Social Motives in Negotiation 1
Conflicting Social Motives in Negotiating Groups
Laurie R. Weingart
David A. Tepper School of Business
Carnegie Mellon University
Pittsburgh, PA 15213
Jeanne M. Brett
J. L. Kellogg School of Management
Evanston, IL 60208
Melbourne Business School
University of Melbourne
Carlton, Victoria, 3053
Philip L. Smith
Department of Psychology
University of Melbourne
Parkville, Victoria, 3052
Tel: 61 3 8344 6343
Fax: 61 3 9347 6618
Keywords: NEGOTIATION, SOCIAL MOTIVES, GROUP PROCESSES
Manuscript under review: November 15, 2006
This research was funded by NSF Grant SBR96-1671 and funds from the Dispute Resolution
Research Center, Northwestern University. The authors would like to thank participants in
seminar series’ at Cornell University, Duke University, New York University, University of
Amsterdam, and Washington University for helpful comments.
Social Motives in Negotiation 2
Negotiators’ social motives (cooperative versus individualistic) influence their strategic
behaviors. This study used multi-level modeling and sequential analysis to test hypotheses
regarding how negotiators’ social motives and the composition of the group influence group
members’ negotiation strategies. Four-person groups negotiating a 5 issue mixed-motive decision
making task were videotaped, transcribed, and coded. Group composition included two
homogeneous conditions (all cooperators and all individualists) and three heterogeneous
conditions (3 cooperators/1 individualist; 2 cooperators/2 individualists; 1 cooperator/3
individualists). Results showed that cooperative negotiators adjusted their use of integrative and
distributive strategies in response to the social motive composition of the group, but
individualistic negotiators did not. Results from a sequential analysis showed that cooperators
responded more systematically to others’ behaviors than individualists. They also redirected the
negotiation depending on group composition.
Social Motives in Negotiation 3
Conflicting Social Motives in Negotiating Groups
Cross-functional teams are shepherding new products to market; cross-cultural teams are
searching for economies of scale in the global workplace; and teams of experienced managers
and brash young MBAs are trying to recapture the momentum of internet companies. All of these
teams must negotiate to perform their tasks. Negotiation is social interaction in which two or
more parties in conflict seek a better outcome through joint action than they could realize by
acting independently (Lax & Sebenius, 1986). Conflict occurs within teams because the diverse
backgrounds and responsibilities of team members lead them to have different interests and goals
that must be resolved while working interdependently. Thus, negotiation as a way to resolve
conflict is a mixed-motive task where group members must cooperate to reach agreement and
compete to fulfill their personal interests (McGrath, 1984).
In this study we develop and test hypotheses about the effects of group members’ social
motives and the social motive composition of the group on group members’ negotiation
behavior. Social motives reflect the relative importance people place on own versus joint
outcomes in social interaction. People who are primarily concerned about their own outcomes are
individualists; those who are concerned with both their own and the other parties’ outcomes are
cooperators (Messick & McClintock, 1968). Both of these motives occur frequently in
populations of decision makers and both include a high concern for one’s own outcomes, a
common state for negotiators (Pruitt, 1981). A group’s social motive composition is one aspect
of its diversity. Groups can be composed of all-cooperative members, all individualistic
members, or some combination of cooperative and individualistic members. In this research, we
study all possible combinations of 4-person groups with cooperatively- and individualistically-
Social Motives in Negotiation 4
Social motives are important in negotiations because they affect negotiators’ strategies
(De Dreu, Weingart, & Kwon, 2000; Deutsch, 1973). A negotiation strategy is a set of behaviors
that serves a common function in negotiation. Prior research identifies two primary negotiation
strategies: distributive strategy consisting of behaviors that serve to divide resources among the
parties (e.g., persuasion and value claiming), and integrative strategy consisting of behaviors that
serve to maximize the amount of resources available (e.g., information exchange and formulating
mutually beneficial tradeoffs) (Walton & McKersie, 1965; Pruitt, 1983). As we discuss below,
research on negotiating dyads has consistently shown that cooperative negotiators use integrative
strategy more and distributive strategy less than individualistic negotiators (DeDreu et al., 2000).
Our research is important because it uses multiple theoretical perspectives to fill in the
gaps in our knowledge about how negotiators respond to others’ strategies. McGinn and Keros
(2002) suggest that negotiators jointly improvise to develop a coherent, mutually agreed upon
approach during negotiations; our research examines the process through which negotiators with
conflicting social motives actually adjust their behavior during negotiation. Multiple theoretical
perspectives suggest how negotiators might respond to the social motive composition of the
group. We rely on two related but distinct literatures, negotiation and experimental games, to do
this. The negotiation literature has investigated the effects of social motives on strategic behavior
in dyads, but has done little to address how the dynamic might change when multiple parties are
involved, much less multiple parties with contrasting social motives. The experimental games
literature is useful because it has used prisoners and social dilemmas to examine dyads and
groups with contrasting social motives. We use this theory to more fully develop hypotheses
about how individual negotiators might behave in varying contexts, a question that negotiation
theory has not addressed to the same level of detail.
Social Motives in Negotiation 5
We employ both of these theories to make predictions in domains that have received little
empirical attention. The majority of negotiation strategy research has focused on two-party
interactions (for exceptions see Beersma & DeDreu, 1999, 2002; Weingart, Bennett, & Brett,
1993). Strategic behavior in multi-party or group negotiations may differ from two-party
negotiations just as “findings at one level of analysis do not generalize neatly and exactly to other
levels of analysis, except under very restrictive circumstances” (Firebaugh, 1979, as cited in
Klein & Kozlowski, 2000, p. 213). In dyads, negotiators try to influence one another as they
interact (Ben-Yoav & Pruitt, 1984a; Ben-Yoav & Pruitt, 1984b; Pruitt, Carnevale, Ben-Yoav,
Nochajski, & Van Slyk, 1983). In groups, negotiators also try to influence each other, but instead
of deciding to respond or not respond to influence, in groups, negotiators have a third option:
they also can choose whose influence to respond to. Thus, negotiators in groups are influenced by
a multitude of stimuli, making their behavior more difficult to predict.
Our research also considers how the mix of social motives in groups affects negotiators’
behaviors. Although there has been some research on mixed social motives and negotiation
strategy in dyads (Olekalns & Smith, 1999, 2003b; Schei & Rognes, 2001) and in groups (Ten
Velden, Beersma, & De Dreu, 2004), the emphasis of this research has been negotiations in
which members had the same social motive. In addition, as we discuss below, the findings from
studies examining mixed social-motive groups have been equivocal and theoretical guidance
would be helpful. Thus, one of our goals is to extend theory regarding strategic behavior in
mixed-orientation negotiating groups.
Finally, our research contributes to the literature on mixed social motives in negotiation
via our research methodology. First, we employ a multi-level analysis to examine the joint
impact of negotiator social motive and social motive composition of the group on actual
Social Motives in Negotiation 6
negotiator behavior. This analysis allows us to determine the differential impact of individual-
and group-level phenomena. Second, we employ a sequential analysis of negotiator behavior to
determine how negotiators respond to one another given their social motive and the composition
of the group. We are the first to employ these methods in the study of mixed social motives in
We are able to conduct both analyses because we directly examined negotiation behavior
via videotaping, transcribing, and coding of the group interactions. Most previous research on
mixed social motives in negotiating groups has relied on self-reports of behavior (Schei, 2004;
Schei & Rognes, 2003; 2005; Ten Velden, Beersma, & De Dreu, 2004). Direct examination
allowed us to explore the dynamics of how social motives influence negotiation behavior. Our
research therefore has the potential to break new theoretical and empirical ground by developing
an in-depth view of how group composition affects group members’ negotiation strategies.
