Choice Article

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 24

Group Decis Negot (2013) 22:873896 DOI 10.

1007/s10726-012-9333-3

Individual Priming in Virtual Team Decision-Making


Valerie L. Bartelt Alan R. Dennis Lingyao Yuan Jordan B. Barlow

Published online: 4 January 2013 Springer Science+Business Media Dordrecht 2012

Abstract Virtual teams have different interactions than face-to-face teams because they rely on information and communication technologies, which can impede or assist certain human cognitive processes. Past research has shown that although virtual teams exchange more information than face-to-face teams, poor decisions often result, because team members do not consider the unique information they receive from others. Drawing from cognitive psychology, our research explored a unique way to improve team decision-making through the use of cognitive priming. We proposed that priming group members to pay attention to others or to engage in counterfactual thinking would improve team members cognition and, therefore, team performance. Prior research with individuals and brainstorming teams has shown these forms of priming to improve performance; however, no research has attempted to use priming to improve the outcomes of virtual team decision-making, which requires deeper interaction and cognitive involvement than brainstorming. We performed two lab experiments using primes that have been found to improve the individual decision-making process. We found that priming had some impact, but it did not signicantly improve decision quality. Various reasons are discussed to explain why priming techniques may not be as powerful in teams as in individuals, and future research ideas are suggested to build on our initial work on priming in virtual team decision-making.

V. L. Bartelt Department of Information Systems, College of Business Administration, Texas A&M University-Kingsville, MSC 182, Texas A&M University-Kingsville, Kingsville, TX 78363-8202, USA A. R. Dennis L. Yuan J. B. Barlow (B ) Operations and Decision Technologies Department, Kelley School of Business, Indiana University, BUS 670, 1309 E. Tenth Street, Bloomington, IN 47405, USA e-mail: jbbarlow@indiana.edu

123

874

V. L. Bartelt et al.

Keywords Priming Cognition Virtual teams Decision-making Decision quality Team performance

1 Introduction Virtual teams are groups of geographically dispersed individuals who communicate through information and communication technologies (ICT) (Bell and Kozlowski 2002). Such teams are now prevalent in organizations (Chudoba et al. 2005), and the effectiveness of virtual team interactions is a major research stream in IS and collaboration literature (Kahai et al. 2007). Prior research suggests that ICT use can reduce workloads by 50 percent or more (Reinig et al. 1995-1996). The use of ICT can affect various virtual team processes (Schmidt et al. 2001), including the decision-making process. In particular, decision quality often suffers in virtual teams (Dennis et al. 19971998; Martz and Shepherd 2004). Virtual teams often face the challenge of bringing together multiple perspectives from differing team members, yet virtual teams tend to have poor information exchange (Dennis 1996). Prior research has identied possible culprits affecting the information exchange process in virtual teams that may result in poor decisions. One main possibility is that team members do not fully consider the information they receive from others (Dennis 1996; Heninger et al. 2006; Hilmer and Dennis 2001). Thus a primary reason for poor team performance may not lie in the group interaction processes of social psychology, but rather in the individual thought processes of cognitive psychology (Kolfschoten 2011; Nagasundaram and Dennis 1993). Prior research shows that many types of individual cognitive challenges are common to group work (OLeary 2011) and to virtual teams in particular (Powell 1994; Whitworth et al. 2000). There is also evidence to show that changes in individual cognition can change group level outcomes (e.g., Curseu and Schruijer 2008). Thus, we believe that improving individual-level cognition can improve group-level performance. There are many approaches researchers can take to better understand the collaboration process and the effects of individual cognition in that process. We adopt the concept of priming from cognitive psychology. Priming is a phenomenon where exposure to a stimulus causes concepts related to that stimulus to be activated in an individuals working memory, which in turn subconsciously inuences subsequent behavior. For example, if a person is primed with words associated with the concept of old age, he or she walks slower than those not primed (Bargh and Chartrand 2000) because when the concept of old age is introduced into an individuals working memory, he or she subconsciously acts older and yet is completely unaware of the change in behavior (Bargh and Chartrand 2000). Priming techniques have been extensively studied in psychology research and have successfully been used to increase individuals decision-making performance (Bargh and Chartrand 2000; Bargh et al. 2001; Koriat and Feuerstein 1976; Reb and Connolly 2009). In information systems research, priming has also been found to improve group electronic brainstorming (Dennis et al. 2012). Despite the research on using priming

123

Virtual Team Decision-Making

875

to change individual behavior, priming has not been studied in group-level decisionmaking tasks that require group collaboration and deep interaction among individuals. We propose that priming at the individual level will improve group decision making by improving individual cognition during group processes. A major barrier to higher decision quality in virtual team decision making lies in individual cognitiona failure to fully consider the information shared by others (Heninger et al. 2006). Specically, we propose that priming can be used to increase participants desire to share their information and consider others information during virtual team discussions, thus improving decision quality.

2 Theory and Prior Research 2.1 Virtual Teams and Decision-Making Collaborative work creates unique cognitive challenges beyond those of individual tasks. Many research streams have investigated the creation of systems, processes, and training that allow groups to reduce excessive cognitive loads during the collaborative process (e.g., Kolfschoten et al. 2011; Kolfschoten and Lee 2010). Increasingly, employees are working in virtual teams, which face additional challenges (Bell and Kozlowski 2002; Chudoba et al. 2005). Virtual teams often span geographical, temporal, and organizational boundaries (Bell and Kozlowski 2002; Chudoba et al. 2005). These barriers add additional problems and cognitive challenges to those faced by traditional teams. Even though ICT are helpful in virtual team formation, functioning and ultimate performance, prior research has found that these lean tools reduce the amount of emotional content shared (Reinig et al. 1995-1996) which is also vital for healthy team interactions (McGrath 1991). Prior research also shows that group tasks requiring convergence are less likely to be successful virtually than face-to-face (Kerr and Murthy 2004). In an effort to improve virtual team interactions, researchers have focused on various electronic collaboration outcomes including attitudinal outcomes (Cohen and Bailey 1997) such as satisfaction and condence (e.g., Reinig 2003; Reinig and Shin 2002; Cred and Sniezek 2003; Reinig et al. 1995-1996). Many different process techniques derived from social psychology have been developed and tested in groups (Briggs et al. 2004). In this paper, we take a different approach by focusing on individual cognitive processes based on research from cognitive psychology. Team decision-making presents unique cognitive challenges due to the need to negotiate multiple team members opinions (Dean and Sharfman 1996). Team members must process information concerning not only the task at hand, but also interpersonal dynamics and team consensus (Whitworth et al. 2000). Teamwork is a complex process, in which the ideas and comments from many different participants needs to be considered. The need to both contribute information and consider the information from other team members increases cognitive load, often to a point where team members have insufcient cognitive resources to do both successfully (Heninger et al. 2006). The cognitive challenges of collaborative teams often lead to poor decision quality because team members do not pay attention to information from others on their team

123

876

V. L. Bartelt et al.

(Dennis 1996; Dennis et al. 19971998; Heninger et al. 2006; Hilmer and Dennis 2001). Because of the cognitive demands of virtual teamwork, team members often cut cognitive corners. Prior research indicates that many individuals experience conrmation bias during teamwork, meaning they seek information to conrm their initial impressions (Webster et al. 1996) rather than being open to new ideas from other team members. Conrmation bias is a phenomenon where people interpret new information based on their current beliefs and ignore contradictory information (Ask and Granhag 2005). This cognition bias occurs as the result of cognitive dissonance (Festinger 1957; Jonas et al. 2001), which is built on the premise that people are not comfortable with two opposing thoughts existing at the same time. Decision quality tends to decrease if people justify their a priori choices and do not fully consider other information options raised during the team discussion (Dean and Sharfman 1993). In general, people often seek easily available information instead of digging deeper to nd contrary evidence (Ask and Granhag 2005). This cognitive need for closure often results in them locking into their initial decisions, strengthening the cognitive bias toward their own impressions and stereotypes. Under the inuence of these cognition biases, people tend to make individual decisions based on previous information, often anchoring to their initial conclusions that were made before team discussions (Ask and Granhag 2005). Previous research has examined the role of individual cognition in face-to-face groups (e.g., Kersten and Cray 1996; OLeary 2011) and virtual teams (e.g., Antunes and Ferreira 2011; Espinosa et al. 2011; Stahl 2006; Powell 1994; Whitworth et al. 2000), but few studies have attempted to inuence individual cognition in order to improve virtual team collaboration and performance. Curseu and Schruijer (2008) used framing as a method for changing cognition in face-to-face teams completing negotiation tasks. In their study, teams who read instructions framing the information as dangerous used more defensive strategies during negotiations, and ultimately had lower performance than neutral teams. We propose that priming, a similar concept, may be used to inuence individual cognition in order to improve group level outcomes and improve team performance, particularly in virtual teams.

