Content uploaded by Guihyun Park
Author content
All content in this area was uploaded by Guihyun Park
Content may be subject to copyright.
A Multilevel Model of Minority Opinion Expression and Team
Decision-Making Effectiveness
Guihyun Park and Richard P. DeShon
Michigan State University
The consideration of minority opinions when making team decisions is an important factor that
contributes to team effectiveness. A multilevel model of minority opinion influence in decision-making
teams is developed to address the conditions that relate to adequate consideration of minority opinions.
Using a sample of 57 teams working on a simulated airport security-screening task, we demonstrate that
team learning goal orientation influences the confidence of minority opinion holders and team discussion.
Team discussion, in turn, relates to minority influence, greater decision quality, and team satisfaction.
Implications for managing decision-making teams in organizations are discussed.
Keywords: teams, goal orientation, minority influence, team effectiveness
Team discussion is a common and important component of
decision making in organizations. The widespread use of team
discussion for decision making is due, in part, to the belief that
teams make better decisions than individuals because teams use a
larger pool of relevant information (Baron & Kerr, 2003; Forsyth,
2006; Hinz, 1990), increase opportunities to identify and correct
mistaken assumptions, factual errors, and reasoning errors (Zim-
bardo, Butler, & Wolfe, 2003), and increase opportunities to learn
from other members’ perspectives, build shared understandings of
the task, and facilitate coordination of members’ behaviors
(Cannon-Bowers, Tannenbaum, Salas, & Volpe, 1995).
Despite its popularity, team decision making frequently fails to
yield the expected gains in decision accuracy, and the process
often results in lower satisfaction for team members (Hackman,
1988; Ilgen, Major, Hollenbeck, & Sego, 1995). Some of the most
striking examples of ineffective team decisions can be found in the
groupthink literature: the Challenger space shuttle tragedy, the
failed invasion of Cuba by the United States (i.e., the Bay of Pigs),
and the targeting of innocent civilians in Vietnam (Janis, 1982).
More recently, the Enron scandal, which resulted in the loss of
thousands of jobs, more than $60 billion in market value, and more
than $2 billion in pension plans, stemmed in part from the poor
group decision making of its board of directors (“Burden of
Enron,” 2006).
One reason why the expected positive outcomes of team dis-
cussion are often not realized is the failure to consider dissenting
or minority opinions. The expression of minority opinions helps
teams make quality decisions by preventing them from prema-
turely moving toward consensus (Nemeth & Chiles, 1988) and
encourages teams to develop multiple perspectives on issues that
contribute to higher quality decisions (De Dreu & West, 2001). In
addition, when minority opinions are considered, the holders of the
minority opinion perceive greater control in the decision process,
resulting in increased satisfaction and greater willingness to re-
main a part of the team (Drake & Mitchell, 1977; Parker, 1993).
Therefore, to improve team decision-making effectiveness, it is
necessary to better understand the processes that result in the fuller
consideration of minority opinions in team discussions.
Multilevel Process Model
The existing literature clearly demonstrates the importance of
incorporating minority opinions into the team decision-making
process. However, relatively little is known about the variables
that relate to minority influence and how these variables translate
into effective team decision making (e.g., De Dreu, 2007; Martin
& Hewstone, 2001). Given that the expression of minority opin-
ions occurs in the context of team decision making, a multilevel
perspective of the process is needed. Specifically, the team pro-
vides a top-down context for the members of the team. Team level
variables likely influence the willingness of the minority opinion
holder to express the opinion. Bottom-up processes also function at
the same time because individuating characteristics of the minority
opinion holder likely relate to the individual’s decision to express
the minority opinion, and once expressed, the minority opinion
likely influences the ensuing team discussion. The current article
suggests that team goal orientation is likely a potent state related to
the multilevel processes of the minority influence.
Figure 1 presents our multilevel model of minority opinion
expression and team decision-making effectiveness. We argue that
team learning goal orientation influences team discussion, and it
also should influence minority opinion holders’ confidence. The
confidence of minority opinion holders, in turn, affects the likeli-
hood of minority opinion expression and facilitates team discus-
This article was published Online First August 16, 2010.
Guihyun Park and Richard P. DeShon, Department of Psychology,
Michigan State University.
Previous versions of this article were presented at the 23rd annual
conference of the Society for Industrial and Organizational Psychology,
San Francisco, California, April 2008.
We are grateful to Daniel R. Ilgen, Neal Schmitt, and Linn Van Dyne for
their helpful comments on a previous version of this article.
Correspondence concerning this article should be addressed to Guihyun
Park, who is now at the School of Social Sciences, Singapore Management
University, 90 Stamford Road, Singapore 178903. E-mail: parkguih@
gmail.com
Journal of Applied Psychology © 2010 American Psychological Association
2010, Vol. 95, No. 5, 824– 833 0021-9010/10/$12.00 DOI: 10.1037/a0019939
824
sion. Team discussion should result in increased minority influ-
ence. Finally, minority influence is associated with better team
performance and greater team satisfaction. Each of these relation-
ships is expanded on in the following section.
Minority Opinion Holder Confidence
Minority opinions can increase team decision quality by initi-
ating the discussion of alternatives and expanding the team’s
knowledge base (Ilgen, Hollenbeck, Johnson, & Jundt, 2005;
Moscovici, 1985). This process can be beneficial even if the
minority opinion is inaccurate (Schulz-Hardt, Brodbeck, Mojzisch,
Kerschreiter, & Frey, 2006). For this beneficial process to occur,
however, the minority opinion holder must first express the opin-
ion, and then the team must discuss the opinion and related issues.
An important individual-level variable relating to both the likeli-
hood of minority opinion expression and the team’s reception of
the expressed opinion is the minority opinion holder’s confidence
in his or her opinion.
