Exploring the Hidden-Profile Paradigm

Team Member–Level (Composition) Inputs

Team composition or member-level inputs (Mathieu et al., 2008) refer to the individual team members’ attributes, such as competencies (i.e., knowledge, skills, and abilities), personality, attitudes, status, and demographic characteristics (Bell, 2007; Mathieu et al., 2008). Research in different domains of small group research has explored the composition inputs and showed that these factors influence various team processes and emergent states. In the hidden-profile literature, however, many of these factors have received none or very limited attention. Information distribution and task-related expertise are the two main individual-level factors that have been extensively studied. Besides these two factors, status and experience, bio-demographic and task-related diversity, network features and familiarity, and team member personality and attitude have received some attention and have been examined in a few studies. In this section, we will review these attributes and their impact on team information exchange and decision-making processes.

Team-Level Inputs

In this section, we review research on the effect of various team-level inputs on team performance in a hidden-profile task. We have organized the pertinent studies under the following categories: information properties, task characteristics, team leadership, team training, team size, time pressure, and communication technology.

Organization- and Context-Level Inputs

The IMO framework includes the perspective that teams are entities embedded in and influenced by organizational systems and their surrounding environments (Mathieu et al., 2008). Human resource systems, organizational climate, multiteam systems coordination, and culture of the host country are examples of such factors (Mathieu et al., 2008). In our search for studies that relate to organizational and contextual influences on team performance in a hidden-profile setting, we were able to identify only four relevant studies; these studies examine the impact of accountability and incentive systems.

Researchers have argued that accountability should reduce the discussion bias toward shared information and lead to a revealing of more unshared information. Contrary to their expectation, Stewart, Billings, and Stasser (1998) found that accountable groups did not mention and discuss more unshared information. In contrast, Scholten, van Knippenberg, Nijstad, and De Dreu (2007) found that teams that expected to be held accountable for their decision-making process were more concerned about the sufficiency of the available information, repeated unshared information more often, and were more likely to choose the correct decision alternative. Even though both of these studies examined the issue of accountability, they differ in one important detail, which is the focus of accountability; participants in Stewart and colleagues’ (1998) study were held accountable for their decision outcome while those in Scholten and colleagues’ (2007) study were accountable for the decision-making process (see Scholten et al., 2007, for a detailed discussion on the differences between process and outcome accountability and their impact on team decision making). Therefore, as discussed by Scholten and colleagues, while process accountability seems to enhance a team’s decision-making process by increasing systematic information processing, outcome accountability seems to have some detrimental effects on the decision-making process by increasing the “desire to please the audience” (Scholten et al., 2007, p. 541).

In addition to the abovementioned studies, research on team climate (Steinel, Utz, & Koning, 2010; Toma & Butera, 2009) also relates to organization-level inputs in that researchers have manipulated the incentive system (a context-level factor) to induce a cooperative/competitive climate. Considering that team climate is an emergent state, we decided to discuss these studies under the section on emergent states.

Overall, our review of inputs at the individual, team, and organizational levels of analysis suggests that while researchers continue to focus on the central concept of information distribution, inquiries have expanded in this area to include numerous aspects and permutations of this distribution. In addition, research on team-level inputs represents expanded knowledge concerning a wide variety of factors. In particular, and not surprisingly, given the centrality of information in the hidden-profile paradigm, researchers have worked collectively to explicate nuances in information properties and task characteristics and their effects on team outcomes. Comparing individual- and team-level inputs, it is clear that the team-level inputs represent a wider variety in the existing literature. Almost absent, however, are explications at the organizational or contextual levels framed as inputs, suggesting a possible bias for bottom-up versus top-down meso approaches in terms of inputs. In the next section, we review work focused on mediating relationships in the hidden-profile literature.

Emergent States

Emergent states are cognitive, motivational, and affective states (Marks et al., 2001) that emerge as team members interact (Marks et al., 2001); these states are influenced by inputs at different levels of analysis as well as team processes and other emergent states (Marks et al., 2001). Even though some emergent states such as trust, cohesion, and empowerment have been studied for decades, in recent years we have seen a proliferation of studies on various emergent states such as mood, affect, shared mental models, and transactive memory systems (TMS). In this section, we review research on three categories of emergent states that have received attention in the hidden-profile paradigm.

