrating five information pieces. For example, a tar(cid:173)
`get can be a submarine, surface, or air type of craft. Tar(cid:173)
`gets also are civilian or military, having either peaceful
`or hostile intentions toward the team's own ship. On the
`basis ofdetermination ofintent, targets are either cleared
`or shot. Thus, the TANDEM requires accurate decisions
`to be made regarding the true characteristics of incom(cid:173)
`ing targets. Figure 4 illustrates the display viewed by
`TANDEM operators.
`The TANDEM system is particularly effective be(cid:173)
`cause of the degree of flexibility it offers. That is, TAN(cid:173)
`DEM scenarios are created with number of targets, type
`oftargets, information ambiguity, information organiza(cid:173)
`tion, and magnitude of penalties determined by the ex(cid:173)
`perimenter through the use of the TANDEM authoring
`system. This authoring system allows a great deal of
`flexibility in investigating team decision making.
`The TANDEM system is particularly well suited for
`the investigation ofconditions that characterize stressful
`environments. For example, it has apparent utility for the
`investigation of ambiguity and time pressure (Weaver
`et aI., 1993) which are "task characteristics" within the
`TEM of Tannenbaum and his colleagues. In addition,
`this task would also be an effective tool for the investi(cid:173)
`gation ofother task-related factors such as the weighting
`of information and redundancy. Although these four fac(cid:173)
`tors are task characteristics according to Tannenbaum's
`model, it should be noted that this decision-making task
`has utility for investigating these characteristics in re(cid:173)
`lation to "team processes" and "team performance."
`According to the model, such "team process" variables
`might be coordination, communication, and problem
`solving, while "team performance" variables could
`include quality, time, and errors. Additionally, the con(cid:173)
`figurability of the TANDEM system permits the inves(cid:173)
`tigation of other input factors which might impact team
`performance such as "work structure." Although
`TANDEM would allow for the investigation of numer(cid:173)
`ous factors of interest to team researchers, the task is
`only moderately dynamic in that the items of informa(cid:173)
`tion to be integrated remain constant throughout as pro(cid:173)
`grammed within the authoring system. That is, team
`
`Petitioner Riot Games, Inc. - Ex. 1035, p. 17
`
`
`
`18
`
`WEAVER, BOWERS, SALAS, AND CANNON-BOWERS
`
`fm.: 00:14:51
`
`p
`
`c
`
`000
`
`cIiuI:
`
`Figure 4. TANDEM operator display (Weaver, Morgan, HaIl, & Compton, 1993).
`
`members possess the same information throughout a
`scenario. Thus, the largest potential shortcoming of
`TANDEM is its failure to require the integration of dy(cid:173)
`namic information over time.
`
`Team Interactive Decision Exercise for Teams
`Incorporating Distributed Expertise
`Another system developed for the investigation of
`team decision making is the Team Interactive Decision
`Exercise for Teams Incorporating Distributed Expertise
`(TIDE2; Hollenbeck, Sego, lIgen, & Major, 1991). This
`low-fidelity networked simulation was developed at
`Michigan State University for use in the study of team
`decision making and includes the task itself and sub(cid:173)
`programs for the collection, sorting, and analysis of
`data. In particular, TJDE2 was developed to provide a
`methodology for investigating team decision making in
`environments characterized by complexity, uncertainty,
`and ambiguity.
`Although TIDE2 is a networked program developed
`for investigating team decision making in command and
`control, its scenarios can be changed for other uses as
`well. For example, the authors note that "one can move
`from a naval command and control scenario to an in(cid:173)
`vestment banking scenario, or a scenario involving a
`personnel selection decision" (Hollenbeck et al., 1991,
`p. 9). The command and control scenario requires 4
`team members to query nine attributes in order to deter(cid:173)
`mine the intent of incoming targets. These attributes
`must be considered in addition to five rules which de(cid:173)
`scribe how the attributes combine to indicate the level of
`
`threat. Each team member has a different area of exper(cid:173)
`tise manipulated through (1) the ability to measure target
`attributes, (2) knowledge of rules, and (3) the capability
`to translate raw target-attribute data into judgments re(cid:173)
`garding the target's threat level per attribute. Figure 5
`depicts the display viewed by operators ofthe task. Each
`team member has individual responsibilities hierarchi(cid:173)
`cally related to the team task, and there is a designated
`leader responsible for determination of the team's deci(cid:173)
`sion and the decision can be rendered at any time with or
`without the recommendations of other team members.
`The leader is also free to disregard the input of subordi(cid:173)
`nates. Team members communicate with one another by
`sending information via the simulation.
`Because decision making is so often conducted within
`contexts in which team members are hierarchically or(cid:173)
`ganized (e.g., organizations, fire-fighting teams), it is
`imperative that researchers gain an understanding of its
`processes in order to optimize decision-making perfor(cid:173)
`mance. Because TIDE2 was designed to be a flexible sys(cid:173)
`tem easily adaptable to many different types ofdecision(cid:173)
`making contexts, it is particularly amenable to these types
`of investigations. For example, it is possible to manipu(cid:173)
`late group conflict and "task characteristics" such as am(cid:173)
`biguity, time pressure, and task complexity.
`Because the task is useful in different circumstances,
`the physical appearance of the task is somewhat less dy(cid:173)
`namic than that ofother simulations discussed in this re(cid:173)
`port. However, this element does not necessarily jeopar(cid:173)
`dize the task's utility and might actually contribute to its
`effectiveness as a research tool for the investigation of
`
`Petitioner Riot