Nemeth and Kwan (1987) highlighted that the confidence of mi-
nority team members in their opinion relates to a greater willingness
to share their opinions that, in turn, leads to greater minority influence
on team decisions. Further, the confidence with which minority
opinions are expressed can influence the consideration that the
opinion receives from other team members. Kerr (2001) argued
that minority opinion holders who show confidence can compen-
sate for their lack of popularity and exert influence over the
majority faction. He explained that people infer the substance of an
opinion from the presenter’s behavioral style (Kerr, 2001). A
minority opinion receives little attention from other team members
if the opinion holder is not confident and quickly succumbs to the
normative pressure of following the majority opinion. Similarly,
Nemeth and Wachtler (1974) also found that a group is more likely
to consider minority opinions when the opinion is expressed with
confidence. Therefore, we hypothesize the following:
Hypothesis 1: Minority opinion holders’ confidence is posi-
tively related to team discussion.
Team-Level Learning Goal Orientation
Many possible team-level variables influence team discussion
and the internal states of the team members (individual level)
involved in the discussion. As we discuss next, team learning goal
orientation (e.g., DeShon, Kozlowski, Schmidt, Milner, & Wiec-
hmann, 2004) is a particularly promising variable that is likely
related to team discussion both directly and indirectly through its
relationship with minority opinion holder confidence. Teams with
high levels of learning goal orientation have a shared perception
that their teams are aiming to develop their competence by seeking
out challenges and using errors as opportunities to increase task
understanding. Analyzing qualitative coding of text messages ex-
changed during team discussion, LePine (2005) found that learning
goal oriented teams were more likely to communicate openly
regarding alternatives and were eager to determine the causes of
discrepant information. Teams with a high learning goal orienta-
tion are more likely to view the sharing of opinions as an effort to
improve the team decision rather than to interfere with it. In
addition, teams with high learning goal orientation exert less
normative pressure and, as a result, minority opinions are more
likely to be regarded as opportunities for learning than as threats to
the majority. On the basis of these arguments, we hypothesize the
following:
Hypothesis 2A: Team learning goal orientation is positively
related to team discussion.
Team learning goal orientation is also likely related to minority
opinion holder confidence. Teams with a high learning goal ori-
entation focus on developing competence and view errors and
mistakes as diagnostic information rather than indicators of incom-
petence. As a result, they are more likely to perceive different
opinions as opportunities to learn about the task rather than view-
ing minority opinions as threats. LePine (2005) found that teams
with a high learning goal orientation were more likely to be
respectful toward and supportive of team members who make
mistakes. The open and constructive discussion environment that
occurs in high learning goal oriented teams makes it possible for
minority opinion holders to develop and enhance their opinion
confidence and participate in discussions of alternatives (Edmond-
son, 2002). Papaioannou and Kouli (1999) manipulated team
learning goal orientation by making students who worked as
members of a four-person sports team focus on developing task-
related competencies; the results demonstrated that individuals
who perceived their teams as having a high learning goal orienta-
Figure 1. Multilevel model of minority opinion expression and team effectiveness.
825
MINORITY OPINION INFLUENCE
tion showed a high level of self-confidence on the task. In sum-
mary, when teams have a high learning goal orientation and
support different views, individuals are less likely to have norma-
tive concerns, allowing the development of increased confidence.
Therefore, we hypothesize the following:
Hypothesis 2B: Team learning goal orientation is positively
related to minority confidence.
Team Discussion and Minority Influence
Stasser and Titus (1985) suggested that team discussion could
be understood as an information-sampling process from the com-
bined pool of information held by the team members. They sug-
gested that shared information has a greater chance of being
mentioned in the initial stages of such a discussion. However, as
the pool of shared information is depleted in the early stages of the
discussion, the chance of mentioning unique information in-
creases. Supporting this notion, Campbell and Stasser (2006)
found that, as teams spend more time in discussion, the probability
of minority opinions being discussed increases. Larson, Foster-
Fishman, and Keys (1994) found that teams in a training condition
that were instructed to spend time scrutinizing and elaborating
each member’s opinion discussed a greater amount of unshared
information during team discussion. Finally, Kerr and Tindale
(2004) suggested that allowing sufficient time for decision-making
groups’ discussion facilitates information exchange and decreases
bias toward shared information. Thus, low levels of team discus-
sion may result in teams agreeing prematurely on decisions while
ignoring the contribution of unique opinions, thus increasing the
bias toward shared information. Conversely, increased discussion
among team members allows a balanced perspective of opinions of
both majority and minority factions, which relates to a greater
influence of minority opinions on team decision. Thus, we hypoth-
esize the following:
Hypothesis 3: Team discussion is positively related to minor-
ity influence on team decisions.
Minority Influence and Team Effectiveness
Team decision-making processes that adequately consider mi-
nority opinions also result in increased team decision quality and
team member satisfaction with the team. Extensive research on
team decision making has demonstrated the positive effect of
minority influence on team decision quality (e.g., De Dreu &
Beersma, 2001; Nemeth & Chiles, 1988; Schwenk, 1990). Minor-
ity dissension encourages teams to consider multiple perspectives
and alternatives, thereby helping team members increase their
understanding of the task (De Dreu & Beersma, 2001). Nemeth
and Chiles (1988) suggested that confronting the minority opinion
prevents teams from prematurely reaching a consensus, which
facilitates the discovery of better alternatives. In addition, minority
opinions give teams the opportunity to question current beliefs and
help detect errors in their assumptions (Schwenk, 1990). There-
fore, members on teams that allow the minority to influence team
decisions have a greater chance of finding more correct answers,
leading to greater performance within the team.
In addition, minority influence on team decisions facilitates
team members’ feeling of esteem and control and increases mem-
bers’ satisfaction with the team as a whole. Minority opinion
members’ influence on team decisions indicates reduced confor-
mity pressure in the team in that members do not feel obligated to
follow others’ opinions because they happen to be endorsed by the
majority faction (De Dreu & De Vries, 1996). Thus, greater
minority influence on team decisions implies reduced conformity
pressure, which provides members with the opportunity to express
their opinions and freely engage in the discussion without feeling
intimidated by the judgments of other members (Edmondson,
1999). As such, members have the opportunity to maintain their
self-esteem and develop feelings of being respected by other
members, which should positively influence their satisfaction
(Miller & Monge, 1986). Morrison and Milliken (2000) suggested
that, when individuals are subject to social cues that discourage
their opinions, they develop a perceived lack of control over the
situation and decreased satisfaction. Similarly, Parker (1993)
found that nurses who believed their dissenting opinions would be
heard have a greater sense of control and show less intention of
leaving the organization. Therefore, members on teams that allow
the minority to influence team decisions have greater feelings of
being respected by other members, which relates to greater satis-
faction with the team. Thus, we hypothesize the following:
Hypothesis 4A: Minority influence is positively related to
team performance.