Social Motive Composition and Negotiation Strategy
Our research is directed by three general research questions. First, at the group level of
analysis we are interested in how a group’s social motive composition influences the overall
strategic approach taken by the group. Second, crossing levels of analysis, we are interested in
how the social motive composition of the group influences the strategic approach taken by
cooperative and individualistic group members. Drawing from the negotiation and experimental
games literatures, we generate hypotheses to investigate these two research questions. Finally, we
explore the processes through which group members respond to the strategic behavior of others
as a function of their own social motives.
A negotiation strategy is a set of behaviors that serves a common function in negotiation.
Prior research has identified two primary functions in negotiations, creating joint value and
Social Motives in Negotiation 7
claiming individual value (Lax & Sebanius, 1986). Creating joint value increases the total
amount of resources available to the negotiators (i.e., increases the size of the pie) while claiming
individual value divides those resources among the parties. Behaviors that lead to the creation of
joint value typically focus on improving one’s understanding of the other parties’ needs and
desires and using that information to craft mutually beneficial agreements. These behaviors are
called integrative and include seeking information about the other parties’ preferences and
priorities among issues and options, sharing information about one’s own preferences and
priorities, and in some way integrating this information into a mutually acceptable agreement.
Hence the label, “integrative.” Behaviors that lead to claiming high amounts of individual value
are those that influence the other party (or parties) to make concessions (or distribute resources)
in one’s own favor. Hence the label, “distributive.” Negotiators intent on value claiming rely on
persuasion, argument, and even threats. Empirical research generally confirms the theoretical
distinction between integrative and distributive strategy (Olekalns & Smith, 2000, 2003a, 2003b;
Pruitt & Lewis, 1975; Putnam & Wilson, 1989; Thompson, 1991; Weingart, Thompson,
Bazerman, & Carroll, 1990; Weingart, Hyder, & Prietula, 1996) with the caveats that some
behaviors may serve both integrative and distributive functions depending upon how they are
employed (Putnam & Wilson, 1989; Weingart et al., 1990), and that the same negotiators may
use both integrative and distributive strategies in the same negotiation (e.g., Putnam, 1990).
Social motives reflect peoples’ goals for profit maximization, and negotiation research
has consistently shown that social motives drive strategic behavior. Social motives can be
influenced by situational cues (e.g., rewards, instructions, expectations of future cooperative
interaction, see DeDreu et al., 2000), disposition toward cooperation versus competition (e.g.,
social value orientation, see McClintock & Liebrand, 1988), or culture (e.g., individualism vs.
Social Motives in Negotiation 8
collectivism, see Cai, Wilson, & Drake, 2000). However, regardless of their source, their effects
on strategic behavior are largely consistent (De Dreu et al., 2000). Given that cooperatively-
motivated negotiators work toward a goal of maximizing joint gain and individualistically-
motivated negotiators work toward a goal of maximizing own gain, it is not surprising that
cooperative dyads tend to rely more on integrative strategies whereas individualistic dyads tend
to rely more on distributive strategies (De Dreu et al., 2000). However, little is known about what
happens when negotiators with different motives interact with one another.
Six studies have examined mixed-orientation negotiations – three within dyads and three
within groups. Results are inconsistent. One study suggests that mixed dyads are more focused
on value claiming than individualistic dyads, finding that mixed-orientation dyads used fewer
integrative behaviors (restructuring and priority information statements) and more distributive
behaviors (positional statements) (Olekalns & Smith, 1999). Another study suggests that mixed
dyads are more focused on value creating. Olekalns & Smith (2003b) found that mixed dyads
exchanged priority information (an integrative behavior) more frequently and positional
information (a distributive behavior) less frequently than expected by chance. In a third study
focusing on outcomes, Schei and Rognes (2003) examined the role of information asymmetry in
mixed orientation dyads, finding that mixed dyads performed as well as cooperative dyads when
the cooperators were informed that their partners were individualists (but the individualists were
unaware). In contrast, mixed dyads performed as poorly as individualistic dyads when the
individualists knew their partners were cooperators (but the cooperators were unaware). In this
situation, individualists outperformed their cooperative partner, showing that when given the
information advantage, individualists exploited their cooperative partners. However, this study
did not directly examine negotiator behavior, making it unclear whether individualists or the
Social Motives in Negotiation 9
cooperators (or both) changed their behavior in response to the other party.
More recent research has begun to examine mixed-orientation groups. One study of 3-
person groups, in which members knew the orientations of others in their group, found that
cooperators reported using distributive strategy less (less dominating and more yielding) when
negotiating in groups when cooperators were in the majority (2 or 3 cooperators) than in groups
when individualists were in a majority (2 individualists and 1 cooperator) (Schei, 2004). These
results suggest that cooperators are sensitive to the orientations of others, in that cooperative
negotiators reported changing their strategic behavior when there were fewer individualists in the
group. However, this study does not reveal how others in the group or the group as a whole
negotiated as a function of social motive composition. In addition, this study relied on self-report
of behavior, which can be useful in capturing general strategic approaches (as was done in this
study), but is less useful for capturing specific tactical behaviors and patterns of behavior (as we
do in our study). A similar study of negotiated outcomes of 3-person groups (analyzed at the
individual and group level, but not using multi-level modeling) found no differences in joint
outcomes associated with social motive composition, found that individualists obtained a larger
share of the outcome than cooperators, but did not examine negotiator strategic behavior (Schei
& Rognes, 2005). Finally, a third study of 3 person groups, also analyzed at the group level,
examined the social motive composition of majority coalitions, where coalition formation (i.e., a
majority of 2 members versus a minority of 1 member) was based on shared preferences for
outcomes (Ten Velden, Beersma, & De Dreu, 2004). Results showed that the cooperative or
individualistic social motive of the majority (outcome preference) coalition influenced the
strategic approach taken by the group. Mixed-orientation majority coalitions (one cooperator/one
individualist) used a greater mix of strategies than either individualistic or cooperative majorities.
Social Motives in Negotiation 10
However, we would argue that the inclusion of a stable coalition in the Ten Velden et al. (2004)
study transforms the dynamic from a multi-party negotiation to a dyadic negotiation, as two
sides, rather than three, existed as a result of the aligned preferences of two parties. It is also
unclear how the third party’s social motives impacted the groups’ overall negotiation strategy.
Thus, it is unclear how these dynamics might generalize to group settings without a priori
coalitions based on preferences (as we study here).
It is very difficult to draw conclusions about mixed orientation group negotiations from
these studies. First, the results are not consistent across studies. Mixed-orientation groups are
sometimes more competitive than individualistic groups, sometimes similar to cooperative
groups, and other times somewhere in between cooperative and individualistic groups in their
strategic behavior. Second, some of these studies were analyzed at the individual level and others
at the group level, and none included a multi-level analysis. In any given study, it’s difficult to
identify the primary source of variance – the group composition or the behaviors of individuals
as a function of their group membership. The current study attempts to overcome these
limitations by examining group dynamics using a multi-level analytic approach.
Extending the basic idea that social motives of individuals drive their behavior in social
settings, we infer that the group’s dominant social motive will influence its members’ use of
negotiation strategy. For example, the more cooperators in the group, the more integrative
behaviors should be exhibited in the group; the more individualists in the group, the more
distributive behaviors should be exhibited. There are several reasons why this should be so. First,
individuals’ social motives affect individuals’ behaviors (De Dreu et al., 2000). Thus, a
cooperatively-motivated group member is likely to initiate and reciprocate integrative behaviors.