Team Processes

As mentioned previously, the IMO framework builds on Marks and colleagues’ (2001) conceptualization of team processes and categorizes team processes as transition, action, and interpersonal processes (Mathieu et al., 2008). Transition processes involve activities such as mission analysis formulation, planning, goal specification, and strategy formulation (Marks et al., 2001). The role of transition processes in solving hidden profiles has received very little attention, with the work of Larson and colleagues (1994), in which participants were instructed to spend the first 5 min of team discussion on planning, being the only study that we can identify as being pertinent to this topic.

A Note Regarding Temporal Factors

One of the distinctive characteristics of the IMO framework (Mathieu et al., 2008) is its time-sensitive conceptualization that emphasizes the importance of developmental models (Gersick, 1988; Tuckman, 1965) and episodic approaches (Marks et al., 2001) in our understanding of team functioning. In general, empirical research on temporal aspects of team functioning remains scant (Mathieu et al., 2008) and the hidden-profile paradigm is not an exception. However, a collection of studies by Larson and colleagues (Larson, 1997; Larson et al., 1996; Larson, Christensen, et al., 1998; Larson, Foster-Fishman, & Franz, 1998; Larson et al., 1994) offers compelling insights on how team information processing behaviors unfold over time. Larson and colleagues advanced the idea of dynamic information sampling model and challenged the assumption that the probability of mentioning shared and unshared information remains stable over the course of a team discussion. The model predicts that shared information, compared with unshared information, will be more likely to be discussed earlier in the discussion. In addition, Larson and colleagues argued that with the progress of team discussion, the probability of mentioning new (not yet discussed) unshared information increases while the likelihood of introducing new shared information decreases. The empirical evaluation of the model provided strong support for these predictions (Larson, 1997; Larson et al., 1996; Larson, Christensen, et al., 1998; Larson, Foster-Fishman, & Franz, 1998; Larson et al., 1994). Later, Stewart and colleagues (2004) speculated whether similar patterns of information sharing could be obtained in a recall task (instead of a decision-making task). Their findings were mixed and did not provide strong support for the dynamic information sampling model in recall tasks.

Taken together, the existing work on mediators in hidden-profile situations represents much more attention to emergent states—with a fairly evenly distributed focus on mood, climate, and cognition—as compared with the attention given to transition, action (other than information processing), or interpersonal processes. As noted by others (Cronin, Weingart, & Todorova, 2011), this dearth of work on process may be due to the difficult and laborious nature of studying dynamic temporal processes in groups and teams over time. Room certainly exists for future work that explicates and integrates the influence of any team dynamic, whether emergent or process, on hidden-profile outcomes. In the following section, we summarize future research directions and recommendations derived from our literature review.

Theoretical Issues

With respect to theoretical issues, one notable observation from the current review concerns the interdisciplinary nature of the hidden-profile literature. The studies in this review were drawn from various disciplines/fields including social psychology, industrial/organizational psychology, communication, management, and organizational behavior, and reflect a number of theoretical perspectives. We would suggest that this attention to hidden profiles from so many fields largely reflects two factors. First, the central mechanism, information sharing, is universal and is applicable to seemingly all groups of individuals involved in making decisions, regardless of context. Second, and relatedly, the influences on this mechanism are seemingly varied and ubiquitous, residing at various levels of analysis and under the purview of various fields of study. This attention from so many fields, and the intersection among those fields in studying hidden profiles, is notable, and certainly appropriate. This said, we also could imagine other disciplines contributing to the discourse on this topic. For example, fields such as economics and anthropology (see, for example, Gatewood, 1984) likely have unique theories and insights to lend, and we would encourage integration of them.

A major goal of this review is to provide some conclusions regarding the question, “For which big theoretical questions in the hidden-profile literature have we made notable gains in understanding and for which do significant theoretical gaps remain?” We turn first to perhaps the most critical theoretical question, which is, “Why do individuals fail to disclose unique information?” As described above, scholars have forwarded various explanations for the tendency of individuals to mention shared information and withhold unique information. On one hand, there has been significant progress made in documenting one proposed effect, namely, that individual members’ initial preferences strongly influence the groups’ final decisions (Gigone & Hastie, 1993, 1997). For instance, Faulmüller and colleagues (2012) provided evidence to support this effect and suggested that members’ initial preferences influence their information sharing due to a desire to be understood by other group members, which, in terms of our model in Figure 1, would involve member-level inputs. Greitemeyer and Schulz-Hardt (2003) have also shown the importance of initial preferences, but they documented a different mechanism in explaining it, that is, that preference-consistent information is processed more thoroughly than is preference-inconsistent information. This mechanism would involve a mediating effect (see Figure 1) rather than a member-level input such as the desire to be understood. Together, these studies provide support for the importance of members’ initial preferences in driving information sharing and suggest that at least two mechanisms may be responsible for that effect. Thus, going forward, we would call for research further explicating these mechanisms. This may include conducting studies to compare these (and other) mechanisms or developing models and tests integrating them as we do not see them as necessarily contradictory.