Hypothesis 4B: Minority influence is positively related to
team satisfaction.
Method
Participants
Participants were 180 undergraduate students who voluntarily
took part in the study. Nine participants did not complete the study
or did not follow instructions and were dropped from the analyses,
resulting in a total of 171 participants (57 teams) for use in data
analysis. Of the participants, 67% were women, and 77% were
White. In addition, 85% were between the ages of 18 and 20 years,
and 46% were freshmen.
Task Overview
Participants performed a passenger luggage inspection task as a
team of three randomly assigned individuals.
1
A total of 20 x-ray
images of suitcases were utilized in the experiment. Some of these
images contained weapons. The key decision was whether to
search the bag if it was believed to contain a weapon or to clear the
bag if it was believed that it contained no weapon. For each image,
team members received 10 s to inspect the image individually and
then made an individual decision without talking to their team-
mates. After making the initial individual decision, team members
1
The task used in this research is a modified version of a luggage-
screening task developed by Daniel R. Ilgen and his students with stimulus
materials from the Transportation Security Administration.
826 PARK AND DESHON
discussed their opinions and then provided a single team decision.
2
Four blocks of five decision events were performed during the
experiment. Immediately after a practice image and every five
decisions thereafter, participants completed a survey booklet con-
sisting of relevant measures, including team goal orientation and
satisfaction.
Procedure
On arrival participants were randomly assigned to a three-
person team. Participants were told that they were to assume the
role of an airport security team and were going to process x-ray
images of passenger luggage. Participants were given the oppor-
tunity to interact with each other before engaging in the perfor-
mance trials to develop a shared understanding of the team task
and a sense of being a member of a team. Following a brief
introduction to the experiment, participants were given 3 min to
get to know their team members and to provide a nickname for
their team. After the teams decided on their nicknames, they were
provided with a training manual and asked to study together for 3
min in preparation for a practice trial.
The experimental task was presented on a computer, and the
three team members viewed the same computer display. All team
member actions and interactions were videotaped over the course
of the entire experiment. We projected x-ray images of passenger
luggage onto the team computers for each trial, and teams were
asked to make and input the team’s decision with a mouse con-
nected to the computer. A practice trial was provided to familiarize
the team members with the task and to provide an opportunity for
team processes to begin functioning before the actual data collec-
tion. During the practice trial, participants were given an unlimited
amount of time to discuss their decisions, and the experimenter
answered any questions regarding the task procedures. Also, par-
ticipants were informed that they could use the training manual as
a reference throughout the experiment. To increase participants’
involvement, the experimenter told participants they would be
tested on the manual as well as on the task at the end of the
experiment.
For each performance trial, an x-ray image of luggage was
projected onto the screen. Team members first examined the image
and then silently noted individual decisions on their paper about
whether the luggage should be cleared or searched and how
confident they were about their individual decision. The initial
individual decision phase lasted 10 s, and participants were not
allowed to see other participants’ responses during the period.
After the individual decisions, team members freely discussed the
luggage to reach their final decision. Team members were encour-
aged to make their team’s final decision collectively and enter their
decision in the computer. There was no time limit on team dis-
cussion phase. Immediately after teams entered their final decision
for each trial, the computer gave “Correct!” or “Incorrect!” as
feedback. Teams earned 10 points for each correct decision and 0
points for each incorrect decision. No penalty was enforced for
incorrect decisions. No points were given for the practice trial.
After the practice trial, participants completed their first question-
naire. A total of 20 performance trials (four blocks of five trials)
were conducted and sessions typically lasted about 1 hr and 20
min.
Measures
Confidence. For each individual decision made during the
initial individual decision phase (the first 10 s for each trial),
confidence was measured with a 5-point scale measure developed
by Henry and Sniezek (1993). The item asked participants to rate
their confidence in their search/clear decision according to a scale
ranging from extremely confident (5) to not confident at all (1).
Minority confidence. Minority opinion individuals’ confi-
dence in their individual decision was assessed for each trial. For
example, if Person A decided to search a suitcase and Persons B
and C decided to clear a suitcase in the third trial of the second
block, we used Person A’s confidence rating for that trial to
represent minority confidence. In a small number of cases (6% of
the total blocks), all members consistently agreed for an entire
block of trials, yielding no minority opinion. These cases were
treated as missing values for that block of trials.
Team cognitive ability. Past research suggests team cognitive
ability is related to team effectiveness (e.g., Bell, 2007; Stewart,
2006). Therefore, we used team cognitive ability as a control
variable in all of the analyses. At the beginning of the study,
participants reported their ACT or SAT scores, which were then
converted to zscores by means of their respective national nor-
mative data. The team cognitive ability was then formed for each
team by taking the average of the members’ scores.
Team learning goal orientation. State team goal orientation
was measured with a modified version of goal orientation items
developed by Elliot and McGregor (2001). The modification en-
tailed changing the referent from the individual to the team so that
it incorporated a set of referent-shift items (Chan, 1998). This
measure was administered immediately after the practice trial and
prior to Blocks 2, 3, and 4. Three items were used: “Right now, my
team wants to learn as much as possible on this task.” “Right now,
my team hopes to gain a broader and deeper knowledge on this
task.” “Right now, my team prefers material that really challenges
us so we can learn new things on this task.” Team members
indicated their agreement with each statement on a 5-point Likert
scale ranging from strongly disagree (1) to strongly agree (5).
Coefficient alphas for this scale ranged from .87 (Block 1) to .90
(Block 4). The intraclass correlations (ICC[1]) obtained at Block 1
(.19,
2
⫽.48), Block 2 (.19,
2
⫽.47), Block 3 (.16,
2
⫽.45),
and Block 4 (.18,
2
⫽.47) provided justification for aggregating
this measure to the team level of analysis by calculating the
average value as a meaningful variable.