2.2 Priming Priming is a cognitive phenomenon where exposure to a stimulus causes a behavioral response to a subsequent stimulus. Priming activates implicit memory, which involves subconscious and unintentional cognition (Graf and Schacter 1985; Schacter 1987). Prior research shows that much of human behavior uses subconscious cognition (Bargh and Chartrand 2000), and priming can be used to manipulate subconscious cognition. Usually, priming is implemented in research by presenting stimuli to a participant in order to activate internal representations of related concepts or attitudes. The activation of these concepts then changes subsequent behavior (Bargh and Chartrand 2000). Primes can be delivered through two different levels of awarenesssubliminal or supraliminal. With subliminal priming, individuals are not aware of any priming. For instance, a word may ash in front of them without their knowledge of it. In

123

Virtual Team Decision-Making

877

supraliminal priming, individuals are provided with a conscious task, but they are not aware of the primes effects (Bargh and Chartrand 2000; DeCoster and Claypool 2004). Prior research shows that supraliminal priming can inuence an individuals attitudes and behavior in subsequent tasks (Chalfoun and Frasson 2011). Researchers have developed many sophisticated supraliminal priming techniques, including the use of scenarios (Connolly and Reb 2003) and semantic (word) priming (Bargh and Chartrand 2000; Higgins and Chaires 1980). Semantic priming works by activating semantic networks in an individuals cognition. Semantic networks are networks of words and concepts built in the brain through prior experience (Martin and Chao 2001). Research has shown that accessing semantic networks through words can activate abstract concepts and categories that ultimately affect behavior (Bargh et al. 2001). The usefulness of priming rst emerged as researchers explored ways of altering the environment to cause people to unknowingly think differently (Bargh and Chartrand 2000; Duncker 1945). Subsequent research shows that priming can affect judgments, goal motivation (Stajkovic et al. 2006), and individual decision-making (Bargh and Chartrand 2000). We believe priming is a potential method that may inuence group level outcomes by changing individual cognition. Priming is usually conducted on the individual level, and, to our knowledge, only one study has used priming at the group level (Dennis et al. 2012). The study by Dennis et al. (2012) used priming to affect individual motivation and creativity, leading to an improvement of performance on a group electronic brainstorming task. In their study, group performance (measured as the number of ideas generated by a group) was signicantly better for groups under a priming treatment than for groups under the placebo treatment. However, it would be difcult to conclude that priming improves group level performance based on the results of one study. During the Dennis et al. (2012) experiment, participants performed a brainstorming task where group performance was largely a function of the sum of individual performance. Other types of group activities, such as decision-making, require different cognitive and collaborative processes than those required in additive tasks such as brainstorming (Steiner 1972); thus, priming may have differing effects. In our research, we focus on priming two distinct concepts (one positive and one negative), to see which, if either, has a better result on group decision-making processes in virtual teams. First, we discuss achievement priming, a concept that has been effective for both individual performance and simple group brainstorming tasks. Second, we discuss regret priming, a concept that has been effective for individual decisionmaking tasks. 3 Study 1: Achievement and Attention Priming 3.1 Hypothesis Development Previous research shows that priming positive concepts can improve individual decision-making and task performance by improving individual cognition (Bargh et al. 2001; Galinsky and Moskowitz 2000). Bargh et al. (2001) postulated that priming the

123

878

V. L. Bartelt et al.

achievement concept would improve individual task performance. They found that achievement priming leads individuals to subconsciously create high goals for themselves in their current situation. Subjects were exposed to two different sets of priming words in a word search puzzle game, one related to achievement (e.g., compete, win, succeed) and another to neutral words (e.g., ranch, carpet, shampoo). Individuals primed with the achievement words performed better in the subsequent tasks than those primed with neutral words. Being primed with words associated with achievement activated the achievement concept through the semantic networks of individuals cognition, which led to individuals subconsciously creating goal pursuits of achievement (Bargh et al. 2001). Achievement priming was also used in a study by Shantz and Latham (2009) to improve individual brainstorming. In this case, the priming was done using an image of a woman winning a race, rather than by using words. Achievement priming was also positively related to group level performance for idea generation tasks. Dennis et al. (2012) conducted an experiment to investigate the priming effect on group level electronic brainstorming. In the experimental group, participants were primed with words related to achievement. After the priming, participants completed a task where they were asked to brainstorm ideas in groups of ve. The achievement prime activated an automatic goal in pursuit of achievement, which improved their motivation and ability to brainstorm as a team. Groups primed for achievement generated more ideas that were also more relevant, novel, and workable than ideas produced by the control groups. We believe the achievement prime should have the same effect in team decisionmaking tasks. We consider that the activation of achievement concepts in the minds of individuals should motivate them to work hard during the group task to overcome cognitive loads and biases that normally affect group decision-making tasks. Because decision-making tasks involve more complex team processes than simple brainstorming tasks, including the processes of disseminating and integrating crucial information, so the achievement concept alone may not be powerful enough to change group behavior. One of the reasons that computer-mediated groups sometimes have low decision-making performance is that group members dont pay attention to other group members information or opinions (Dennis 1996; Heninger et al. 2006; Hilmer and Dennis 2001). Paying attention to other team members is even more critical in situations where each group member is in possession of some unique information that cannot otherwise be accessed by other members. This condition makes it necessary for group members to collaborate on solving the group task. Thus, for a prime to work in this complex environment, it should not only activate high performance goals in team members, but also stimulate group members to pay attention to one another. In this manner, not only will individual behavior be changed, but the key group-level concept of integrating information should be improved. Despite the importance of attention in cognitive processes and collaborative work, we found no previous work on the effect of priming attention in individuals or groups. However, our reasoning suggests that the attention priming effect should strengthen the achievement priming effect for virtual team decision-making processes where attention is necessary for success. We propose that achievement priming and attention priming together should alter individual cognition (increased focus on information),

123

Virtual Team Decision-Making

879

Priming

Individual Cognition

Group Information Processing

Group Decision Quality

Fig. 1 Process of priming effects on group decision quality

which will be reected in the group process (increased dissemination and integration of information), which will lead to higher quality decisions, as shown in Fig. 1. Hypothesis 1 Individual team member exposure to achievement and attention priming prior to team decision-making will improve decision quality. 3.2 Methodology We conducted a lab experiment to test the effect of achievement and attention priming on virtual team decision-making. 3.2.1 Participants One hundred thirty undergraduate business students from a large state university were assigned to 26 ve-person teams and participated in this study. 92 percent of the participants were within the age range of 1921. The remaining 8 percent were within the 2327 age range. Thirty-ve percent of the participants were female. There were 11 teams in the neutral treatment, and 15 teams in the achievement/attention priming treatment. All participants received extra credit for their participation. 3.2.2 Task The task was to choose at most three students to admit to the university from a set of ve possible candidates. All teams selected exactly three candidates to admit. Researchers have used the task in previous experiments on virtual team decision-making (Dennis 1996; Robert and Dennis 2005). We selected this particular task because the application process is familiar to most participants. The student admission task was also appropriate because it is a hidden prole task as dened by Stasser (1992). In a hidden prole task, each team member makes independent decisions on incomplete information, and then they come together to discuss and make a team decision. Incomplete information consists of both common information, which is known to every participant, and unique information, which is known to only one participant. Because all information must be shared to reach the best decision, the hidden prole task is designed to facilitate discussion among team members. The team can only come to the correct decision if all the information is shared and attended to. Hidden prole tasks have been commonly used in team research because they are similar to the tasks real groups face (Stasser 1992); in real groups, not all participants have complete information to make the best decision individually. Participants were rst asked to read and learn facts about the candidates whose information would be shared. Four minutes were allowed for participants to read and