The other major set of explanations for the tendency to discuss shared versus unique information draws from social-psychological principles. According to these explanations, members discuss shared information because they want to be validated and liked, for example, versus looking foolish or ignorant by offering unique information (e.g., Wittenbaum et al., 1999). In 2004, Wittenbaum and colleagues recognized the dearth of direct tests of these mechanisms and called for more research focusing on the specific member-level input of the motivation to share or not share information. However, our review suggests that this call has gone largely unheeded. Unlike the case with the initial preferences explanation, direct tests of these alternative propositions remain few in number. For example, we could only locate two studies testing the idea of mutual enhancement (Henningsen & Henningsen, 2004; Wittenbaum et al., 1999) and three studies testing the notion of social validation (Boos et al., 2013; Mojzisch et al., 2008; Parks & Cowlin, 1996). Furthermore, we were unable to locate any studies directly comparing the major explanations.

More generally, instead of focusing directly on the cognitive and motivational factors that affect information sharing, the hidden-profile literature historically has centered on the impact of team-level inputs and mediating emergent states. The present review suggests that this trend has accelerated in recent years. We would suggest, though, that higher level inputs ultimately influence important intrapersonal processes. As Stasser and colleagues (1995) noted, “the probability of an individual mentioning an item, p(M), can be conceptually decomposed into three sequential events: encoding the item before discussion, recalling it during discussion, and selecting it for discussion once it is recalled” (p. 247). Various factors affect these processes, but these processes themselves require greater exploration. Studying intervening cognitive and motivational mechanisms can not only provide theoretical insight into group decision making, but, more practically, doing so can also illuminate the role and potential modification of those input factors that impinge on these intrapersonal processes. Thus, we offer some suggestions for such endeavors below.

One practical recommendation we would make is for more integration of research on psychological safety and trust into examinations of individuals’ cognitions regarding information sharing. Despite the oft-cited rationale that members are hesitant to share information due to fear of appearing ignorant or not being socially validated (Wittenbaum et al., 1999), studies explicitly incorporating and measuring states such as psychological safety or trust within the hidden-profile paradigm seem to be virtually absent. As one of our anonymous reviewers appropriately suggested, what would be especially valuable would be to address more nuanced questions with regard to safety and trust. Instead of simply assessing whether psychological safety or trust affect sharing of unique information, researchers should endeavor to understand “about what it is” that members are feeling unsafe or untrusting. Do they fear being excluded from the group, not feeling accepted by the group, the group making the wrong decision, and/or something else? Also, researchers should examine the notion that different types of safety or trust concerns may relate to different subsequent behaviors (e.g., members being reticent vs. trying to convince other members of their correctness).

Moreover, one could ask how these various considerations interact. As an illustration of the likely complex state of affairs, one could imagine an individual who views her group as quite competent and is therefore hesitant to share unique information. She may hesitate to share because, given the competence of the group, she perceives (a) that someone else already would have voiced this information and/or (b) others will think less of her for sharing something irrelevant or inaccurate. At the same time, though, she may reason that, given its competence, the group ultimately will arrive at the correct decision and also would appropriately consider (or discount) her unique information in doing so. In this case, she trusts the group in one sense (in terms of arriving at the correct decision) but not in other sense (how members may treat her). Obviously, one could make this example even more intricate and likely realistic. For instance, her trust probably varies across target group members and also over the course of the decision-making episode. Practically, this complexity suggests adopting different methodological approaches. In particular, we would call for researchers both to model these factors (e.g., through experimental manipulations or computational modeling) and also to adopt more qualitative approaches—for example, by interrupting group discussions and privately querying members about their cognitions and motivations. We elaborate on these ideas below.