Team discussion. All team discussions were videotaped, and
two coders, the first author and an undergraduate research assis-
tant, rated each individual’s contribution to the team discussion on
the basis of (a) the number of task relevant utterances an individual
made during discussion and (b) the extent to which the individual
discussed aspects of the decision task with other members in the
2
The data were collected as a part of a larger study that assessed two
different decision-making structures. In one condition, individuals made a
collective decision, and anyone could enter the decision into the computer.
In the other condition, individuals made a collective decision as a team, but
a single team member was responsible for entering the team decision into
the computer. This distinction is not relevant to the current focus; therefore,
the difference between the conditions was controlled for with a dummy
coded variable in all reported analyses.
827
MINORITY OPINION INFLUENCE
team in an interactive and engaged manner. Utterances were coded
in terms of the total number of complete phrases that an individual
spoke during each block that were task relevant (e.g., asking
questions, answering questions, talking about objects in the lug-
gage, and talking about the regulation of harmful objects). Partic-
ipation ratings were made on a scale of 1 (low)to3(high). Each
coder independently coded half of the video recordings of the team
discussion. To estimate rater agreement and reliability, the two
raters independently coded the discussions from a randomly se-
lected set of five teams. On average, the two raters agreed on 97%
of the coded utterances made by each individual by each block.
The interrater reliability correlation from the two raters was .85 for
participating ratings. For each block of trials, team utterance and
team participation variables were formed by summing team mem-
ber utterances and team member participation ratings for each
team.
Next, the team utterance and the team participation ratings were
standardized and summed to form a team discussion variable. The
correlation between the team utterance and team participation
variables was significant (average correlation within blocks ⫽.45;
correlation across all blocks ⫽.45). Utterance and participation
each represent quantity and quality of team discussion: Utterances
reflect general levels of task-relevant activity during team discus-
sion and capture the average quantity of the team member’s task
contributions, whereas participation reflects the average level of
team member engagement in coordinated team discussion and
reflects the quality of a team member’s contributions. It is possible
that an individual might speak often during discussion but not do
so in a coordinated, engaged, and interactive manner. Therefore,
we believe that the combination of the two measures provides a
more construct-relevant representation of team participation than
either would alone.
Minority influence. The extent to which team members with
minority opinions influenced the team decision was calculated as
the percentage of times that the minority opinion holder’s initial
opinion was selected as a team decision instead of the majority
opinion holders’ initial opinion. When all team members agreed on
their initial individual opinion for a trial, then minority influence
could not exist for that trial. Therefore, minority influence is based
on the subset of occasions where there was disagreement among
team members’ initial individual decisions. To be concrete, con-
sider a scenario where, for a block of five trials, a team made two
decisions aligned with the minority opinion, two decisions aligned
with the majority opinion, and one decision based on a unanimous
opinion. Minority influence is then calculated for that block of
trials by forming the ratio of two (the number of times that
minority opinion was selected as the team decision) divided by
four (the total number of trials where individuals disagreed on their
initial opinions).
Team performance. For each correct decision made, teams
were awarded 10 points. Team performance was measured by the
number of points that the team earned. A total of 20 trials of
luggage screening were carried out; team performance was calcu-
lated for each block of five trials.
Team satisfaction. Team satisfaction was measured with a
4-item Team Satisfaction Scale adapted from Cook, Hepworth,
Wall, and Warr (1981). To reduce participants’ workload, this
measure was administrated only three times—at the beginning
(after Block 1), in the middle (after Block 3), and at the end (after
Block 4) of the experiment. The scale included the following
items: “All in all, how satisfied are you with your members in your
team?” “All in all, how satisfied are you with your team’s perfor-
mance on the task?” “How satisfied are you the progress you made
in the task?” “Considering the effort you put into the task, how
satisfied are you with your team’s performance?” Teams re-
sponded to the items with a 7-point Likert scale ranging from
extremely dissatisfied (1) to extremely satisfied (7). The factor
analysis supported a one-factor solution. Coefficient alphas for this
scale were .88, .88, and .89 at Blocks 1, 3, and 4, respectively. The
intraclass correlations (ICC[1]) obtained at Block 1 (.31,
2
⫽
.57), Block 3 (.34,
2
⫽.58), and Block 4 (.24,
2
⫽.52) provided
justification for aggregating this measure to the team level of
analysis by calculating the average value within teams to represent
the team level of satisfaction.
Results
The frequencies of observed decision across all teams are pro-
vided in Table 1. Unanimous decisions occurred when all team
members agreed on the initial individual decision and then made a
team decision consistent with the unanimous individual decisions.
Majority decisions occurred when two of the three team members
made the same individual decision (majority), and the final team
decision was consistent with the majority individual decisions.
Finally, minority decisions occurred when two of the three team
members made the same individual decision (majority), but the
final team decision was consistent with the minority individual
decision. On average, teams arrived at unanimous decisions in
62% to 66% of trials in each block (roughly three out of five
decisions per block). Teams also arrived at minority decisions in
15% to 20% of trials in each block (roughly one out of five
decisions per block).
Table 1
Frequency of Majority, Minority, and Unanimous Decision Across Blocks
Decision
Block 1 Block 2 Block 3 Block 4
MSD%MSD%MSD%MSD%
Majority
a
0.70 0.57 14 0.98 0.47 20 0.97 0.46 19 0.93 0.56 19
Minority
b
0.98 0.67 20 0.76 0.75 15 0.93 0.95 19 0.75 0.69 15
Unanimous
c
3.32 0.83 66 3.25 0.83 65 3.10 0.83 62 3.31 0.76 66
Note. Each block consisted of five decisions.
a
When a majority opinion was selected.
b
When a minority opinion was selected.
c
When everyone agreed on a decision.
828 PARK AND DESHON
Table 2 presents descriptive statistics for the study variables. It
is interesting to note that team discussion varied more between
teams than within teams over time. Performance, on the other
hand, varied more within teams over time than between teams.
Also, the general pattern of correlations is consistent with our main
hypotheses. Between teams, learning goal orientation was related
to both team discussion and minority influence, minority confi-
dence was related to discussion, and minority influence, perfor-
mance, and satisfaction were related to each other. Within teams,
minority confidence was related to discussion and minority influ-
ence, discussion was related to minority influence, and perfor-
mance was related to satisfaction. It appears that the relationships
among team learning goal orientation, minority confidence, and
discussion were primarily between-team phenomena, whereas the
relationships among minority confidence, discussion, minority in-
fluence, and performance existed both between and within teams
over time.