123

880

V. L. Bartelt et al.

come to an individual decision based on incomplete information. Then, 15 min were allowed for participants to work together as a team to share, read, and discuss their information with others in order to reach a team decision. Prior research determined this was a sufcient amount of time for this task. Eighteen pieces of information were distributed among the team members. Consistent with prior research, each participant was provided with six pieces of common information: the candidates names, gender, GPA, verbal SAT, and math SAT scores, and commitment to major (Dennis 1996; Robert and Dennis 2005). Two unique pieces of information were provided to each of the ve participants (including intended major, quality of the high school, commitment to their major, motivation to attend, parents afliation with the university, letters of recommendation, extra-curricular activities, missing courses, class rank, advanced placement courses, residence, and additional comments). Additional information included circumstantial information about the candidates, such as whether their high school grades have steadily improved or declined. Two facts contained in the unique information should have led to a denial of the applicants (one student was missing two courses and another student had a lower class rank from a poor quality school). Three directors of the undergraduate admissions ofce individually ranked the universitys admissions criteria. An average of the directors scores was referenced when designing the task of admitting the students to the university. After the task was designed, the director of admissions determined which candidates to reject and accept. This was used as a scoring key for the panel of applicants that the participants were given in the experiment. Participants used Gmail Chat to work together to discuss the task. This software is similar to other group chat software in that there is one lower window in which participants type ideas, with a larger scrolling window on top in which the ideas contributed by group members are displayed. The participants sat in separate cubicles in front of individual personal computers during the study so no verbal communication occurred. 3.2.3 Treatments There were two treatments: achievement/attention priming and neutral priming. We used the Scrambled Sentence Test (SST), a form of supraliminal priming (Bargh and Chartrand 2000; Srull and Wyer 1979). This technique has been used in prior priming research (Bargh and Chartrand 2000), including prior research on electronic brainstorming in teams (Dennis et al. 2012). We used the same priming procedure used by Dennis et al. (2012)a modied version of the SST delivered as a computer game. Participants were instructed to create an interesting headline for the university newspaper by using four out of the ve words that had been provided by the game. Each word was assigned a point value, and participants were encouraged to create a headline that generated the highest score. The participants were allowed 8 min to play the priming game. Prior research has determined both the strength and the duration of prime to be important factors for optimal priming effects, and 8 min was considered sufcient for the priming to take effect (Bargh and Chartrand 2000; Srull and Wyer 1979). The longer length of the game play design ensured that participants would not nish early, countering the possibility of decaying effects among participants. Prior research

123

Virtual Team Decision-Making

881

has found the immediacy of the task following the prime to affect the similarity of participants responses to the intended prime (Bargh and Chartrand 2000). The achievement/attention priming version of the game included one achievement word drawn from Bargh et al. (2001) in each set of ve words. Achievement words included win, award, honor, and so forth. In addition, we added one word related to paying attention in each set of ve words. Attention words included listen, noticed, and so forth. Both types of priming words were obtained as instructed in Bargh and Chartrands (2000) article. Synonyms of the intended meaning (either achievement or attention) were included in the word list. Neutral words were used for the control groups. The neutral words were taken from prior research (Bargh et al. 2001) and validated using the Affective Norms of English Words (ANEW) database (Bradley and Lang 1999). In this database, words are rated based on arousal and valance levels. Neutral words were selected from the database that produced moderate arousal and valence levels. Some examples of neutral priming words included, worker, room, and short. A full list of the priming words can be found in the appendix. 3.2.4 Dependent Variable Decision quality was determined for each team based on the correct admit/reject decision for each of ve candidates. The director of admissions at the university where the study was conducted assessed each candidate and determined whether that individual would be admitted. The decision performance scale ranges from 05. 3.2.5 Procedures Participants were randomly assigned to treatments and were provided with predetermined usernames to ensure anonymity among team members. Participants began the experiment by playing the priming game. Upon nishing the priming game (or after 8 min had elapsed), participants were provided 4 min to read the admission task information that they had been given, which contained both unique and common information about the candidates. They were then asked to make their individual decision to admit no more than three candidates to the university. Participants were told that each team member had been given both common information known to everyone and unique information only they knew. They were informed that it was important for them to contribute the information they knew and carefully read the information they received from others in order to make a good team decision. Next, participants worked together as a team using Gmail Chat to arrive at a team decision of no more than three candidates to admit. Then, participants completed a survey indicating their teams decision and individual demographic information. 3.3 Results We analyzed the data with one-way between-groups ANOVA. Teams primed for achievement and attention failed to have signicantly higher decision quality than teams receiving the neutral priming (F(1,24) = 0.700, p = 0.411).

123

882 Table 1 Achievement/attention prime results (Number of correct decisions) n Mean SD ANOVA F (1,24) ANOVA p value 0.411

V. L. Bartelt et al.

KruskalWallis p value 0.929

Achievement/attention Neutral

15 11

3.87 3.73

0.990 1.009

0.700

In order to examine one possible reason that the ANOVA tests were not statistically signicant, we tested whether the data met the assumptions of an ANOVA test. First, Levenes test was run to test homogeneity of variance on the dependent variable, decision quality. This test was used because the normality of the data is not assumed when using this procedure (Glass 1966). Decision quality was not signicant (Levene statistic = 0.687; df = 1,24; p = 0.416), indicating an equality of variance in the data. Second, to determine the normality of the data, the ShapiroWilk test was used on each treatment due to its appropriateness for sample sizes that are less than 50. Decision quality was found to be signicant ( p < 0.01) in all treatments, indicating that the data was not normal. Because the data was not normal, we used the KruskalWallis test, the nonparametric alternative to ANOVA, to test differences between groups. The results of the KruskalWallis test were not statistically signicant ( p = 0.929), indicating no difference in decision quality between groups. Therefore, Hypothesis 1 was not supported. Results are summarized in Table 1. Since decision quality was not signicant, we conducted a post-hoc analysis of the transcripts to identify whether the prime had any effect on earlier steps of the process presented in Fig. 1. We did not have the information available to test whether the priming changed how well the individuals paid attention and integrated information from others. However, we downloaded the transcripts from the experiment and examined the information exchanged by team members to see if the amount of information contributed to the discussion was different between the two treatments, and whether the amount of information shared in teams truly led to better decision quality. One rater coded the number of correct pieces of information in the transcript of each group. A second rater independently coded half of the transcripts. Inter-rater reliability was 81 percent (calculated as 1 - the number of disagreements / total number of codings), which is adequate. We used a one-way ANOVA to test whether the amount of information shared differed between treatments. The results indicate no difference between priming treatments (F = 3.428; df = 1,24; p = 0.076). We again tested whether the data met the assumptions of an ANOVA test. First, Levenes test was run to test homogeneity of variance. Information shared was not signicant (Levene statistic = 1.080; df = 1,24; p = 0.309), indicating an equality of variance in the data. To determine the normality of the data, the ShapiroWilk test was used on each treatment. Information shared was not signicant ( p > 0.05) in all treatments, indicating normality of the data. Thus, the assumptions of the ANOVA test are met, so we did not use a KruskalWallis test. Results are summarized in Table 2.