In addition to incorporating states like trust and psychological safety—the integration of which seems fairly intuitive—we also would call for research borrowing from distinct other literatures. For instance, with respect to sharing information, part of the tendency to disclose unique material likely stems from motivational considerations such as fear of appearing incompetent by proffering irrelevant or wrong information (Littlepage, Perdue, & Fuller, 2012; Wittenbaum et al., 1999). Thus, we could imagine scholars incorporating social-psychological work on metaperceptions (i.e., people’s perceptions of how others [will] view them) and the conditions that foster more or less concern among individuals about how they are seen by others (see Kenny & DePaulo, 1993).

Incorporating metaperceptions, and the closely related social relations model (e.g., Kenny & DePaulo, 1993), could help address the following question: To what degree is a member’s apprehension to share unique information a function of (a) the individual unique member (i.e., some members are more apprehensive than others), (b) shared concerns about a specific other group member(s’) reaction (i.e., some members are more feared than others), and/or (c) members being differentially being concerned about specific other members’ reactions (e.g., one member being apprehensive of Mike’s reaction and another apprehensive of Jill’s reaction)? Taking this a step further, one could look at, for example, whether those concerns are a function of factors such as demographics (e.g., being the only African American member of an otherwise all-White group), or personality traits (e.g., some members are apprehensive of everyone else’s reactions due to their own personal traits while other members are apprehensive of only specific others’ reactions due to those others’ traits; see Frey & Tropp, 2006; King, Kaplan, & Zaccaro, 2008). Clearly, there are many possibilities here.

A second important question remaining to be explored in the hidden-profile literature concerns the relative importance of specific inputs. As reflected in our review, scholars have continued to investigate a number of these factors. Here, we would conclude that, despite uncovering and reviewing a sizable body of work on various inputs, we were, at times, unable to draw firm conclusions about their impact because studies have yielded inconsistent results. For instance, with respect to the question of whether ethnic diversity improved teams’ information sharing activities, Phillips and colleagues (2006) found it did, whereas Kooij-de Bode and colleagues (2008) found it did not. As another example of contradictory findings, Bramesfeld and Gasper (2008) found that happy moods led to a focus on unique information and to better team performance, whereas van Knippenberg and others (2010) showed that a sad mood improves team performance by facilitating information elaboration. Here, though, Emich (2014) attempted to reconcile inconsistencies regarding the effects of mood by postulating and documenting that the mix of (positive and negative) moods in the group affected information processing and exchange. In general, then, we would call for more such studies in which researchers propose and test ideas that can reconcile previous inconsistent findings. This approach implies going beyond the question that has been traditionally asked - whether Factor X impacts the disclosure of unique information and the group’s decision making quality - and instead investigating under what conditions, using what operationalizations, and at what levels these factors influence these outcomes.

Another area where we see great potential is examining the interplay and relative importance of the various input factors that can influence the sharing of unique information. As the current review demonstrates, studies generally focus on one factor to the isolation of other potential contributors. In reality, multiple factors are obviously at play, and we would call for future research to take steps to model this complexity. For example, studies addressing whether status or personality (e.g., assertiveness) reign supreme in determining whether individuals disclose unique information and/or addressing the interaction between these factors would seem theoretically and practically useful. Obviously, the number of permutations one could address is practically infinite, and we are not advocating that researchers examine sets of factors in an atheoretical, haphazard manner. Rather, theory and real-world evidence should guide selection of the particular variables under consideration.

Another observation gleaned from our review, and consistent with others’ conclusions (e.g., Cronin et al., 2011), is that studies of emergent team states predominate relative to those examining dynamic actions in the hidden-profile literature. In particular, we perceive a significant gap in terms of specifying the dynamic communication behaviors that mediate the linkage between inputs and outcomes in these decision-making teams (see Figure 1). Communication is the means through which cognitive factors ultimately affect team decision-making outcomes as these factors matter insofar as information does or does not get expressed. As such, studies examining communication elements such as the amount of communication, the nature of it (e.g., more or less conflictual), which group members communicate when and how often, and the frequency of turn-taking are all candidates that may relate to sharing of unique information. These processes and elements warrant study both as unique aspects of the group decision-making context and also as indicators, and constituents, of emergent team norms and team climate which, ultimately, can then influence such information exchange (see also Humphreys & Aime, 2014).