Therefore, the more cooperative members in a group, the more likely a member will initiate
Social Motives in Negotiation 11
integrative behavior. Once integrative behavior is initiated, the more cooperative members in the
group the more likely that integrative behavior will be reciprocated with integrative behavior,
reciprocity being a general phenomenon in negotiation (Brett, Shapiro, & Lytle, 1998; Putnam &
Jones, 1982; Weingart, Prietula, Hyder, & Genovese, 1999). The same rationale explains why the
more individualistic members in a group, the more likely the group’s strategy will be dominated
by distributive behaviors. An individualistically-motivated group member is likely to initiate
distributive behavior, and the more individualists in the group, the more likely this initiative and
the more likely the initiative will be reciprocated.
Hypothesis 1: The more cooperatively-oriented (as opposed to individualistically-
oriented) members of a group, the more likely the group will use integrative strategies.
Hypothesis 2: The more cooperatively-oriented (as opposed to individualistically-
oriented) members of a group, the less likely the group will use distributive strategies.
Experimental game research. Hypotheses 1 and 2 propose that as group composition
becomes more homogenous either in the direction of cooperative social motives or individualistic
ones, the relative frequency with which group members use integrative and distributive strategies
will change in a linear fashion. These hypotheses treat both cooperative and individualistic social
motives similarly. However, the experimental games research suggests that perhaps they should
not be. For example, in groups composed of both individualistic and cooperative negotiators,
cooperative negotiators may shift their approach and reciprocate distributive initiatives as a
defensive move. However, individualists may not shift and reciprocate integrative initiatives, but
instead use the information being shared by cooperative group members to further their own self-
interest goal. We draw on experimental game theory to develop hypotheses about when these
changes will occur depending on the individual’s social motive and the social motive
Social Motives in Negotiation 12
composition of the group.
Experimental game research uses two-player games (such as the Prisoner's Dilemma,
Chicken, and Trust) and multi-player social dilemmas to study decisions to maximize joint
versus individual gain. Typically, these decisions are framed as a choice to cooperate (maximize
joint gain) or defect [either to maximize own gain (greed) or to defend oneself (fear)] however,
each party’s actual outcome is dependent on the choice made by all parties. The game's point
structure determines the type of interdependence between the parties. Participants typically play
repeated rounds of the game, obtaining feedback about the other parties’ behavior after each
round (see Komorita & Parks, 1995 for a review). Results generally suggest that cooperatively-
motivated players make more cooperative choices than do individualistically- or competitively-
oriented players (i.e., those motivated to maximize the difference between own outcome and the
other party) (e.g., Kelly & Stahelski, 1970b; Kramer, McClintock, & Messick, 1986; Kuhlman &
Marshello, 1975), however, the effects may be influenced by the structure of the game (Komorita
& Parks, 1995; Kuhlman & Wimberley, 1976; Parks, 1994).
Although there are some similarities between negotiation tasks and experimental games,
generalization between the two literatures requires caution. The similarities include:
interdependence regarding outcomes (in negotiations both parties have to agree), cooperative
versus non-cooperative choices (in negotiations integrative versus distributive behaviors), and
opportunity to reciprocate behavior. However, negotiation tasks involve direct interaction
between the parties, a wider array of behaviors to choose among to implement a strategic
perspective, multiple issues, contextual information, and incomplete information about the other
party's payoffs. Experimental games simplify and/or hold constant these complicating factors. In
addition, whereas there is both self-interest and collective-interest in experimental games and
Social Motives in Negotiation 13
negotiation, the nature of the interdependence differs. Unlike social dilemmas where cooperating
in service of the collective interest is largely against one’s self interest, in multi-party (and
dyadic) multi-issue negotiations, increasing collective gain via expanding available resources
increases the potential for individual gain.
Research using experimental games (especially two-player prisoner dilemma games)
suggests that a strategy driven by a cooperative motive is more fragile or changeable than one
driven by a competitive motive (to maximize one’s own gains relative to others’ gains) (Kelley &
Stahelski, 1970a; Kelley & Stahelski, 1970b; McClintock & Liebrand, 1988; Schlenker &
Goldman, 1978). Kelley and Stahelski’s (1970b) “triangle hypothesis” proposed that cooperative
players would be more likely to shift from cooperation to non-cooperation (defection) when
confronted with an opponent making non-cooperative choices, than a competitive player would
be likely to shift from non-cooperation to cooperation when confronted with an opponent making
Kelly and Stahelski (1970b) labeled this pattern of behavior the “triangle hypothesis”
because a graph of own orientation (ranging from cooperative to competitive) versus
expectations of other’s orientations (also ranging from cooperative to competitive) results in a
triangular pattern (see Figure 1). They suggested this asymmetric behavior occurs because
cooperative players are more sensitive than competitive players to the social motives of others
(Kelley & Stahelski, 1970b; Van Lange, 1992). Cooperative players encourage cooperation by
signaling their cooperative motive with an early cooperative move, and, if the move is
reciprocated, they experience prolonged cooperative behavior. Thus, they have the potential to
recognize the existence of cooperative others. If their cooperative signal is not reciprocated,
cooperators are quick to shift their strategy to avoid exploitation. Thus, cooperative players are
Social Motives in Negotiation 14
facile at shifting their strategy depending on the actions of the other player (Kelley & Stahelski,
In contrast, competitive players are less sensitive to the motives of others because they
seldom face prolonged cooperative behavior and thus assume that all players are competitively
motivated, like themselves. Since there is no opportunity to exploit a non-cooperative strategy,
and there even may be a strategic advantage to pursuing a non-cooperative strategy, competitive
players are unlikely to shift to cooperation even when faced with a cooperative opponent.
Competitive players maintain their non-cooperative strategy regardless of the strategy of their
opponent (Kelley & Stahelski, 1970b; McClintock & Liebrand, 1988; Van Lange, 1992).
We expect the triangle hypothesis to extend from situations involving competitively-
motivated negotiators to those involving individualistically-motivated negotiators. Although the
two social motives are theoretically distinct, early research on the triangle hypothesis confounded
them. For example, Kelley & Stahelski (1970a, 1970b) focused on competitive motives in their
theorizing, however they induced it using instructions that evoked both competitive and
individualistic motives, “I will work for myself, against the other player, and will be concerned
only with my own score” (1970b, p. 68) and “I will work independently, for myself, against the
other player and the experimenter. I will be concerned only with my own chips” (1970a, p. 384).
Subsequent research has differentiated between individualistic and competitive motives and their
effects on behavioral choice. Most notably, Kuhlman and Wimberley (1976) demonstrated that
the structure of the conflict situation influences whether social motives will result in behavioral
differences. Specifically, they found that individualists behaved similarly to competitors when
resources were limited; such as in prisoner dilemma games. Limited resources made the moves
of individualists (maximize own gain) and competitors (maximize relative gain) largely
Social Motives in Negotiation 15
indistinguishable (Kuhlman & Wimberley, 1976). Since negotiations, by definition, involve
situations of limited resources (hence the need to negotiate a resolution), and in our context the
possibility of increasing joint gain though cooperation, we expect the triangle hypothesis to
extend from competitors to individualists in negotiations in much the same way as it does in
Extending experimental game theory to group negotiations involving individualists and
cooperators, we propose that negotiators’ strategies will depend not only on their own social
motive, but on the social motive composition of their group. We expect that cooperators in
groups with other cooperators will use integrative strategy, but in groups where there are some
individualists, the cooperators will reduce their reliance on integrative strategy and increase their
use of distributive strategy. Our reasoning is based on the prior game theoretic research that
identifies cooperators as sensitive to the social motives of others. Prior research examining the
effect of social motives on concession making and on deception also supports our hypothesis. De
Dreu & Van Lange (1995) found that when negotiating, cooperators responded to concessions
with concessions of greater magnitude than did individualists or competitors. Cooperators
playing an information provision game or in another study, negotiating, were more likely than
individualistic or competitive negotiators to engage in distributive strategy (both withholding
information and providing false information to the other party) when they knew the other party
was individualistically/competitively oriented rather than cooperatively-oriented (Steinel & De
Dreu, 2004).We propose therefore that cooperators will be sensitive to the social motivational
composition of the group and adjust their negotiation strategy to avoid potential exploitation. In
short, we expect that the distributive behaviors of one or more individualists in a group will put
cooperators in a defensive mode, with the result being that the cooperators will use distributive
Social Motives in Negotiation 16
strategy more than when they are in groups with fewer individualists.