Analysis Overview
To evaluate our hypotheses, we used a random intercepts model
with an autoregressive error structure. The random intercepts ac-
count for nonindependence of observations due to team member-
ship (i.e., ICC[1]) and the autoregressive error structure accounts
for nonindependence of errors over time within persons. Our basic
model may be represented as follows:
Yij ⫽b0关i兴⫹兵b1共Abilityi兲⫹b2共Conditioni兲其
⫹b3共Timeij兲⫹b4共Xij兲⫹eij,
b0关i兴⬇N共0,b0兲,
where i⫽1...N;j⫽1...T;Y
ij
represents one of the dependent
variables in the study (e.g., discussion, minority confidence, team
performance, etc . . .) for the ith team at the jth time period; b
0[i]
is
a random effect reflecting heterogeneity in team intercepts; Ability
i
and Condition
i
represent the values of the covariates for the ith
team; Time
ij
is a variable corresponding to a value that ranges
between 0 and 3 for the ith team measured at the jth time point; X
ij
represents specific independent variables used to examine the
hypothesized relationships(e.g., minority confidence, team learn-
ing goal orientation, and minority influence) for the ith team
measured at the jth time point; and b
1
,b
2
,b
3
, and b
4
represent
fixed effects that capture the common relationships between the
model predictors and the Y
ij
.
The results of these analyses are summarized in Tables 3 and 4,
including the parameter estimates (␥) and significance tests for
each predictor. ␥represents the average or fixed effect estimate
and is roughly analogous to conducting a separate regression for
each individual or team and averaging the resulting regression
weights across individuals or teams. Following Rosenthal and
Rubin’s (2003) approach, we computed the R
2
value for each fixed
effect in Tables 3 and 4 by squaring the R
equivalent
estimate.
Hypothesis Testing
Hypothesis 1. Consistent with Hypothesis 1, minority confi-
dence was positively related to team discussion: ␥⫽0.40; F(1,
155) ⫽9.04, p⬍.01. This result indicates that those teams with
confident minority opinion members were more likely to engage in
team discussion.
Hypothesis 2. Consistent with Hypothesis 2A team learning
goal orientation was positively related to team discussion, ␥⫽
0.47; F(1, 174) ⫽4.83, p⬍.05. Consistent with Hypothesis 2B,
team learning goal orientation was positively related to minority
opinion members’ confidence in their opinions, ␥⫽0.23; F(1,
171) ⫽5.92, p⬍.05. This indicates that when teams have a high
learning goal orientation, individuals with minority opinions felt
greater confidence in their opinion. Extending the findings from
Hypotheses 1, this result suggests that team learning goal orienta-
tion facilitates minority confidence in their opinions and team
discussion.
Hypothesis 3. Consistent with Hypothesis 3, team discussion
was positively related to minority opinion member influence, ␥⫽
Table 2
Between- and Within-Team Correlations Among Study Variables
Variable M
SD
1234567Between team Within team
1. Ability 0.65 0.47 — 1.00 — — — — — —
2. Team learning GO 3.39 0.35 0.28 .03 1.00 ⫺.06 .01 ⫺.14 ⫺.14 ⫺.07
3. Minority confidence 3.01 0.47 0.49 .25 .22 1.00 .20 .33 ⫺.03 .03
4. Team discussion 0.12 1.42 0.89 .17 .37 .27 1.00 .33 .07 .10
5. Minority influence
a
0.42 0.22 0.28 .18 .38 ⫺.02 .48 1.00 .08 .08
6. Performance 28.53 5.57 10.47 .02 ⫺.04 ⫺.23 .28 .37 1.00 .70
7. Satisfaction 5.22 0.70 0.46 ⫺.08 .12 ⫺.21 .33 .35 .60 1.00
Note. The correlations below the diagonal represent between-team correlations that are computed with teams’ aggregated scores. The correlations above
the diagonal represent the average within-team correlations over time. The values greater than .26 (between team) and .15 (within team) are significant at
␣⫽.05. GO ⫽goal orientation.
a
When examining Table 2, care should be taken to distinguish between the notion of minority decisions represented in Table 1 and the notion of minority
influence represented in Table 2. The minority decision percentages in Table 1 are based on blocks of five trials. These percentages represent the ratio of
the number of times that the minority opinion was selected as the team’s decision in the set of five trials. Conversely, the average minority influence
percentage in Table 2 represents the average ratio that the minority opinion was selected as the team’s decision for only those trials where disagreement
occurred across four blocks. In other words, the minority influence percentage in Table 2 represents the number of times that teams adopted the minority
opinion holders’ initial individual decision over the majority opinion holders’ initial individual decision. When all members agreed on their individual
decisions, minority or majority influence does not exist for that trial. Therefore, unanimous decisions were not included when calculating minority influence.
829
MINORITY OPINION INFLUENCE
0.09; F(1, 127) ⫽28.06, p⬍.01. This indicates that, as teams
engaged in active discussions, they were more likely to adopt the
minority opinion as their team decision. Specifically, an increase
in one unit of the team discussion rating led to a 10% increase in
the likelihood of the minority opinion being selected as the team
decision. Although we tested the hypotheses with the combined
discussion variables, in the Appendix we also present the results
based on the separated team discussion constituents.
Hypothesis 4. Consistent with Hypothesis 4A, minority opin-
ion member influence was positively related to team performance,
␥⫽6.60; F(1, 204) ⫽10.11, p⬍.01, indicating that greater
minority influence relates to better team performance. Consistent
with Hypothesis 4B, minority influence was positively related to
team satisfaction, ␥⫽0.38; F(1, 132) ⫽5.91, p⬍.01. This
indicates that greater minority influence relates to greater satisfac-
tion with the team.