An additional and fundamental issue that remains unresolved is as much practical as theoretical in nature: that is, how can groups and organizations promote the sharing of unique information? In our opinion, despite the vast amount of research done on hidden profiles, group and organizational scholars are not yet in the position to provide firm conclusions/advice (e.g., to groups or organizations) about how to promote such sharing. For instance, although studies document that training can affect information sharing (e.g., Deiglmayr & Spada, 2011; Klocke, 2007; Larson, Christensen, et al., 1998; Larson et al., 1994; Mennecke, 1997; Stasser et al., 1989; Waddell et al., 2013), they do not generally show that it can improve ultimate decision-making performance.

We attribute this inability to provide more practical recommendations to several factors. First, as discussed above, we still remain fairly ignorant with respect to why group members do or do not choose to share such information and, in particular, about which factors matter most and at what point. Second, many of the input factors examined to date are not amenable to change. Advising organizations to alter many of these factors is either unfeasible (e.g., changing task characteristics, the personality makeup of the group, whether or not a time pressure exists) or impossible (e.g., altering the longevity of the group). Finally, even among those inputs and mediators that organizations could influence (e.g., communication technology, group mood), the literature often cannot provide consistent conclusions and recommendations.

We see a few ways to increase the practical importance of this research. First, replicating and extending research on the intrapersonal processes (as suggested above) will help identify inputs that affect those processes. Second, we would call for an emphasis on inputs over which groups and organizations can have control. For example, in addition to studying the impact of leader style, studies could examine the influence of more specific leader behaviors (e.g., giving praise or structuring the discussion to a greater or lesser degree). These behaviors are the more proximal inputs on members’ behavior. Finally, we would suggest several methodological innovations. In addition to enriching and testing theoretical accounts, adopting various methodological strategies (e.g., studying hidden profiles in more naturalistic contexts) could provide practical insights.

One emergent team phenomenon that would seem germane to hidden profiles is transactive memory system (TMS). Intuitively, when not all information held is perfectly shared, groups should benefit when members are aware of which other members possess or should possess (based on factors like job position) different types or pieces of information (i.e., when the group has a well-developed TMS). Here too though, the literature seems to be primarily mute. Although some studies have linked TMS to information sharing (e.g., Mohammed & Dumville, 2001; Randall, Resick, & DeChurch, 2011), research explicitly examining TMS in the hidden-profile literature is mostly missing (but see Stasser et al., 1995, for one exception). In another recent exception, Engelmann and Hesse (2011) studied three-member groups who were physically separate while working on problem-solving tasks. They found that members of groups who were provided with digital concept maps representing the meta-knowledge structure of their other members discussed unshared information earlier and generally performed better than those without these maps.

Methodological Issues

In addition to suggesting theoretical avenues to pursue, the current review also leads to some conclusions about methodology, both in terms of advances and with respect to gaps in the literature. With respect to advances, one observation from this review is that scholars have used a wide variety of innovative experimental paradigms to try to elucidate the phenomenon. Also, studies also have incorporated physiological measurements in the hidden-profile literature (e.g., Minas et al., 2014), a practice that may yield additional insights into the intrapersonal processes at play. At the same time, the current review also points to some opportunities for additional methodological approaches. Here, we provide some practical recommendations with respect to this issue.

One theme emerging from this review is that the majority of hidden-profile research consists of laboratory studies. Notably, Brodbeck, Kerschreiter, Mojzisch, and Schulz-Hardt made the same observation in 2007 and called for more applied research. The current review suggests that the situation has not appreciably changed since they made this recommendation. The primary reliance on laboratory experiments is understandable given that this context allows researchers to carefully configure studies to address specific questions, such as those involving who possesses what information and the influence of different team roles, structures, and the like. While we certainly are not advocating abandoning such studies, we also would call for increasing the study of group decision making in naturalistic contexts or variations of laboratory studies that better simulate natural settings (for an example, see Sohrab, 2014).

Observing real organizational groups as they make decisions can advance the hidden-profile literature and, more generally, the group decision-making literature, in several ways. First, it can reveal the frequency of unique information not being shared. Although some well-known events reveal that this phenomenon can occur, and can result in catastrophic consequences (e.g., Foushee & Helmreich, 1988), we currently do not know just how common this phenomenon is in practice. We are aware that people can fail to provide unique information, but we do not know how often it does happen relative to all the instances (i.e., group decisions) it could. Simple methodological practices like postevent debriefings (perhaps especially if done in private) can help reveal who knew what, when, and why they did or did not choose to share that information.