The triangle hypothesis allows us to predict that individualists will not alter their behavior
when faced with a cooperator. However, the empirical support for this part of the hypothesis in
game theoretic research is mixed. Individualistic players although less likely to change when
confronted with cooperators, have been shown to be sensitive to the social motives of their
opponents in expanded prisoner dilemma games (including a withdraw option) (Miller &
Holmes, 1975), in a variety of more or less competitive prisoner dilemma games (Schlenker &
Goldman, 1978; van Lange & Visser, 1999), and in n-player social dilemmas with feedback but
without communication (Liebrand, Wilke, Vogel, & Wolters, 1986). Similar to research on social
dilemmas that demonstrates how communication between players increases the occurrence of
cooperation (e.g., Dawes, McTavish, & Shaklee, 1977), negotiations that allow full
communication among parties provide greater opportunity for individualists to ascertain and
respond to the presence of cooperators. Thus, it is possible that in the group context,
individualistic negotiators may use more integrative strategy in groups with more cooperators
than in groups with fewer cooperators. However, we do not expect that individualistic
negotiators will reduce their reliance on distributive strategy in the presence of cooperative group
This reasoning leads to the following hypotheses:
Cooperators are more likely than individualists to change their strategy in response to the
social motive composition of the group:
H3: As the number of cooperators in the group increases (relative to the
number of individualists), cooperators will increase their use of integrative
strategy to a greater extent than will individualists.
Social Motives in Negotiation 17
H4: As the number of individualists in the group increases (relative to the
number of cooperators), cooperators will increase their use of distributive
strategy to a greater extent than will individualists.
Social Motives and the Structuring of Negotiation
These competing sets of hypotheses beg the question of how cooperators actually react to
the behavior of others in a negotiation. Patterns of strategic behavior (actions and reactions)
provide structure to the negotiation, and the way negotiators structure their interaction in relation
to one another influences outcomes (Donohue, 1981; Putnam & Jones, 1982). A simplistic
generalization from the experimental game literature regarding the structuring of negotiation is
that cooperators, being more sensitive to the behavior of others, will be more likely to reciprocate
both integrative and distributive strategies than will individualists. Alternatively, one might
propose, based on the negotiation literature, that cooperators will be more likely to reciprocate
integrative behavior and individualists more likely to reciprocate distributive behavior, in line
with their respective social motives (Olekalns & Smith, 2003b).
However, focusing solely on reciprocity does not acknowledge that negotiators can also
provide structure to the negotiation by shifting from a distributive strategy to an integrative
strategy (i.e., exhibiting transformational sequences) or by complementing prior integrative (or
distributive) behavior by responding in a strategically similar, but not identical way (Brett,
Weingart, & Olekalns, 2004). By structure, we are referring to the predictability of sequential
strategic behaviors. In a multi-party negotiation such as the one we are studying, when Negotiator
A’s use of value creation consistently elicits the same response from another negotiator within
the group, the strategic pattern provides structure to the negotiation. We expect that cooperators,
because of their sensitivity to others’ social motives, are more likely than individualists to display
Social Motives in Negotiation 18
structured communication patterns. We propose that there will be more identifiable regularities
in cooperators’ than individualists’ communication behavior, as defined by their responses to
others’ strategic initiatives. We frame this as a general research proposition, rather than a formal
hypothesis, providing us latitude to explore how this structure might occur.
R1: Cooperative negotiators will provide more structure to the negotiation than their
individualistic counterparts as shown by the patterns of regularity in their responses to others.
We developed the task, Towers Market, for the purpose of studying multi-party, multi-
issue negotiations in which there is no a priori majority and in which there is no coalition that
remains together across issues (Beggs, Brett, & Weingart, 1989; Weingart et al., 1993). The task
involves representatives of four stores, a grocery, a bakery, a liquor store, and a florist and
challenges them to reach agreement on 5 issues concerning whether or not they should open a
market together. The issues include: ambient temperature in the building, procedures for hiring
and training of clerks, advertising style and cost distribution, distribution of maintenance costs,
and location of the stores relative to the main entrance. Appendix 1 shows the structure of
outcomes by role, issue, and option. Note that there were many possible sub-optimal agreements
and no stable coalitions across issues. Students find the exercise engaging and it can be used
outside the research context to teach principles of multi-party, multi-issue negotiation. It is
available in the set of DRRC teaching materials
We gave participants a narrative describing the situation, the issues to be discussed, the
options available for each issue, and their employer’s preferences on each issue. We also gave
Social Motives in Negotiation 19
them the part of the table in the Appendix that quantified their own preferences. Their table
allowed negotiators to evaluate potential agreements in terms of personal benefit.
The participants in our study were 144 management students from two universities who
worked in groups of four (36 groups) doing a group-decision making exercise for their required
MBA-level organizational behavior course. Students were required to participate in the exercise,
but could opt out of the research. They were 75 percent male and 39 percent international
students and had a minimum of two years work experience. Participants were in their first
semester of their MBA program, and neither school had an extensive orientation program at the
time, thus they were unlikely to know each other well.
To ensure that each participant had a dominant social motive (either individualistic or
cooperative), we measured participants’ social motives (i.e., social value orientation) several
weeks before the class exercise. Participants’ confidential role materials also reinforced their
social motive with instructions from the company’s president to maximize either their own
outcomes (if they were individualists) or their own and the group’s outcomes (if they were
cooperators). Therefore the manipulation in this study provided a dispositionally-consistent goal,
but did not include instructions regarding how to achieve the goal, diminishing the possibility of
demand effects regarding the use of specific tactical behavior.
We assigned participants to groups using a blocked random procedure. We composed 4-
person groups to form five experimental conditions: (1) 4 cooperators [4C] (6 groups), (2) 3
cooperators, 1 individualist [3C] (7 groups), (3) 2 cooperators, 2 individualists [2C] (7 groups),
(4) 1 cooperator, 3 individualists [1C] (7 groups), and (5) 4 individualists [0C] (9 groups). We
Social Motives in Negotiation 20
assigned participants to groups selecting cooperative or individualistic students at random from
the two pools of students, according to the composition requirements of the group.
At the experimental session, we gave participants general and confidential role
instructions and time to read both. The role instructions said that the points in their preference
table represented the profits that their store would earn if the group agreed on that alternative for
that issue. We told participants that their preference table was confidential and that they were not
to discuss issues in terms of points. We told groups they had one hour to try to reach an
agreement among all 4 members on all 5 issues, but that they were not required to reach
We used De Dreu and Van Lange’s (1995) measure of social value orientation that
presents nine decomposed games each with a cooperative, an individualistic, and a competitive
option (Messick & McClintock, 1968). This measure has been shown to have good internal
consistency (e.g., Liebrand & Van Run, 1985) and test-retest reliability (Kuhlman, Camac, &
Cunha, 1986). We used participants who made 6 of 9 consistent cooperative or individualistic
choices and assigned those who made competitive choices and those without consistent social
value orientations to non-research groups. From our sample, approximately 50% of participants
were categorized as individualists, 26% as cooperators, 5% as competitives, and 19% were
The instructions from their company’s president used to reinforce individualistic
participants’ social motive stated:
The agreement you reach today will have a major impact on the profitability of
Social Motives in Negotiation 21
[Jacqui’s Bakery]. Points are indicators of profitability. The more points the
group’s decision provides for [Jacqui’s Bakery], the higher [Jacqui’s Bakery]
profitability will be. You should only be concerned with how well [Jacqui’s
Bakery] is doing.