Testing of the mediation model. To evaluate the proposed
model, we examined the significance of each hypothesized indirect
relationships with the z⬘method, as recommended by MacKinnon,
Lockwood, Hoffman, West, and Sheets (2002). Results support the
mediated relationships in our model. Specifically, the mediated
effects of team discussion were significant: team discussion me-
diated the relationship between team goal orientation and minority
influence, (z⬘)⫽2.03, p⬍.01, and team discussion mediated the
relationship between minority confidence and minority influence,
(z⬘)⫽2.61, p⬍.01. Minority confidence mediated the relation-
ship between team goal orientation and discussion, (z⬘)⫽1.88,
p⬍.01. Finally, the mediation effects of minority influence were
significant, such that the minority influence mediated the relation-
ship between discussion and performance, (z⬘)⫽2.71, p⬍.01,
and between discussion and satisfaction, (z⬘)⫽2.20, p⬍.01.
Thus, we found support for the model that minority confidence,
discussion, and minority influence play mediating roles between
learning goal orientation and team effectiveness.
Discussion
This research is the first to approach the issue of minority
influence in decision-making teams from a multilevel perspective
by examining how team learning goal orientation relates to the
processes of minority influence that, in turn, relate to team effec-
tiveness. Our results suggest that team learning goal orientation
facilitates minority influence through increased team discussion.
Specifically, when team members viewed their goal as learning
and mastering the task, having different opinions and seeking
diverse opinions appeared to be encouraged in a manner that
promoted higher confidence among minority opinion holders.
Likewise, the members of teams that had a high learning goal
orientation were more involved in active team discussion.
Consistent with previous findings, the current study showed a
relationship between team discussion with minority influence and
team decision quality. Findings from the current study also suggest
that team discussion relates to minority influence and team satis-
faction. Thus, the current findings replicate and extend previous
studies on minority influence by examining team discussion as a
Table 3
Relation of Team Learning Goal Orientations, Minority Confidence, and Team Discussion
Model ␥SE Num df Den df F R
2
DV: Minority confidence
Ability 0.12 0.10 1 49 1.38 .03
Team learning GO 0.23 0.10 1 171 5.92
ⴱ
.03
DV: Team discussion
Ability 0.43 0.36 1 44 1.40 .03
Team learning GO 0.47 0.21 1 174 4.83
ⴱ
.03
Minority confidence 0.40 0.13 1 155 9.04
ⴱⴱ
.06
Note. ␥⫽average regression weight; Num ⫽numerator; Den ⫽denominator; DV ⫽dependent variable;
GO ⫽goal orientation.
ⴱ
p⬍.05.
ⴱⴱ
p⬍.01.
Table 4
Relation of Discussion on Minority Influence and Effectiveness
Model ␥SE Num df Den df F R
2
DV: Minority influence
Ability ⫺0.02 0.06 1 45 0.11 .00
Team discussion 0.09 0.01 1 127 28.06
ⴱⴱ
.17
DV: Team performance
Ability 0.04 1.65 1 77 0.00 .00
Minority influence 6.60 2.08 1 204 10.11
ⴱⴱ
.05
DV: Team satisfaction
Ability ⫺0.13 0.18 1 51 0.56 .01
Minority influence 0.38 0.16 1 132 5.91
ⴱⴱ
.04
Note. ␥⫽average regression weight; Num ⫽numerator; Den ⫽denominator; DV ⫽dependent variable.
ⴱⴱ
p⬍.01.
830 PARK AND DESHON
precursor of minority influence on team decisions and team effec-
tiveness.
Limitations
Despite these contributions, several limitations of this research
should be kept in mind when interpreting these findings. All teams
in the current study were newly formed and worked on a novel task
of screening the luggage. It remains unclear whether these findings
will generalize to different organizational contexts in which team
members are more experienced with each other and the task and
are aware of each other’s different sources of expertise.
The effect sizes found in the present research are not large.
Despite the small effect sizes, the widespread use of decision-
making teams and the importance of the decisions being made by
the teams suggest that the present results can have substantial
practical implications (e.g., Abelson, 1985; Cortina & Landis,
2008). The security screening task used in this research is a good
example of when small team decision errors can result in disas-
trous outcomes.
As is generally the case for nonexperimental investigations, the
causal mechanism underlying our observed results remains ambig-
uous, particularly with respect to the relative magnitude of the
direct and indirect effect of team learning goal orientation on
minority confidence. More advanced analytic techniques (e.g.,
multivariate time series analysis) with direct manipulation of team
learning goal orientation that could be used to model the underly-
ing dynamics require far more intensive longitudinal data than we
were able to collect in the present research (Stone-Romero &
Rosopa, 2008).
Although the current study focuses on mainly positive implica-
tions of minority influence, such as increased problem-solving
abilities, it is important to note that minority influence can also
imply negative group processes. For example, minority opinions
can cause conflict among group members and yield suboptimal
outcomes (e.g., Lott & Lott, 1965; Ten Velden, Beersma, & De
Dreu, 2007). Future studies should examine factors that distinguish
the positive impacts of minority influence from the negative im-
pacts.
Implications
The current study suggests emphasizing team learning goal
orientation as a viable decision-making intervention for organiza-
tions. Previously, selecting a member in the team to take the role
of dissenter, or the devil’s advocate, during team discussions was
suggested as a technique to increase minority influence and team
decision quality by inducing cognitive conflicts of team members
(Schweiger, Sandberg, & Rechner, 1989; Schwenk, 1990). How-
ever, this technique can cause stress for team members, especially
those who are selected as the devil’s advocate, and can create
unnecessary friction among team members, which may decrease
their satisfaction (Nemeth & Staw, 1989; Van Dyne & Saavedra,
1996). The results of this study suggest that the emphasis of
learning goals can increase team decision quality without sacrific-
ing members’ satisfaction.
Many different characteristics of both individuals and teams
affect minority influence in decision-making teams; future studies
need to consider the diverse characteristics of both individuals and
teams to broaden the understanding of minority influence pro-
cesses. For example, a team-level climate on procedural justice
may influence the extent to which team members respect other
members’ opinions and unique suggestions (Van Prooijen, Karre-
mans, & Van Beest, 2006). Further research is needed to identify
other individual and team characteristics that improve team
decision-making processes and outcomes.
References
Abelson, R. P. (1985). A variance explanation paradox: When a little is a
lot. Psychological Bulletin, 97, 129–133.