123

Virtual Team Decision-Making Table 2 Achievement/attention prime results (Information shared) n Achievement/attention Neutral 15 11 Mean 13.47 19.27 SD 7.376 8.580 ANOVA F (1,24) 3.428

883

ANOVA p value 0.076

A simple linear regression testing the effect of information sharing on decision quality was not signicant ( p = 0.534). Because the data is not linear, we also used a Kendalls tau correlation test, a common nonparametric alternative to linear regression, to see if there was a signicant correlation between information sharing and decision quality. This result was also not statistically signicant ( p = 0.417), indicating that even when team members shared more information, other elements of group information processing were not strong; participants did not fully integrate the information shared, and decision quality remained low.

3.4 Discussion The results from the achievement and attention priming experiment did not support Hypothesis 1, which means that achievement and attention priming does not improve group decision-making. The positive emotions evoked by achievement priming may have affected individuals in some ways, but were not sufcient to cause a signicant difference in overall group performance. The key question is: why didnt achievement and attention priming have an effect? We hypothesized that this priming would increase motivation and attention to information, but it did not. Priming did not lead to increased sharing of information, and even when team members shared more information for any reason, other team members did not integrate that information to make a quality decision. It could be that the achievement priming encouraged team members to behave in a more individualistic manner to achieve more; that is, it encouraged them to become more anchored on their original decisions and focus more on what they already knew, actually reinforcing the conrmation bias, rather than consider the new information received from other team members. How can we encourage team members to think more about what others contribute? Prior research in individual decision making has found self-blame regret priming to induce counter factual thinking (Reb and Connolly 2009; Zeelenberg and Pieters 2007). Counter factual thinking is the process of actively challenging the facts you currently have; that is, second-guessing what you know and asking yourself what if the facts you have are not true. This seedling of doubt that is planted in your mind causes you to ask questions that consider possibilities other than the initial position you rst held. Counter-factual thinking would result in seeking more information, which would lead to more informed decision-making and better decisions. Although its inuence may induce negative emotions, regret priming has been found to improve individual decision-making (Reb and Connolly 2009; Zeelenberg and Pieters 2007). We next describe a study we conducted to explore the effects of a negative prime, regret, on group performance.

123

884

V. L. Bartelt et al.

4 Study 2: Regret Priming 4.1 Hypothesis Development Regret is a counterfactual, or negative, cognitively-based emotional experience which results from visualizing a better situation if an alternate decision had been made (Gilbert et al. 2004; Reb and Connolly 2009; Zeelenberg and Pieters 2007). Regret usually involves some sort of self-berating due to wishing that the alternate decision had been made (Reb and Connolly 2009; Zeelenberg 1999). Regret is often associated with a feeling of self-blamea distinguishing factor between regret and disappointment (Gilbert et al. 2004; Zeelenberg et al. 1998). Research has begun to examine the role of negative emotions in the decisionmaking process, especially regret (Connolly and Zeelenberg 2002). Research has found anticipated regret to affect individual decision-making (Connolly and Zeelenberg 2002; Zeelenberg 1999). When individual decision makers do not consider alternate outcomes, a series of poor decision-making typically ensues (Reb and Connolly 2009). People prefer to avoid future regret (Reb 2008; Reb and Connolly 2009; Wroe et al. 2004), and especially avoid regret-inducing feedback (Reb and Connolly 2009; Zeelenberg and Beattie 1997; Zeelenberg et al. 1996). People even prefer to forego the possibility of positive outcomes in their desire to avoid negative experiences (Reb and Connolly 2009; Taylor 1991). Individuals who want to minimize the experience of regret about a future outcome often expend greater cognitive attention during the decision-making process (Connolly and Zeelenberg 2002; Reb 2008; Reb and Connolly 2009; Simonson 1992). Because individuals would like to avoid feeling regret in the future, they use more careful cognitive processes during decision-making, increasing attention to information in order to ensure that they do not lose key information that could result in a wrong decision. Self-blame regret may cognitively trigger participants to pay attention to and more thoroughly consider new information proposed by others during the decisionmaking process. As a result, regret priming has improved individual decision-making (Reb and Connolly 2009). Regret priming techniques have been used to increase the desire to read and consider available information. These techniques encourage individuals to engage in hypothetical thinking (Byrne and Girotto 2009), where they consider situations in which they wished they had behaved differently (Kahneman and Miller 1986; Wong et al. 2009). For example, prior research involving a decision task with monetary incentives found self-blame regret to improve decision quality (Reb and Connolly 2009). In that study, those primed with self-blame regret accepted more feedback and paid attention to it, which led to better decisions (Reb and Connolly 2009). Since prior literature has found regret priming to improve individual decisionmaking ability by increasing attention during the process (Byrne and Girotto 2009; Connolly and Zeelenberg 2002; Wong et al. 2009), it follows that regret priming should improve virtual team decision-making as well by affecting the cognition of team members so that they pay more attention to key information, again as depicted in Fig. 1.

123

Virtual Team Decision-Making

885

Hypothesis 2 Individual team member exposure to regret priming prior to team decision-making will improve decision quality. 4.2 Methodology We conducted a second lab experiment to test the effect of regret priming on virtual team decision-making. 4.2.1 Participants Participants were 200 undergraduate business students from a large state university assigned to 40 ve-person teams. There were 20 teams in the neutral prime treatment, and 20 teams in the regret prime treatment. All participants received extra credit for their participation. Demographic information for Study 2 was not available, but such information is likely very similar to the participants of Study 1. 4.2.2 Task The experiment used the same task as outlined in Sect. 3.2.2. All teams again selected exactly three candidates to admit. 4.2.3 Treatments In this experiment, participants played a computer game modeled after the Scrambled Sentence Test (SST), as described in Sect. 3.2.3. However, instead of using words related to achievement and attention, the version of the game used for Study 2 included regret words drawn from Reb and Connolly (2009). Regret words, using synonyms related to regret and remorse, were used to put participants in this mindset. The set of neutral words used for the neutral treatment was the same as used for the control group in Study 1. The full list of priming words can be found in the appendix. 4.2.4 Dependent Variable As in Study 1, decision quality was determined for each team based on the correct set of candidates who were accepted to the university. 4.2.5 Procedures The same procedures were followed for this experiment as those used in Study 1, as described in Sect. 3.2.5. 4.3 Results ANOVA analysis showed that the regret priming treatment had no signicant effect on decision quality (F (1,38) = 0.066, p = 0.798). We again tested whether the data met

123

886 Table 3 Regret prime results (Number of correct decisions) n Mean SD ANOVA F (1,38) ANOVA p value 0.0798

V. L. Bartelt et al.

KruskalWallis p value 0.813

Regret Neutral

20 20

3.30 3.20

1.174 1.281

0.066

Table 4 Regret prime results (Information shared) n Mean SD ANOVA F (1,38) ANOVA p value 0.011 KruskalWallis p value 0.050

Regret Neutral

20 20

41.60 27.75

20.699 10.672

7.074

the assumptions of an ANOVA test. First, Levenes test was run to test homogeneity of variance on the dependent variable, decision quality. Decision quality was not signicant (Levene statistic = 0.036; df = 1,38; p = 0.852), indicating an equality of variance in the data. To determine the normality of the data, the ShapiroWilk test was used on each treatment. Decision quality was found to be signicant ( p < 0.01) in all treatments, indicating that the data was not normal. The results of the KruskalWallis test were not statistically signicant ( p = 0.813), again indicating no difference in decision quality between groups. Therefore, Hypothesis 2 is not supported. Results are summarized in Table 3. We again conducted a post-hoc analysis of the transcripts to count the amount of information contributed to the discussion. An ANOVA indicates that the amount of information shared in the primed groups was signicantly higher than the information shared in control groups (F = 7.074; df = 1,38; p = 0.011). We next tested whether the data met the assumptions of an ANOVA test in order to conrm these results. First, Levenes test was run to test homogeneity of variance. Information shared was signicant (Levene statistic = 21.922; df = 1,38; p < 0.001), indicating a violation of the equality of variance assumption. To determine the normality of the data, the ShapiroWilk test was used on each treatment. Information shared was not signicant ( p = 0.113) in the control groups, indicating normality of the data. However, in the primed groups, information shared was signicant ( p = 0.009) indicating lack of normality. The KruskalWallis test was performed as an alternative to ANOVA. Results were statistically signicant ( p = 0.050), conrming the difference between groups. Results are summarized in Table 4. A simple linear regression testing the effect of information sharing on decision quality was not signicant ( p = 0.329). Because the data is not linear, we also used a Kendalls tau correlation test, a common nonparametric alternative to linear regression, to see if there was a signicant correlation between information sharing and decision quality. This result was also not statistically signicant ( p = 0.495), indicating that even though the prime caused some change at the individual level, the group level processing of information that was shared was not affected.