In today’s meeting you should act purely out of self-interest.
Your primary objective should be to maximize the points for [Jacqui’s
Bakery]. You are to get the best agreement for [Jacqui’s Bakery] as you
The instructions used to reinforce cooperative participants’ social motive stated:
The agreement you reach today will have a major impact on your
profitability and on the profitability of Towers Market as a whole. Points
are indicators of profitability. If the Market is to succeed, all the stores are
going to have to be profitable. The greater the points for each store in the
Market, the more profitable the total Market will be. You should be
concerned with how well the other stores are doing, as well as how
well you are doing for [Jacqui’s Bakery].
In today’s meeting you should not act purely out of self-interest.
Your primary objective should be to maximize the group’s outcome as well as [Jacqui’s
Bakery’s] outcome. You are to get the best agreement for Towers Market as a whole as
Participants were not informed of the social motives of other members of their group.
We checked participants’ understanding of their social motive instructions with four
items on a brief post-negotiation questionnaire. We asked participants to identify their primary
Social Motives in Negotiation 22
objective from four options (maximize own outcome, maximize own and group outcomes,
maximize group outcomes, other), to report how committed they were to their objective (1 = very
uncommitted to 5 = very committed), how difficult it was to achieve their objective (1 = very
easy to 5 very difficult), and how much influence their primary objective had on their behavior
during the negotiation (1 = not at all to 5 = a lot).
Measuring Negotiation Strategy
Data collection. We asked students for their written consent to video tape their group
interaction and use the transcripts of the tapes for research. Names were not associated with
video tapes or transcriptions in any way. Our transcriber identified each speaker by role, the
speaking turn (consecutively numbered), and the content of what was said. We next unitized the
transcripts at the level of the thought unit. A thought unit typically contained at least one subject-
verb-object set and focused on one idea or thought. One coder unitized the transcripts and a
second coder unitized 716 units to check for reliability. The level of unitizing reliability was high
(Geutzkow’s U = .06).
We used a coding scheme based on those previously used by Weingart et al. (1996),
Weingart et al. (1993), and Pruitt and Carnevale (1982). The 32 categories of behavior that we
coded are listed in Table 1. A 33rd category, “miscellaneous”, including unclassifiable statements,
was also included in the coding scheme, but not included in any of the analyses. One person
coded all the transcripts, generating approximately 32,000 coded thought units within
approximately 19,000 speaking turns. A second person coded a subset (approximately 500 units
across different groups). Inter-rater reliability was good (Cohen’s kappa > .80 for all categories).
Periodically the first coder checked for drift by recoding a previously coded transcript. Category
by category kappas were good with 94% of the kappas calculated > .80.
Social Motives in Negotiation 23
Data reduction. We used correspondence analysis (Greenacre, 1993) to determine
whether the 32 individual behavioral codes could be reduced to a smaller number of interpretable
clusters of negotiation strategy. We used a group by behavior frequency table and the ANACOR
clustering procedure in SPSS. Cluster analysis (rather than factor analysis) is the appropriate data
reduction procedure when data are frequencies (Greenacre, 1993). Like other data reduction
procedures, cluster analysis results are plotted in a reduced dimensional space. We chose two
dimensions a priori because our coding scheme was based on the distinctions identified in prior
research between integrative and distributive strategies.
Table 1 shows the results of the correspondence analysis. We interpreted six strategy
clusters based on a visual inspection of the mapping guided by an understanding of the coded
behaviors and the theoretical distinctions between integrative and distributive negotiation
strategies. Four of these clusters reflected integrative or distributive strategy; two other clusters
reflected process management behaviors (see Weingart, Olekalns, & Smith, 2004). Because our
hypotheses concerned integrative and distributive behaviors, we did not use the two process
Integrative and distributive behaviors clustered into information behaviors and action
behaviors related to settlement. Information behaviors included both asking for and providing
information; behaviors acting on that information were attempts to create or claim value –
primarily by proposing various types of offers. Combining the dimensions of strategic orientation
(integrative versus distributive) and strategic function (information versus action) generated four
empirically sound categories of strategic behavior: distributive information, integrative
information, value claiming, and value creating (see Table 1). Pruitt and Lewis (1975) and other
researchers (Weingart et al., 1990) have demonstrated the importance of information exchange as
Social Motives in Negotiation 24
a building block of offers that create value, so the distinction between information and action is
not novel. However, research that focuses on strategy and outcome generally collapses across the
information and action dimension (Weingart et al., 1990). Retaining the information-action
distinction in this study allowed us to probe more deeply into our research questions of how the
social motive composition of a group affects group members’ negotiation strategies.
The behaviors in the integrative information cluster included requests for and provision of
information necessary to understand differential preferences and priorities. Examples can be
found in Table 1. The behaviors in the value creating cluster included attempts to discover joint
gain through multi-issue offers, suggestions for packaging and compromise, and demonstrating
insight into the other parties’ concerns. Behaviors relaying multi-issue offers and recognition of
others’ concerns increase joint gain by aiding in the discovery of combinations of issues that
allow for mutually beneficial tradeoffs (Pruitt, 1981; Weingart et al., 1996; Weingart et al.,
1993). The clustering of these behaviors is highly consistent with the operationalization of
integrative strategy in other negotiation research (see Pruitt, 1981).
The behaviors in the distributive information cluster included asking for and providing
factual information about the situation (as provided in the role instructions), focusing on single
issues at a time, identifying differences in positions, and asking questions about others’ reasons
for their positions (see Table 1). It was a bit surprising to us that questions about others’
substantiation of positions were clustered with other distributive behaviors rather than with the
behaviors in the integrative information cluster. We had expected those questions to provide
additional insight into the other party’s position. However, examination of the behavioral units
coded in this category showed that negotiators were challenging the logic or factual basis of the
information provided and assumptions underlying that information. Such challenges logically
Social Motives in Negotiation 25
serve a more distributive function. Finally, the behaviors in the value claiming cluster included
single issues offers, substantiation of position, threats, power-based arguments, references to
bottom line, and the use of mutuality and creative solutions in an attempt to convince the other
side to accept one’s offer. This cluster is highly evocative of other operationalizations of
distributive strategy (see Pruitt, 1981).
To operationalize the dependent variables we generated a proportion for each strategic
category of behavior: each negotiator’s use of a strategic category of behavior compared to the
total number of strategic behaviors used by that negotiator. Because of the nature of the
distribution of proportions, we then log transformed each proportion.
Data analysis. Table 2 presents the means, standard deviations, and correlations among
our independent and dependent variables based on a sample size of 144. We used hierarchical
linear modeling (HLM) to test Hypotheses 1-4 because our 144 participants were nested into 36
groups. HLM allowed us to test hypotheses about individual behaviors while controlling for
group membership (Bryk & Raudenbush, 1992; Kenny, Kashy & Bolger, 1998).
We began by testing whether there was group level variability in each of our dependent
variables: integrative information, value creation, distributive information, and value claiming.