Baron, R. S., & Kerr, N. L. (2003). Group process, group decision, group
action. Buckingham, PA: Open University Press.
Bell, S. T. (2007). Deep-level composition variables as predictors of team
performance: A meta-analysis. Journal of Applied Psychology, 92, 595–
615.
Burden of Enron’s collapse on Skilling as sentencing approaches. (2006,
October 22). USA Today. Retrieved from http://www.usatoday.com/
money/industries/energy/2006-10-22-skilling-sentence_x.htm
Campbell, J., & Stasser, G. (2006). The influence of time and task demon-
strability on decision-making in computer-mediated and face-to-face
groups. Small Group Research, 37, 271–294.
Cannon-Bowers, J. A., Tannenbaum, S. I., Salas, E., & Volpe, C. E. (1995).
Defining competencies and establishing team training requirements. In
R. A. Guzzo & E. Salas (Eds.), Team effectiveness and decision making
in organizations (pp. 333–381). San Francisco, CA: Jossey-Bass.
Chan, D. (1998). Functional relations among cohesions in the same content
domain at different levels of analysis: A typology of composition mod-
els. Journal of Applied Psychology, 83, 234–246.
Cook, J. D., Hepworth, S. J., Wall, T. D., & Warr, P. B. (1981). The
experience of work: A compendium of 249 measures and their use.
London, England: Academic Press.
Cortina, J. M., & Landis, R. S. (2008). When small effect sizes tell a big
story, and when large effect sizes don’t. In C. E. Lance & R. J.
Vandenberg (Eds.), Statistical and methodological myths and urban
legend: Doctrine, verity and fable in the organizational and social
sciences (pp. 287–308). New York, NY: Routledge.
De Dreu, C. K. W. (2007). Minority dissent, attitude change, and group
performance. In A. R. Pratkanis (Ed.), The science of social influence:
Advances and future progress (pp. 247–270). New York, NY: Psychol-
ogy Press.
De Dreu, C. K. W., & Beersma, B. (2001). Minority dissent in organiza-
tional teams: Implications for group innovation. In C. K. W. De Dreu &
N. K. De Vries (Eds.), Group consensus and minority influence: Impli-
cations for innovation (pp. 258–283). London, England: Blackwell.
De Dreu, C. K. W., & De Vries, N. K. (1996). Differential processing and
attitude change following majority versus minority arguments. British
Journal of Social Psychology, 35, 77–90.
De Dreu, C. K. W., & West, M. A. (2001). Minority dissent and team
innovation: The importance of participation in decision making. Journal
of Applied Psychology, 86, 1191–1201.
DeShon, R. P., Kozlowski, S. W. J., Schmidt, A. M., Milner, K. R., &
Wiechmann, D. (2004). A multiple-goal, multilevel model of feedback
effects on the regulation of individual and team performance. Journal of
Applied Psychology, 89, 1035–1056.
Drake, B. H., & Mitchell, T. R. (1977). The effects of vertical and
horizontal power on individual motivation and satisfaction. Academy of
Management Journal, 20, 573–591.
Edmondson, A. (1999). Psychological safety and learning behavior in work
teams. Administrative Science Quarterly, 44, 350–383.
Edmondson, A. (2002). The local and variegated nature of learning in
organizations: A group-level perspective. Organization Science, 13,
128–146.
831
MINORITY OPINION INFLUENCE
Elliot, A. J., & McGregor, H. A. (2001). A 2 ⫻2 achievement goal
framework. Journal of Personality and Social Psychology, 80, 501–519.
Forsyth, D. R. (2006). Group dynamics (4th ed.). Belmont, CA: Thomson.
Hackman, J. R. (1988). The design of work teams. In J. Lorsch (Ed.),
Handbook of organizational behavior (pp. 315–342). Englewood cliffs,
NJ: Prentice-Hall.
Henry, R. A., & Sniezek, J. A. (1993). Situational factors affecting judg-
ments of future performance. Organizational Behavior and Human
Decision Processes, 54, 104–132.
Hinz, V. B. (1990). Cognitive and consensus processes in group recogni-
tion memory performance. Journal of Personality and Social Psychol-
ogy, 59, 705–718.
Ilgen, D. R., Hollenbeck, J. R., Johnson, M., & Jundt, D. (2005). Teams in
organizations: From I-P-O models to IMOI models. Annual Review of
Psychology, 56, 517–544.
Ilgen, D. R., Major, D. A., Hollenbeck, J. R., & Sego, D. J. (1995). Raising
an individual decision-making model to the team level: A new research
model and paradigm. In R. Guzzo & E. Salas (Eds.), Team decision
making in organizations (pp. 113–148). San Francisco, CA: Jossey-
Bass.
Janis, I. L. (1982). Groupthink: Psychological studies of policy decisions
and fiascoes (2nd ed.). New York, NY: Houghton Mifflin.
Kerr, N. L. (2001). Is it what one says or how one says it? Style vs
substance from an SDS Perspective. In C. K. W. De Dreu & N. K. De
Vries (Eds.), Group consensus and minority influence: Implications for
innovation (pp. 201–228). Oxford, England: Blackwell.
Kerr, N. L., & Tindale, R. S. (2004). Small group decision making and
performance. Annual Review of Psychology, 55, 623–656.
Larson, J. R., Foster-Fishman, P. G., & Keys, C. B. (1994). Discussion of
shared and unshared information in decision-making groups. Journal of
Personality and Social Psychology, 67, 446–461.
LePine, J. A. (2005). Adaptation of teams in response to unforeseen
change: Effects of goal difficulty and team composition in terms of
cognitive ability and goal orientation. Journal of Applied Psychology,
90, 1153–1167.
Lott, A. J., & Lott, B. E. (1965). Group cohesiveness as interpersonal
attraction: A review of relationships with antecedent and consequence
variables. Psychological Bulletin, 64, 259–309.
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., &
Sheets, V. (2002). A comparison of methods to test mediation and other
intervening variable effects. Psychological Methods, 7, 83–104.
Martin, R., & Hewstone, M. (2001). Conformity and independence in
groups: Majorities and minorities. In M. A. Hogg & R. S. Tindale (Eds.),
Handbook of social psychology: Group processes (pp. 209–234). Mal-
den, MA: Blackwell.