123

Virtual Team Decision-Making

887

4.4 Discussion Although prior research has found regret priming to affect individual decision-making (Reb and Connolly 2009), this study shows that this priming had a different effect on teams. It increased the amount of information contributed to the discussions, but it had no impact on decision quality. The prime was strong enough to affect individual behavior, but not strong enough to combat the cognitive biases and complex processes of information processing in a group decision-making task. We conclude that regret priming may not be an appropriate prime for improving team decision quality because of its negative effect on emotion. It may be that negative emotion escalates during team discussion, more so than during individual decision making. When all team members are primed negatively, it may be that the negative emotions feed the discussion and the negative contributions build on each other, intensifying the negative tone. This negativity may therefore be reinforced and become stronger in team discussions than in individual decision making. This strong negative tone may have a detrimental impact on team decision making (Baumann and Kuhl 2005) that interferes with the counter factual thinking that is triggered in individual decision making.

5 Discussion 5.1 Comparison to Previous Research A key reason that virtual teams often make poor decisions is that individuals within teams fail to consider unique information provided by others (Heninger et al. 2006; Hilmer and Dennis 2001). This may be due to the dual task interference that is created when participants attempt to both contribute information to discussions and read the large volume of information contributed by others (Heninger et al. 2006). This interference and resulting cognitive load may increase conrmation bias, which occurs when team members seek information that supports their initial pre-discussion individual decisions, rather than fully considering information that challenges those pre-discussion decisions (Ask and Granhag 2005). For these reasons, we aimed to understand whether achievement/attention and/or regret priming could affect individual cognition in such a way that individuals would pay attention to and use the information provided by other team members to improve group decision making. Both types of priming have been successful in research on individuals (Bargh et al. 2001; Reb and Connolly 2009), and achievement priming has been shown to improve the performance of electronic brainstorming groups (Dennis et al. 2012). Some examples of previous research where priming has been successful in these areas are shown in Table 5. The examples of individual-level priming in the table include only those studies that have used regret or achievement as priming conceptsthere are hundreds of individual-level priming studies on other concepts being primed. However, the table shows all known team-level priming studies, and highlights the need for research on priming in group decision-making.

123

888 Table 5 Examples of previous priming research by level of analysis and task Individual level Brainstorming Stajkovic et al. (2006) Achievement prime Delivered via SST Shantz and Latham (2009) Achievement prime Delivered via pictures Bargh et al. (2001) Achievement prime Delivered via SST Galinsky and Moskowitz (2000) Regret prime Delivered via scenarios Reb and Connolly (2009) Regret prime Monetary delivery Group level

V. L. Bartelt et al.

Dennis et al. (2012) Achievement prime Delivered via SST

Decision making

None

This study was the rst to explore the effects of priming on virtual team decisionmaking. While the results were not statistically signicant, this study helps to inform future research by explaining key processes and issues that must be addressed and giving an example of how priming studies can be carried out for group decisionmaking tasks. 5.2 Alternative Explanations of Results Based on the results of two experiments, we conclude that achievement/attention priming and regret priming do not signicantly improve group decision-making. Nonsignicant results are often not published because they do not present a clear story (Browman 1999). Nonetheless, results that fail to conrm theoretically based hypotheses can be important because they challenge existing theory, whether to suggest its revision or to point out important boundary conditions (Allchin 1999). Thus, the key questions following from our two studies are: Why did individual priming not improve group decision making in these initial experiments? How can priming be rened to have an effect? What is different between group brainstorming and group decision-making, and between individual decision-making and group decision-making that caused priming to have no impact? In the section below, we outline six plausible reasons for our results. The rst possible reason may be that priming inuences individual performance (and by extension, group performance in additive tasks that are essentially a sum of individual performance), but does not as easily affect performance when team members must work together to develop a consensus. Research shows that virtual teams have varying success in electronic settings depending upon the task (Kerr and Murthy 2004; Dennis et al. 2001; Santanen et al. 2004). Thus, priming and other effects of individual cognition also may have differing results. In a study where achievement priming was used at the group level (Dennis et al. 2012), participants performed a brainstorming

123

Virtual Team Decision-Making

889

task. Brainstorming is an additive task (Steiner 1972) in which group performance is largely a function of the sum of individual performance (i.e., the number of ideas produced by a group is dependent on the sum of ideas produced by individuals). Prior research has identied social loang as a common deterrent during brainstorming tasks (Shepherd et al. 1995-1996), yet the additive nature of brainstorming tasks means that if individual performance improves and an individual contributes one additional idea, then the overall group performance improves (assuming that the idea does not inhibit the production of ideas by other group members). If priming was successful on only some individuals in the group (or even one individual), the group would still have a higher number of ideas generated overall. This straightforward effect of individual priming for group brainstorming is not as simple in group decision-making. A second possible reason for the lack of impact on team decision-making may be that the priming techniques were not very strong. For decision-making tasks, it is important for all members of the group to pay attention to key information and arrive at the correct decision. Thus it could be that if priming is not effective for most group members, it may have limited effects on group performance for tasks that require reaching a consensus. A third possible reason is that the primes we used caused unintended effects on the decision-making process. For instance, it is possible that multiple people interacting in a regret mindset caused the discussion to be too critical, and thus the group was not cohesive enough to reach better decisions. Conversely, the regret prime might have triggered empathy with the candidates not chosen, making it harder for participants to make choice. Likewise, multiple people interacting in an achievement mindset might induce a focus on individual achievement rather than team consensus and thus team members may have failed to focus on others comments. Also, the individual achievement may have inadvertently sparked a spirit of competition between team members. A fourth possible reason why individual priming was not effective in group decisionmaking is that we may not have selected the proper primes. Because of the complex nature of decision-making as compared to individual decision-making or group brainstorming tasks, it may be that only particular primes work. Achievement/attention priming and regret priming emphasize the importance of comprehending key information but neglect the importance of organizing information to reach consensus among group members on the shared information. However, the key to high quality team decision-making, also the goal of group collaboration, is organizing shared information according to its importance and building consensus among group members. Fifth, the priming we selected may have reduced individuals condence, either by using negative emotions in Study 2, or by focusing attention on others, thereby reducing condence in self, in Study 1. Coupled with prior ndings indicating that collaborative technology lessens condence in team decisions (Cred and Sniezek 2003), the additional reduction of condence levels due to priming may have affected team decision quality to a greater extent than during individual decision-making. It is possible that implementing a positive prime that is conducive for thoughtfulness, such as accountability or personal responsibility, may improve both condence and team decision quality. Finally, it is possible that the way the prime was delivered may not have been effective for the purposes of this study. The computer game we used to deliver the