We used an unconditional means (random intercept) model that contained no predictors, but
included a random intercept. The results of the HLM random intercept models showed
significant group level variability on each dependent variables, justifying testing of our
hypotheses: integrative information: γ (gamma) coefficient = –0.28, t(35) = –5.76, p < .01; value
create, γ coefficient = –0.81, t(35) = –23.21, p < .01; distributive information, γ coefficient =
−0.84, t(35) = –23.04, p < .01; value claim, γ coefficient = –1.07, t(35) = –21.44, p < .01.
We then fit separate multi-level models for each dependent variable. In these models the
Social Motives in Negotiation 26
level 1 predictor variable was the individual’s social motive and the level 2 predictor variable
was the number of cooperators in the group. We also specified a cross-level interaction between
social motive and number of cooperators in the group. In these models evidence for Hypotheses 1
and 2 would be provided by an effect for the Level 2 variable (number of cooperators) and
evidence for Hypotheses 3 and 4 would be provided by a cross-level interaction between
individual social motives (Level 1) and number of cooperators (Level 2).
To investigate the structure of strategic interaction, we used S-PLUS to fit a set of log-
linear models to the four-way contingency table: Group Composition (5) × Strategy1 × Strategy2
× Strategy3. In this contingency table, Strategy1, Strategy2, and Strategy3 were the strategies used
at three consecutive time steps, such that each strategic behavior counted as an initiating strategy
(Strategy1) allowing sequences to overlap. These sequences were then cumulated across
negotiations. This analysis assumes that the sequential dependencies in negotiations can be
modeled as a second-order Markov chain. In a second-order Markov chain, two prior behaviors
are needed to predict the third behavior. In our previous analyses of dyadic negotiations, second-
order Markov models have sufficed to describe all the significant sequential dependencies
present in data of this kind (Olekalns & Smith, 2003b; Weingart et al, 1999). We included all six
categories of strategic behavior (including the two process categories) in this analysis to capture
an accurate base rate of sequences, however we only report the results as they pertain to the four
categories of interest (integrative information, value create, distributive information, value
claim). A more detailed description of our modeling approach can be found in Smith, Olekalns,
and Weingart (2005).
Social Motives in Negotiation 27
The social motive manipulation was effective. When we asked participants in the post
negotiation questionnaire to identify their primary objective, 87.2% of the individualistic and
83.3% of the cooperative participants reported their objectives accurately. Another 9.1% of
cooperators reported their primary objective as maximizing the group’s outcome, χ2(3) = 94.09,
p < .001.
Individualists and cooperators reported similarly high levels of commitment to their
primary objective (Scale 1 to 5, overall M = 4.18, SD = .61, t(142) = -.57, ns). All participants
reported their primary objective had a strong influence on their behavior (overall M = 4.45, SD =
.76), but individualists (M = 4.56, SD = .64) reported the influence was slightly stronger than
cooperators (M = 4.32, SD = .86), t(142) = 1.96, p = .05. Finally, individualists reported their
primary objective was more difficult to achieve (M = 3.40, SD = .69) than cooperators (M = 2.85,
SD = .83), t(142) = 4.34, p < .001.
Across dependent variables the pattern of results provides strong support for the triangle
hypothesis. Hypotheses 1 and 2 based on prior negotiation research and theorizing about the
effects of social motives on strategy were not supported. Hypotheses 3 and 4 based on theory and
research in the games literature at the dyad level did fit the data. What follows is a detailed
presentation of these results by dependent variable.
Integrative information. Hypothesis 1, that the use of integrative information would vary
as a function of group composition was not supported, γ (gamma) coefficient = .05, t(34) = 1.69,
p = .10. There was also no effect of individuals’ social motives, γ coefficient = −.12, t(140) =
Social Motives in Negotiation 28
−1.14, ns. Hypothesis 3 predicting a cross level effect between social motive and group
composition was supported, γ coefficient = .10, t(140) = 1.93, p = .05. This interaction, shown in
Figure 2, supports Hypothesis 3: cooperators used more integrative information as the number of
cooperators in the group increased. Individualists only slightly adjusted their behavior as a
function of group composition. Overall, this model provides a better fit to the data, χ2(3) = 20.70,
p < .000, than a random effects model. Approximately 45% of the between group variation in
value creating strategies can be accounted for by the cross-level model.
Value creation. Hypothesis 1 that the use of value creation strategy would vary as a
function of group composition was not supported, γ coefficient = .005, t(34) = .17, ns. There was
a significant effect for individual social motives, γ coefficient = -0.20, t(140) = -2.06, p < .05,
such that cooperatively-motivated negotiators were more likely to use value creation strategies
than individualistically-motivated negotiators. Hypothesis 3 predicting a cross-level effect
between social motive and group composition was significant, γ coefficient = 0.10, t(140) = 2.34,
p < .05. This interaction, shown in Figure 3, supports Hypothesis 3: cooperators used more value
creation as the number of cooperators in the group increased. Individualists did not change their
behavior as a function of group composition. Overall, this model provides a better fit to the data,
χ2(3) = 8.54, p < .05, than a random effects model. Approximately 37% of the between group
variation in value creation can be accounted for by the cross-level model.
Distributive information. Hypothesis 2 that the use of distributive information would vary
as a function of group composition was not supported, γ coefficient = -.01, t(34) = -.23, ns. There
was also no effect of individuals’ social motives, γ coefficient = .07, t(140) = .73, ns. Hypothesis
4 predicting a cross-level interaction between individual social motives and group composition
on distributive information also was not significant, γ coefficient = −.07, t(140) = −1.60, p = .11.
Social Motives in Negotiation 29
Value claiming. Hypothesis 2 that the use of value claiming would vary as a function of
group composition was not supported, γ coefficient = −.01, t(34) = −.17, ns. There was a
marginally significant effect for social motives, γ coefficient = 0.21, t(140) = 1.90, p = .06.
Cooperatively-motivated negotiators were less likely to use value claiming strategies than
individualistically-motivated negotiators. Hypothesis 4 predicting a cross-level interaction
between social motives and group composition was supported, γ coefficient = −0.14, t(140) =
−2.81, p < .01. This interaction, shown in Figure 4, supports Hypothesis 4: cooperators engaged
in more value claiming as the number of cooperators in the group decreased. Individualists did
not change their behavior as a function of group composition. Overall, this model provides a
better fit to the data, χ2(3) = 15.65, p < .005, than a random effects model. Approximately 37% of
the between group variation in value creating strategies can be accounted for by the cross-level
In sum, the data provided substantial support for the hypotheses (3 and 4) derived from
the triangle hypothesis and no support for the hypotheses (1 and 2) derived from negotiation and
social motive theory.
Exploratory Analysis of Negotiation Structure
We first tested whether the cooperators and individualists responded differently to the
strategies of others depending on the social motive composition of the group. To do this, we
compared the cooperators and individualists to one another within the mixed-orientation
conditions (1C, 2C, 3C). We could not include groups from the 0C and 4C conditions in this
analysis because there were no cooperators or individualists (respectively) in those conditions.
We fit a set of log-linear models to the five-way contingency table: Group Composition (3) ×
Social Motive (2) × Strategy1 × Strategy2 × Strategy3. Results from a comparison test of nested
Social Motives in Negotiation 30
log-linear models showed differences between cooperators and individualists in sequential
strategy responses, ∆G2(50) = 56.86, p < .05.1 However, the best fitting model included group
composition and individual negotiator’s social motive, ∆G2(200) = 239.42, p < .01, suggesting
that both individualists and cooperators adapted their strategy responses in response to group
composition, but they did so differently.