Miller, K. I., & Monge, P. R. (1986). Participation, satisfaction, and
productivity: A meta analytic review. Academy of Management Journal,
29, 727–753.
Morrison, E. W., & Milliken, F. J. (2000). Organizational silence: A barrier
to change and development in a pluralistic world. Academy of Manage-
ment Review, 25, 706–725.
Moscovici, S. (1985). Social influence and conformity. In G. Lindzey & E.
Aaronson (Eds.), The handbook of social psychology (pp. 347–412).
New York, NY: Random House.
Nemeth, C., & Chiles, C. (1988). Modeling courage: The role of dissent in
fostering independence. European Journal of Social Psychology, 18,
275–280.
Nemeth, C., & Kwan, J. (1987). Minority influence, divergent thinking,
and detection of correct solutions. Journal of Applied Social Psychology,
17, 788–799.
Nemeth, C., & Staw, B. M. (1989). The tradeoffs of control and innovation
in groups and organizations. In L. Berkowitz (Ed.), Advances in exper-
imental social psychology (Vol. 22, pp. 175–210). New York, NY:
Academic Press.
Nemeth, C., & Wachtler, J. (1974). Creating the perceptions of consistency
and confidence: A necessary condition for minority influence. Sociom-
etry, 37, 529–540.
Papaioannou, A., & Kouli, O. (1999). The effect of task structure, per-
ceived motivational climate and goal orientations on students’ intrinsic
motivation and anxiety. Journal of Applied Sport Psychology, 11, 51–71.
Parker, L. E. (1993). When to fix it and when to leave: The relationship
among perceived control, self-efficacy, dissent and exit. Journal of
Applied Psychology, 78, 949–959.
Rosenthal, R., & Rubin, D. B. (2003). requivalent: A simple effect size
indicator. Psychological Methods, 8, 492–496.
Schulz-Hardt, S., Brodbeck, F. C., Mojzisch, A., Kerschreiter, R., & Frey,
D. (2006). Group decision making in hidden profile situations: Dissent
as a facilitator for decision quality. Journal of Personality and Social
Psychology, 91, 1080–1093.
Schweiger, D. M., Sandberg, W. R., & Rechner, P. L. (1989). Experiential
effects of dialectical inquiry, devil’s advocacy, and consensus ap-
proaches to strategic decision making. Academy of Management Jour-
nal, 32, 745–772.
Schwenk, C. R. (1990). Effects of devil’s advocacy and dialectical inquiry
on decision making: A meta-analysis. Organizational Behavior and
Human Decision Processes, 47, 161–176.
Stasser, G., & Titus, W. (1985). Pooling of unshared information in group
decision making: Biased information sampling during discussion. Jour-
nal of Personality and Social Psychology, 48, 1467–1478.
Stewart, G. L. (2006). A meta-analytic review of relationships between
team design features and team performance, Journal of Management, 52,
29–54.
Stone-Romero, E. F., & Rosopa, P. J. (2008). The relative validity of
inferences about mediation as a function of research design character-
istics. Organizational Research Methods, 11, 326–352.
Ten Velden, F. S., Beersma, B., & De Dreu, C. K. W. (2007). Majority and
minority influence in group negotiation: The moderating influence of
social motivation and decision rules. Journal of Applied Psychology, 92,
259–268.
Van Dyne, L., & Saavedra, R. (1996). A naturalistic minority influence
experiment: Effects on divergent thinking, conflict and originality in
work-groups. British Journal of Social Psychology, 35, 151–168.
Van Prooijen, J. W., Karremans, J. C., & Van Beest, I. (2006). Procedural
justice and the hedonic principle: How approach versus avoidance
motivation influences the psychology of voice. Journal of Personality
and Social Psychology, 91, 686–697.
Zimbardo, P., Butler, L., & Wolfe, V. (2003). Cooperative college exam-
inations: More gain, less pain when students share information and
grades. Journal of Experimental Education, 71, 101–126.
(Appendix follows)
832 PARK AND DESHON
Appendix
Alternative Model With Separate Dimensions of Team Discussion
Received February 13, 2008
Revision received April 13, 2010
Accepted April 22, 2010 䡲
Online First Publication
APA-published journal articles are now available Online First in the PsycARTICLES database.
Electronic versions of journal articles will be accessible prior to the print publication, expediting
access to the latest peer-reviewed research.
All PsycARTICLES institutional customers, individual APA PsycNET威database package sub-
scribers, and individual journal subscribers may now search these records as an added benefit.
Online First Publication (OFP) records can be released within as little as 30 days of acceptance and
transfer into production, and are marked to indicate the posting status, allowing researchers to
quickly and easily discover the latest literature. OFP articles will be the version of record; the
articles have gone through the full production cycle except for assignment to an issue and
pagination. After a journal issue’s print publication, OFP records will be replaced with the final
published article to reflect the final status and bibliographic information.
6.6**
.11**
.25**
.18*
.12
.32**
.23*
Team States
• Learning Orientation
Minority Opinion
Holder States
• Confidence
Team Discussion (Quantity)
• Task Relevant Utterance
Team Discussion (Quality)
• Participation Rating
Minority
Influence
Team Performance
Team Satisfaction
.13**
.38**
Figure A1. This figure presents the pattern of results when the constituents of team discussion (i.e., task
relevant utterances and participation ratings) are kept distinct. The general patterns of relationships were
consistent with our hypotheses. Both task relevant utterances and participation ratings showed significant
relationships with minority influence—this suggests the validity of the two discussion dimensions on predicting
minority influence. Whereas task-relevant utterance was significantly related to team learning goal orientation
and minority confidence, participation rating was significantly related only to minority confidence. This
indicates that when a team has a high learning goal orientation, they are more likely to engage in a greater
amount of task-relevant discussion. However, they are not necessarily participating in their discussion in a more
engaging manner.
ⴱ
p⬍.05.
ⴱⴱ
p⬍.01.
833
MINORITY OPINION INFLUENCE
A preview of this full-text is provided by American Psychological Association.
Content available from Journal of Applied Psychology
This content is subject to copyright. Terms and conditions apply.