123

890

V. L. Bartelt et al.

Scrambled Sentence Test may not have been an effective priming delivery vehicle for group decision-making tasks. We chose text words as the priming technique because they are close to the modality of interest (typed text) and are simple to execute. During the Scrambled Sentence Test, the prime is delivered subconsciously by using a conscious task (Bargh and Chartrand 2000; Srull and Wyer 1979). The prime is also delivered semantically by word association (Bargh and Chartrand 2000). This type of prime may not have had as pronounced an effect during the team decision-making task. It is possible that other types of priming delivery methods would be more effective for team decision-making. 5.3 Implications for Future Research More research needs to be devoted to overall virtual team decision-making to better understand what factors inuence decision quality. First, it is possible that individual team members may have reacted differently to the primes. Different personalities have been affected differently by priming. Extroverts were more affected by incentive motivation and were also found to be more affected by positive priming (Robinson et al. 2010). Other individual differences play a role in priming effects. In an experiment about social judgment, it was found that people with a high need for cognition were greater affected by subtle priming (Petty et al. 2008). Subtle priming was more effective than blatant priming because those with a high need for cognition had a greater tendency to overanalyze situations. Due to greater cognitive resources devoted to the task, they were able to determine the prime if it was too obvious. If the same prime is given to each team member, possible differences in individual reactions to the prime may produce inconsistencies at the group level. A type of prime may be needed that specically targets individuals recall and memory. Over the years, experiments have studied the effect of priming on memory and recall (Graf et al. 1982; Jacoby and Dallas 1981; Tulving et al. 1982). It is possible that either of these types of priming would be more benecial to group decision-making abilities. Second, it is possible that the general team atmosphere and dynamics were not benecial to achieving optimal decisions. The primes we used were chosen based on their effects on individual decision-making abilities (Reb and Connolly 2009; Bargh et al. 2001) and group brainstorming (Dennis et al. 2012). Yet, other priming words have been found to affect behavior and persuasion and may be more appropriate for group interactions. Prior research has found words like partner primed feelings of trust in comparison to the word opponent (Burnham et al. 2000). Additionally, priming for goal-relevant cognition was found to affect persuasion (Strahan et al. 2002). Thus, a different type of priming that more appropriately affects group decisionmaking processes may be needed. Currently, priming experiments have explored group cohesion (Pendry and Carrick 2001; Pichon et al. 2007) and social norms (Blair and Banaji 1996; Otten and Wentura 1999). Prior research has found that those given the conformity prime were more likely to conform to the group than those who were given a neutral prime (Pendry and Carrick 2001). Those with the anarchy prime reacted similarly to those who performed the task in isolation, and were less likely to conform to group pressures

123

Virtual Team Decision-Making

891

(Pendry and Carrick 2001). These individual-level primes were found to be effective on changing group cohesion. However, none has done an analysis at the group level. It is possible that certain individual primes work differently in group situations and should be modied for this intent. Thus, one possible direction for future research is identifying the correct mechanism of priming that inuences group decision-making. For instance, a team that is more competitive may result in better decisions than a team that is more cooperative (Montoya-Weiss et al. 2001). Conversely, social sensitivity has been found to be a key factor in group decision making, at least for face-to-face teams (Woolley et al. 2010). Perhaps priming competition or social sensitivity could improve performance. It would also be interesting to see if group decision-making would improve if team members were given different primes. In this way, each team member would be approaching one another from a different mindset. This may counter the effects of groupthink by causing team members to have a heightened awareness of various mentalities. Third, although the priming we used did not affect decision quality, it is important to consider participants mindsets during the experiment, especially during individual tasks when priming has been found to be effective (Reb and Connolly 2009). A neutralizing prime may need to be introduced prior to running experiments to ensure that each person is operating from an intended mindset. Also, because priming effects have not been considered in IS research, results from previous experiments need to be revisited to ensure consistency of participants mindsets before the experiment commences. According to Bargh and Chartrand (2000), the effect of the priming often lasts 10 min after the prime. We had our participants begin the task immediately after priming ended, but the group discussion did not start until 4 min later. Future research could change the order of presentation, so that the group discussion begins immediately after the priming task. Since regret priming triggers a heightened emotional state (Larsen and Ketelaar 1991), conversations in the priming treatments have elicited more emotionally charged content. Analytical software tools like Linguistic Inquiry and Word Count could be used to determine the amount of emotional and cognitive processes that may have affected the priming treatments. Some prior research shows that negative emotions could impair decision-making abilities (Baumann and Kuhl 2005). If greater negative emotional content is found during the team discussions during the priming treatment, it is possible that these emotions are impairing decision-making abilities. Team discussions may intensify negative mindsets to a debilitating degree in comparison to the effect of the negative prime during individual decision-making. Future research invoking regret priming could also assess the mood of the subjects pre- and post-priming to better understand the effects on emotion. Fourth, other research methods such as qualitative data may help to provide insights on how priming affects team members mindsets and the behaviors that ensue. Alternately, brain studies could be performed to determine what is taking place in the brain throughout the decision-making process (Dimoka et al. 2012). Different brain functioning like working memory, emotions, and attention could be monitored. It is possible that particular moments in the decision-making process are more memorable than others. This could affect decision-making abilities. It is also possible that

123

892

V. L. Bartelt et al.

attention could decrease as the discussion continues, and this could ultimately impair decision-making abilities. Based on these revealing results, different communication tool interventions could be implemented to combat the potential pitfalls during the decision-making process. Fifth, another type of prime delivery mechanism other than the computer game version of the Scrambled Sentence Test (SST) may be more appropriate for team decisions. The SST is a type of implicit word-association prime (Bargh and Chartrand 2000). Our goal was to stimulate increased cognition, primarily with regard to typed text. Words of text are semantic and cognitive, whereas images are more emotional. The idea is to invoke the semantic concepts connected to these words into working memory. We chose text words because they are linked to these cognitive, semantic concepts, and because they are simple to execute. However, explicit primes may be more effective due to the sensory areas that are involved, which may be more appealing to the emotions and team dynamics. Prior research has found contextual (Meyers-Levy and Zhu 2007) and material primes (Kay et al. 2004) to affect decision-making which may have implications for the design of 3D virtual worlds. For instance, ceiling height variations in architectural designs primed consumers to process information differently (Meyers-Levy and Zhu 2007). Based on ceiling height, perceptions of freedom or connement caused consumers to choose relational or item-specic processing, respectively (Meyers-Levy and Zhu 2007). Additionally, presenting participants with business-related objects in the room caused those to perceive games to be more competitive and also affected the amount of money that the participants proposed to keep for themselves (Kay et al. 2004). Seeing an image of a woman winning a race caused employees to more effectively brainstorm due to the activation of the achievement concept (Shantz and Latham 2009). Would these effects also apply to virtual worlds?

5.4 Ethical Consideration A nal consideration of this research study is the ethical issues associated with deliberately manipulating people through subconscious means. We found that regret priming changed the way team members interacted with each other (though participants likely did not notice this change in behavior). Researchers and practitioners should fully evaluate and discuss the ethics of deploying such a technique in the eldthat is, outside of a research laboratory. The critical issue is to consider whether it is ethical to subconsciously induce changes in behavior that participants are not aware of, even though the change is for the better (in that it increased the amount of information exchanged in some cases). We recommend that well-dened rules or procedures be developed and put in place to regulate the usage of any form of priming, especially negative priming, before putting it in practice in the eld. Our participants gave informed consent to participate in a research study, although they remained unaware of its intended and actual effects, even after the study. If we were to deploy priming in the eld (separate and distinct from a research project), it may be wise to advise participants of what is going on and obtain their informed consent to being subconsciously manipulated.