To obtain a better understanding of how group composition affected the behavior of
individualists and cooperators, we fit two parallel sets of models (see Table 3). One set of models
examined the behavior of cooperatively-motivated negotiators in groups containing 1, 2, 3, or 4
cooperators (Table 3a); the second set examined the behavior of individualistically-motivated
negotiators in groups containing 1, 2, 3, or 4 individualists (Table 3b). In each table, Model 1
includes controls for frequency effects, first-order sequential dependencies, and second-order
dependencies, and is used as a baseline, independence model.2 In Model 2 we also include the
effects of (i.e., interaction between) group composition (number of cooperators) on frequency of
single strategy use. In Model 3 we add in the effects of group composition on first-order
sequences. Tables 3a and 3b show the overall model fit, G2, as well as the changes to model fit,
∆G2, that resulted from including the interactions. Because of the sparseness of the contingency
table, ∆G2 is a more informative statistic than G2.
Table 3a shows that for cooperative negotiators, group composition affected both strategy
use and strategy sequences. In Table 3a, Model 2 fit improved over Model 1 when we included
an interaction between group composition and use of single strategies, ∆G2(45) = 310.7, p <
.001. In Table 3a, Model 3 fit improved over Model 2 when we added an interaction between
group composition and strategy sequences, Model 3: ∆G2(150) = 197.2, p < .001. This implies
that how cooperative negotiators structured their communication differed depending the social
Social Motives in Negotiation 31
motive composition of the group.
Similarly, as can be seen in Table 3b, group composition also affected both strategy use
and strategy sequences for individualistic negotiators. In Table 3b, Model 2 fit improved over
Model 1 when we included an interaction between group composition and use of single
strategies, ∆G2(45) = 125.01, p < .001. In Table 3b, Model 3 fit improved over Model 2 when we
included an interaction between group composition and use of strategic sequences, ∆G2(150) =
236.41, p < .001. This implies that how individualistic negotiators structured their use of strategy
also differed depending the social motive composition of the group.
To examine where the improvements in model fit occurred, we calculated the
standardized residuals by comparing each model that was associated with a significant
improvement in fit to the immediately preceding model. In this way, we were able to determine
those areas of the contingency table that resulted in improved model fit by identifying the
strategies or strategic sequences that occurred more (or less) frequently than expected by chance.
Specifically, to examine the improvements in fit associated with specific strategic behaviors
[obtained by adding group dependencies in frequency of strategy use (Model 2)] we examined
the standardized residuals in the Group × Strategy1 marginal table for Model 1. Similarly, to
examine the improvements in fit associated with specific sequences [obtained by adding group
dependencies in the sequencing of strategies (Model 3)], we examined the standardized residuals
in the Group × Strategy1 × Strategy2 marginal table for Model 2.3
The standardized residual for any cell j is equal to (nj-mj)/mj1/2, where nj and mj are,
respectively, the observed frequency and fitted (i.e., expected) values for the cell (Agresti, 1990).
Cells with standardized residuals greater than +/- 2.0 (the expected value of the Pearson chi-
square statistic with one degree of freedom when the null hypothesis is true) are important in
Social Motives in Negotiation 32
diagnosing the cause of a model’s failure to fit. Our results show that although some sequences
had large standardized residuals, in absolute terms they occurred relatively infrequently (on
average, less than once per dyad). In subsequent discussion, we concentrate on those sequences
that meet two criteria: (a) the standardized residuals are at least +/- 2.0 and (b) the expected
frequency for the sequence is at least 5 across groups within social motive composition
We proposed that cooperative negotiators would exhibit more regularity (or structure) in
their patterns of use of strategic sequences than would individualistic negotiators. Results support
the proposition. Table 4 shows that cooperators used 9 different sequences of behavior at a level
greater than chance whereas individualists only used 1. Similarly, cooperators used 5 strategic
sequences significantly less than chance, whereas individualists only used 1 at a level less than
chance. At this very general level, the data suggests that cooperators were responding in a more
systematic way than individualists to the strategic behavior exhibited by others in their group.
Table 4 allows us to interpret the use of strategic sequences by cooperators. The major
insight is that cooperators in the all-cooperator groups (4C) structured sequences much
differently than the cooperators in the 1C and 2C conditions. This interpretation is highlighted by
the dashed arrows linking strategic sequences used in 1C and 2C groups and those used in 4C
groups. Three sequences were used by 1C or 2C groups more frequently than chance and by 4C
groups less frequently than chance: infoI ? claim, claim ? infoI, and infoI ? infoI. One
strategic sequence was used in 1C and 2C groups less frequently than chance but more frequently
than chance in 4C groups: infoD ? infoD.
Using theory helps to interpret these differences in patterns of sequences. Whether the
group was using transformational (e.g., infoI ? claim), reciprocal (e.g., infoI ? infoI), or
Social Motives in Negotiation 33
complementary (infoD ? claim) sequences more or less frequently depended on group
composition. In the all-cooperative condition, cooperators were less likely to reorient (via
transformational sequences) the negotiation from integrative to distributive (infoI ? claim
standardized residual = −4.25) and vice versa (claim ? infoI standardized residual = −3.99), they
were more likely to do so in the 1C and 2C conditions (e.g, in the 2C condition infoI ? claim
stand. resid. = 4.55; claim ? infoI = 2.19). In the 1C and 2C conditions cooperators were more
likely to reciprocate integrative information (stand. resid. = 3.28 and 2.03 respectively) and less
likely to reciprocate distributive information (stand. resid. = −3.14 and −3.20 respectively), but
the opposite occurred in the 4C condition where they reciprocated distributive (stand. resid. =
5.02), but not integrative information (stand. resid. = −4.13). Finally, complementary sequences
(different behaviors within strategy type) were only used by cooperators in the 4C condition
(infoD ? claim and claim ? infoD). Together these patterns of strategic sequences show that
cooperators largely reversed their approach when there were all cooperators versus when
cooperators were in the minority or faced an equal number of individualists.
Another way to interpret the ways cooperators structured the negotiation depending on
group composition is to examine sequences that moved toward/reinforced integration versus
distribution. We identified these sequences by considering the strategic orientation of the second
behavior in each sequence – a sequence that ended in a distributive behavior was classified as
distributive and a sequence that ended in an integrative behavior was classified as integrative. We
then counted the number of sequences that resulted in integrative (distributive) behavior at a
level greater than chance and subtracted the number of sequences that resulted in integrative
(distributive) behavior at a level less than chance (e.g., decreased the probability of that behavior)
and calculated a net value for both integrative and distributive sequences within each condition.
Social Motives in Negotiation 34
Results showed that when there was only one cooperator in the group of four, cooperators’
responses moved the negotiation toward integrative behavior (net 2 integrative behavior
responses) and away from distributive behavior (net 1 distributive behavior response). When
there were two cooperators in the group, cooperators both moved the negotiation toward
integration (net 2) and toward distribution (net 1). Interestingly, when the group was composed
of all (4) cooperators, cooperators moved the negotiation toward distribution (net 2) and away
from integration (net 2). These results suggest that cooperators were responding within each
condition by providing a counterpoint to the dominant strategic approach within the group. In
situations where there were few cooperators, they structured the negotiation by moving toward
integrative behavior. In situations where there were no individualists, they structured the
negotiation by moving toward distributive behavior (and away from integrative information
exchange). In groups where there were equal numbers of cooperators and individualists they did
a little of both, providing structure toward integrative and distributive strategy.
Note that there were no significant strategic sequences exhibited by cooperators in 3C
groups. This suggests that their behavior could be adequately characterized by the frequency with
which they used strategies as opposed to the more structured strategic sequences. There were also
few significant strategic sequences identified as in use by individualists as a function of group
composition. As can be seen in Table 4 when individualists dominated cooperators in 1C groups
they used reciprocal sequences of infoI ? infoI less frequently than chance. In groups where
there was but one individualist, (3C), that individualistic negotiator instigated complementary
sequences of distributive strategy (claim ? infoD). This sparse number of strategic sequences
among individualists also indicates that their behavior is for the most part better accounted for by
the frequency of use of strategy rather than by the structuring of sequences of behavior.