Perhaps more importantly, current research does not document just how the phenomenon of unique information remaining unshared actually unfolds in practice. When, exactly, do group members forget or decide not to share this information? What are the conditions immediately preceding this forgetting or decision to withhold such information? As the current review makes clear, research tends to focus on variables and factors that influence these processes, but it primarily does not pinpoint the immediately proximal environmental conditions (e.g., what was just said by whom) and resultant intrapsychic processes responsible for this effect. Knowing how often and precisely how, why, and when, these information sharing lapses occur could, in turn, guide laboratory studies that ultimately reveal theoretical insights into this phenomenon. It also could lead to the development of specific recommendations and practices that organizations and groups could implement to counter it.

Finally, studying information sharing in more natural contexts could allow for the design of laboratory studies or dynamic simulations that more closely mimic naturalistic group decision-making scenarios. Our suspicion is that real decision-making contexts often differ from those typically conducted in the hidden-profile literature in meaningful ways with respect to factors such as time, accessibility to information, stability of information, and the nature of the problem space. For instance, many important group decisions likely take place over the course of days, weeks, or months. These longer time frames may result in fundamental differences in how information is accessed, recalled, and shared. In terms of accessing information, in many hidden-profile studies, participants receive a folder with a large amount of information in the beginning of the experiment and then return the folder to the researchers before the team commences discussion. In actual decision-making scenarios, though, such as those that top management teams or operating room teams encounter, members must most likely access the information throughout the discussion. As such, in these latter scenarios, questions about encoding, retrieving, and sharing information likely take a somewhat different form than do those we typically address. For instance, we could envision studies examining the frequency with which individuals revisit their information sources containing unique versus shared information and how these actions influence decision making. Finally, in terms of the problem space, in the traditional hidden-profile paradigm, participants are presented with a specific problem to solve or decision to reach. In reality, though, part of the group decision process often entails detecting a problem in the first place and also framing the problem in the appropriate manner (e.g., identifying an underlying cause or problem of which the seeming problem is just an indicator).

With respect to this last point, many of the studies in this field replicate the settings of the original study by Stasser and Titus (1985, 1987) wherein participants receive a list of alternatives (e.g., job candidates) and are instructed to choose one of these predetermined alternatives. In addition, participants in these studies start the task by going through their individual packages without having an opportunity to discuss any information with their team members. This setting reinforces the formation of individual preferences that are shown time and again to play an important role in biased information sharing and discussion (Faulmüller et al., 2012). Although a large number of studies with the exact same design enriches our understanding of factors that contribute to information processing in that particular setting, it limits our knowledge of information processing in the absence of those conditions. Questioning the necessity of these design features in our understanding of how teams process asymmetrically distributed information, Sohrab (2014) studied teams who worked under conditions of asymmetric information distribution but were not given a list of alternatives and individual preferences were not formed before team discussion began. Sohrab found that even in the absence of such conditions, unshared information is mentioned and repeated less than shared information. Surprisingly, the current extensive body of literature offers a very limited explanation of why unshared information receives less attention even when participants are not biased with their initial preferences and our knowledge of information processing in the absence of a list of predetermined alternatives is scant. Hence, we believe that to gain a deeper understanding of how teams process asymmetrically distributed information, future research needs to question some of the common design elements and explore novel research designs.

In addition to conducting field research with naturally occurring groups, we also would call for studies examining hidden profiles in high-fidelity simulations, such as those that are (becoming) standard practice in industries such as aviation, power plant operations, and medicine (e.g., Salas, Rosen, Held, & Weissmuller, 2008; Stachowski, Kaplan, & Waller, 2009). These simulations are designed to mimic the real situations and complexities that the constituent teams may face (Waller, Lei, & Pratten, 2013). In addition to being able to create situations that may more closely reflect those that most decision groups face, simulations would also allow for real-time assessment of the spread and retention of information and of motivational considerations in sharing it or not. For example, by stopping simulations at certain points (Uitdewilligen, Waller, & Pitariu, 2013) and through postscenario debriefings (Salas et al., 2008), researchers could ask specific members what information they possess and why they did or did not choose to disclose a given piece of information. Working backwards, one could then attempt to identify the conditions that led to the failures in encoding, storing, and/or sharing that information.