123

Virtual Team Decision-Making

893

6 Conclusion This research explored the possibility of improving virtual team decision-making performance through priming. Although priming has been successful in individual and group-brainstorming tasks, priming did not signicantly inuence decision quality in two separate experiments of virtual team decision-making. Based on the results of our studies, we were able to gain key insights that will inform future research in this area. Such research is needed to better understand the effects of priming at the team level. First, we demonstrated that the priming techniques and words used in individual-level and group brainstorming studies were not effective in the same form for decision-making groups. Future researchers will need to be more creative and thoughtful in designing primes to work for decision-making groups. Second, we provide discussion and interpretation of the reasons that priming in decision-making groups is different than other situations where priming has been successful. We hope that future studies will take into account this discussion when considering alternative ways to prime decision-making groups.
Acknowledgments The authors would like to thank Rosh Dhanawade for his assistance in generating potential attention priming words and collecting data during Study 1.

References
Allchin D (1999) Negative results as positive knowledge, and zeroing in on signicant problems. Mar Ecol 793 Prog Ser 191:303305 http://www.int-res.com/articles/theme/m191p301.pdf. Antunes P, Ferreira A (2011) Developing collaboration awareness support from a cognitive perspective. In: Proceedings of the 44th annual hawaii international conference on system sciences, January 47, 2011, Computer Society Press, pp 110 Ask K, Granhag PA (2005) Motivational sources of conrmation bias in criminal investigations: the need for cognitive closure. J Investig Psychol Offender Proling 2(1):4363 Bargh JA, Chartrand TL (2000) A practical guide to priming and automaticity research. In: Reis H, Judd C (eds) Handbook of research methods in social psychology. Cambridge University Press, New York, pp 253285 Bargh JA, Gollwitzer PM, Lee-Chai A, Barndollar K, Trtschel R (2001) The automated will: nonconscious activation and pursuit of behavioral goals. J Pers Soc Psychol 81(6):10141027 Baumann N, Kuhl J (2005) Positive affect and exibility: overcoming the precedence of global over local processing of visual information. Motiv Emot 29(2):123134 Bell BS, Kozlowski SWJ (2002) A typology of virtual teams: implications for effective leadership. Group Organ Manag 27(1):1449 Blair IV, Banaji MR (1996) Automatic and controlled processes in stereotype priming. J Pers Soc Psychol 70(6):11421163 Bradley MM, Lang PJ (1999) Affective norms for english words (ANEW): instruction manual and affective ratings. Technical Report C-1. The Center for Research in Psychophysiology, University of Florida Briggs RO, de Vreede GJ, Nunamaker JF (2004) Collaboration engineering with thinklets to pursue sustained success with group support systems. J Manag Inf Syst 19(4):3164 Browman HI (1999) Negative results. Mar Ecol Prog Ser 191: 301302. http://www.int-res.com/articles/ theme/m191p301.pdf Burnham T, McCabe K, Smith VL (2000) Friend-or-foe intentionality priming in an extensive form trust game. J Econ Behav Organ 43(1):5773 Byrne RMJ, Girotto V (2009) Cognitive processes in counterfactual thinking. In: Markman KD, Klein WMP, Surh JA (eds) Handbook of imagination and mental simulation. Psychology Press, New York, pp 151160

123

894

V. L. Bartelt et al.

Chalfoun P, Frasson C (2011) Subliminal cues while teaching: HCI technique for enhanced learning. Adv Hum-Comput Interact 2011:115 Chudoba KM, Wynn E, Lu M, Watson-Manheim MB (2005) How virtual are we? measuring virtuality and understanding its impact in a global organization. Inf Syst J 15(4):279306 Cohen SG, Bailey DE (1997) What makes teams work: group effectiveness research from the shop oor to the executive suite. J Manag 23(3):239290 Connolly T, Reb J (2003) Omission bias in vaccination decisions: wheres the omission? Wheres the bias? Organ Behav Hum Decis Process 91(2):186202 Connolly T, Zeelenberg M (2002) Regret in decision making. Curr Dir Psychol Sci 11(6):212216 Cred M, Sniezek JA (2003) Group judgment processes and outcomes in video-conferencing versus faceto-face groups. Int J Hum-Comput Stud 59(6):875897 Curseu PL, Schruijer S (2008) The effects of framing on inter-group negotiation. Group Decis Negot 17(4):347362 Dean JW Jr, Sharfman MP (1993) Procedural rationality in the strategic decision making process. J Manag Stud 30(4):587610 Dean JW Jr, Sharfman MP (1996) Does decision process matter? A study of strategic decision-making effectiveness. Acad Manag J 39(2):368396 DeCoster J, Claypool HM (2004) A meta-analysis of priming effects on impression formation supporting a general model of information biases. Pers Soc Psychol Rev 8(1):227 Dennis AR (1996) Information exchange and use in group decision making: you can lead a group to information, but you cant make it think. MIS Q 20(4):433455 Dennis AR, Hilmer KM, Taylor NJ (19971998) Information exchange and use in GSS and verbal group decision making: effect of minority inuence. J Manag Inf Syst 14(3):6168 Dennis AR, Minas RK, Bhagwatwar A (2012) Sparking creativity: improving electronic brainstorming with individual cognitive priming. In: Proceedings of the 45th annual Hawaii international conference on system sciences, January 47, 2012, Computer Society Press, pp 139148 Dennis AR, Wixom BH, Vandenberg RJ (2001) Understanding t and appropriation effects in group support systems via meta-analysis. MIS Q 25(2):167193 Dimoka A, Banker RD, Benbasat I, Davis FD, Dennis AR, Gefen D, Gupta A, Ischebeck A, Kenning P, Pavlou PA, Mller-Putz G, Riedl R, vom Brocke J, Weber B (2012) On the use of neurophysiological tools in information systems research: Developing a research agenda for NeuroIS. MIS Q, forthcoming Duncker K (1945) On problem solving. Psychological monographs. Am Psychol Assoc 58(5) Espinosa JA, Armour F, Boh WF (2011) The role of group cognition in enterprise architecting. Proceedings of the 44th annual Hawaii international conference on system sciences, January 47, 2011, Computer Society Press, pp 110 Festinger L (1957) A theory of cognitive dissonance. Stanford University Press, Stanford Galinsky AD, Moskowitz GB (2000) Counterfactuals as behavioral primes: priming the simulation heuristic and consideration of alternatives. J Exp Soc Psychol 36(4):384409 Gilbert DT, Morewedge CK, Risen JL, Wilson TD (2004) Looking forward to looking backward: the misprediction of regret. Assoc Psychol Sci 15(5):346350 Glass GV (1966) Testing homogeneity of variances. Am Educ Res J 3(3):187190 Graf P, Mandler G, Haden P (1982) Simulating amnesic symptoms in normal subjects. Science 218(4578):12431244 Graf P, Schacter DL (1985) Implicit and explicit memory for new associations in normal and amnesic subjects. J Exp Psychol Learn Mem Cognit 11(3):501518 Heninger WG, Dennis AR, Hilmer KM (2006) Individual cognition and dual-task interference in group support systems. Inf Syst Res 17(4):415424 Higgins ET, Chaires WM (1980) Accessibility of interrelational constructs: implications for stimulus encoding and creativity. J Exp Soc Psychol 16(4):348361 Hilmer KM, Dennis AR (2001) Stimulating thinking in decision making: cultivating better decisions with groupware one individual at a time. J Manag Inf Syst 17(3):93114 Jacoby LL, Dallas M (1981) On the relationship between autobiographical memory and perceptual learning. J Exp Psychol Gen 110(3):306340 Jonas E, Schulz-Hardt S, Frey D, Thelen N (2001) Conrmation bias in sequential information search after preliminary decisions:an expansion of dissonance theoretical research on selective exposure to information. J Pers Soc Psychol 80(4):557571 Kahai SS, Carroll E, Jestice R (2007) Team collaboration in virtual worlds. SIGMIS Database 38(4):6168

123

Virtual Team Decision-Making

895

Kahneman D, Miller D (1986) Norm theory: comparing reality to its alternatives. Psychol Rev 93(2):136 153 Kay AC, Wheeler SC, Bargh JA, Ross L (2004) Material priming: the inuence of mundane physical objects on situational construal and competitive behavioral choice. Organ Behav Hum Decis Process 95(1):8396 Kerr DS, Murthy US (2004) Divergent and convergent idea generation in teams: a comparison of computermediated and face-to-face communication. Group Decis Negot 13(4):381399 Kersten GE, Cray D (1996) Perspectives on representation and analysis of negotiation: towards cognitive support systems. Group Decis Negot 5(46):433467 Kolfschoten GL (2011) Cognitive load in collaborationbrainstorming. In: Proceedings of the 44th annual Hawaii international conference on system sciences, January 47, 2011, Computer Society Press, pp 19 Kolfschoten GL, Lee C (2010) Self-guiding group support systems: can groups use GSS without support? In: Proceedings of the 43rd annual Hawaii international conference on system sciences, January 58, 2010, Computer Society Press, pp 16 Kolfschoten GL, Vreede G-J, Pietron LR (2011) A training approach for the transition of repeatable collaboration processes to practitioners. Group Decis Negot 20(3):347371 Koriat A, Feuerstein N (1976) The recovery of incidentally acquired information. Acta Psychol 40(6):463 464 Larsen RJ, Ketelaar T (1991) Personality and susceptibility to positive and negative emotional states. J Pers Soc Psychol 61(1):132140 Martin A, Chao LL (2001) Semantic memory and the brain: structure and processes. Curr Opin Neurobiol 11(2):194201 Martz WB, Shepherd MM (2004) Group consensus: the impact of multiple dialogues. Group Decis Negot 13(4):315325 McGrath JE (1991) Time, interaction, and performance (TIP): a theory of groups. Small Group Res 22(2):147174 Meyers-Levy J, Zhu R (2007) The inuence of ceiling height: the effect of priming on the type of processing that people use. J Consum Res 34(2):174186 Montoya-Weiss MM, Massey AP, Song M (2001) Getting it together: temporal coordination and conict management in global virtual teams. Acad Manag J 44(6):12511262 Nagasundaram M, Dennis AR (1993) When a group is not a group. Small Group Res 24(4):463489 OLeary D (2011) The emergence of individual knowledge in a group setting: mitigating cognitive fallacies. Group Decis Negot 20(1):318 Otten S, Wentura D (1999) About the impact of automaticity in the minimal group paradigm: evidence from affective priming tasks. J Soc Psychol 29(8):10491071 Pendry L, Carrick R (2001) Doing what the mob do: priming effects on conformity. Eur J Soc Psychol 31(1):8392 Petty RE, DeMarree KG, Briol P, Horcajo J, Strathman AJ (2008) Need for cognition can magnify or attenuate priming effects in social judgment. Pers Soc Psychol Bull 34(7):900912 Pichon I, Boccato G, Saroglou V (2007) Nonconscious inuences of religion on prosociality: a priming study. Eur J Soc Psychol 37(5):10321045 Powell CA (1994) Cognitive hurdles in the use of decision support systems to enhance problem understanding. Group Decis Negot 3(4):413421 Reb J (2008) Regret aversion and decision process quality: effects of regret salience on decision process carefulness. Organ Behav Hum Decis Process 105(2):169182 Reb J, Connolly T (2009) Myopic regret avoidance: feedback avoidance and learning in repeated decision making. Organ Behav Hum Decis Process 109(2):182189 Reinig B (2003) Toward an understanding of satisfaction with the process and outcomes of teamwork. J Manag Inf Syst 19(4):6583 Reinig B, Briggs RO, Shepherd MM, Yen J, Nunamaker JF, Jr (19951996) Affective reward and the adoption of group support systems: productivity is not always enough. J Manag Inf Syst 12(3):171185 Reinig B, Shin B (2002) The dynamic effects of group support systems on group meetings. J Manag Inf Syst 19(2):303325 Robert L, Dennis AR (2005) The paradox of richness: a cognitive model of media choice. IEEE Trans Prof Commun 48(1):1021

123

896

V. L. Bartelt et al.

Robinson MD, Moeller SK, Ode S (2010) Extraversion and reward-related processing: incentive motivation in affective priming tasks. Emotion 10(5):615626 Santanen EL, Briggs RO, de Vreede GJ (2004) Causal relationships in creative problem solving: compaing faciliatation interventions for ideation. J Manag Inf Syst 20(4):167197 Schacter DL (1987) Implicit memory: history and current status. J Exp Psychol Mem Learn Cognit 13(3):501518 Schmidt JB, Montoya-Weiss MM, Massey AP (2001) New product development decision-making effectiveness: comparing individuals, face-to-face teams, and virtual teams. Decis Sci 32(4):575600 Shantz A, Latham GP (2009) An exploratory eld experiment of the effect of subconscious and conscious goals on employee performance. Organ Behav Hum Decis Process 109(1):917 Shepherd MM, Briggs RO, Reinig BA, Yen J, Nunamaker JR (19951996) Invoking social comparison to improve electronic brainstorming: Beyond anonymity. J Manag Inf Syst 12(3): 155170 Simonson I (1992) The inuence of anticipating regret and responsibility on purchase decisions. J Consum Res 19(1):105118 Srull TK, Wyer RS Jr (1979) The role of category accessibility in the interpretation of information about persons: some determinants and implications. J Pers Soc Psychol 37(10):16601672 Stahl G (2006) Analyzing and designing the group cognition experience. Int J Co-op Inf Syst 15(2):157178 Stajkovic AD, Locke EA, Blair ES (2006) A rst examination of the relationships between primed subconscious goals, assigned conscious goals, and task performance. J Appl Psychol 91(5):11721180 Stasser G (1992) Information salience and the discovery of hidden proles by decision-making groups: a thought experiment. Organ Behav Hum Decis Process 52(1):156181 Steiner ID (1972) Group process and productivity. Academic Press, New York Strahan EJ, Spencer SJ, Zanna MP (2002) Subliminal priming and persuasion: striking while the iron is hot. J Exp Soc Psychol 38(6):556568 Taylor SE (1991) Asymmetric effects of positive and negative events: the mobilization-minimization hypothesis. Psychol Bull 110(1):6785 Tulving E, Schacter DL, Stark HA (1982) Priming effects in word-fragment completion are independent of recognition memory. J Exp Psychol Learn Mem Cognit 8(4):336342 Webster DM, Richter L, Kruglanski AW (1996) On leaping to conclusions when feeling tired: mental fatigue effects on impressional primacy. J Exp Soc Psychol 32(2):181195 Whitworth B, Gallupe B, McQueen R (2000) A cognitive three-process model of computer-mediated group interaction. Group Decis Negot 9(5):431456 Woolley AW, Chabris CF, Pentland A, Hashmi N, Malone TW (2010) Evidence for a collective intelligence factor in the performance of human groups, Science 330 (29 October):686688 Wong EM, Galinsky AD, Kray LJ (2009) The counterfactual mind-set: a decade of research. In: Markman KD, Klein WMP, Surh JA (eds) Handbook of imagination and mental simulation. Psychology Press, New York, pp 161174 Wroe AL, Turner N, Salkovskis PM (2004) Understanding and predicting parental decisions about early childhood immunizations. Health Psychol 23(1):3341 Zeelenberg M (1999) The use of crying over spilled milk: a note on the rationality and functionality of regret. Philos Psycol 12(3):325340 Zeelenberg M, Beattie J (1997) Consequences of regret aversion 2: additional evidence for effects of feedback on decision making. Organ Behav Hum Decis Process 72(1):6378 Zeelenberg M, Beattie J, van der Pligt J, de Vries NK (1996) Consequences of regret aversion: effects of expected feedback on risky decision making. Organ Behav Hum Decis Process 65(2):148158 Zeelenberg M, Pieters R (2007) A theory of regret regulation 1.0. J Consum Psychol 17(1):318 Zeelenberg M, van Dijk WW, Manstead ASR, van der Pligt J (1998) The experience of regret and disappointment. Cognit Emot 12(2):221230

123

You might also like