Social Motives in Negotiation 35
What happens when diverse groups negotiate? Does one strategic orientation dominate or
do group members converge to a style that reflects the diversity of the group? In answering these
questions, we examined whether and how individual social motive and group composition with
respect to others’ motivations shaped individuals’ negotiation strategies. We found that the
behavior of cooperatively-motivated negotiators was more strongly affected by the composition
of the group than the behavior of individualistically-motivated negotiators. Our results also
suggest that cooperatively-motivated negotiators, much more than individualistically-motivated
negotiators reacted not only by adjusting the total amount of integrative and distributive behavior
they used, but also by responding to others’ behavior in systematic ways, thus providing structure
to the negotiation. Our results contribute to negotiation theory by demonstrating that the pattern
of strategic adjustment to the strategic initiatives of others observed in the experimental gaming
literature also occurs in multi-party negotiations. Our results also identify a pitfall of negotiating
in mixed-motive groups: the presence of individualistically-motivated negotiators triggers a
decrease in use of integrative strategies. However, this decrease in use of integrative strategies
and increase in use of distributive strategies is itself strategic. Just as in games such a prisoner’s
dilemma, cooperators in groups appear to be protecting themselves from individualists’ value
claiming by reducing their reliance on integrative behavior. In addition, cooperators are also
simultaneously using integrative behaviors strategically to redirect the group toward cooperation.
We developed hypotheses at both the group and individual level. At the group level, we
argued for a majority influence effect, hypothesizing that the dominant group strategy would be
determined by the proportion of cooperative versus individualistic group members. Hypotheses 1
Social Motives in Negotiation 36
and 2 were based on research on social motives in negotiating dyads which shows that
individuals’ social motives influences those individuals’ own strategic behavior and our
prediction that individuals’ strategic behavior would be reciprocated by another group member.
However, we also proposed an extension of the triangle hypothesis (Kelley & Stahelski, 1970a;
1970b), that cooperatively-oriented individuals would adapt their strategic behaviors to the
strategies of other members of the group. This extension is reflected in Hypotheses 3 and 4 that
predicted a cross-level or individual-within-group effect on strategic behavior.
Our results provided strongest support for this individual-within-group pattern for three
of the four strategies we examined. Consistent with our hypotheses, the use of integrative
information and value creating strategies by cooperative group members progressively decreased
and the use of value claiming strategies by cooperators progressively increased, the more
individualistic members there were in the group. In addition, we found that the strategic use of
distributive information by cooperators versus individualists across negotiating contexts
exhibited the same pattern as claiming value, but did not reach statistical significance. These
findings have two implications for theory.
The first is our ability to separate the behavior of the group from the behavior of the
individual. In reviewing literature, we noted past research examining mixed-motive groups not
only yielded inconsistent results but also varied in level of analysis (either examining behavior at
the individual or group level). Our analysis enabled us to disentangle the effects of group
composition from those of an individual’s social motive, and also to examine the impact of group
composition on individual behavior. Our results suggest that the group level effects found by
other researchers may be attributable to unilateral strategic adaptation by cooperators rather than
mutual strategic adaptation by both cooperators and individualists.
Social Motives in Negotiation 37
We demonstrated that cooperative negotiators were highly sensitive to the negotiating
context, detecting and responding to the behaviors of individualistic negotiators. This finding
extends the triangle hypothesis, developed by Kelley and Stahelski (1970a, 1970b) in the context
of two-party Prisoner’s Dilemma games, to multi-party negotiation. Importantly, it demonstrates
the same sensitivity to others’ behavior in a context where the intentions and motives of the other
party are unknown and difficult to discern. Unlike PDGs, where participants clearly signal their
strategic intent in their initial moves, negotiators in our study were not aware of other group
members’ social motives and did not receive clear signals at the beginning of the interaction.
Instead, it was the strategic behaviors of the individualists that precipitated the cooperators’
change in strategic behavior. Since strategic behaviors in negotiation are much less transparent
than strategic moves in PDGs, the robust nature of these findings is important. Multi-party
negotiations are very complex decision making situations. Participants must determine the
positions and interests of others in the group, as well as their intentions for cooperation. Our
results suggest that cooperative negotiators are much more sensitive than individualists to the
intentions of others, leading them to cooperate when negotiating with group members who are
cooperators and compete when group members are individualists.
It is interesting to note that the strategic use of distributive information by cooperators
versus individualists across negotiating contexts did exhibit the same pattern as claiming value,
but did not reach statistical significance. It may be that the distributive information exchanged in
this study – factual information about the case – was not as strategically relevant to task
performance as the other three strategies. Factual information exchange allows group members to
develop a shared understanding of the objective situation at hand. Cooperative negotiators might
not have significantly adjusted their reliance on this strategy in response to the composition of
Social Motives in Negotiation 38
the group because they did not perceive it as giving them a competitive advantage in the presence
The frequencies of integrative and distributive behaviors exhibited by individualistic
negotiators indicate that they did not alter their strategic behavior in response to the motivational
context in which they found themselves negotiating. Integrative and distributive information,
value creating and claiming were all used in equal proportions by individualists regardless of
group composition. This suggests that individualistic negotiators adopted an egocentric strategic
approach, while the cooperative negotiators adopted a more carefully constructed variable
approach depending on the motivations of others in the group. One explanation for why
individualists do not adjust their behavior according to context is that individualists do not
perceive the motivational nuances in their context. That is, individualists are so egocentric that
the cooperative actions of others pass unnoticed. Our sequential results support this explanation.
Our sequential analysis supported the results of the HLM analysis showing the distinction
between cooperative and individualistic negotiators in their responsiveness to the strategic
behavior of others. Individualists exhibited significantly less regularity in their responses to the
behavior of others than did cooperators. It appears that individualists were not responding to the
strategic cues of other group members, but rather maintaining their own predetermined strategy.
In contrast, the regularity in cooperators’ responses suggests strategic adaptation. When
cooperators were few in the group, they used more distributive behaviors, but in shifting away
from cooperation in response to potential external threats from individualistic group members,
cooperators did not abandon integrative strategy altogether. Instead they used integrative
strategies to nudge the group back to cooperation. Interestingly, when surrounded by other
cooperators, cooperative negotiators introduced distributive behaviors into the negotiation.
Social Motives in Negotiation 39
Distributive behaviors within an integrative context may have ensured that the group didn’t
compromise, sub-optimize, or just give in. This suggests that although cooperators used more
cooperation when negotiating with all cooperators, they used distributive behaviors strategically
to avoid compromise.
These results add to our understanding of negotiators with cooperative social motives.
Previous research has conceptualized cooperators as being concerned for own and others’
outcomes and for equality in outcomes (Van Lange, 1999). Our results show how cooperators
enact their motive in complex negotiation settings – via a flexible and apparently strategic use of
behaviors. It appears that the focus on others’ interests, as well as their own, cues negotiators to
attend to the other negotiators’ strategic behaviors. It is possible that cooperators’ strategic
flexibility is purely due to self-preservation, but the evidence that cooperators introduce
distributive strategy when in a group with all cooperators cuts against that interpretation. Future
research is needed to explore the motivation behind cooperators’ responsiveness.
Not all group negotiation tasks have the same interdependent structure as the Towers
Market task. Some group negotiation tasks can be resolved with the interests of the majority
prevailing (majority-rule), others with the interests of a single group member prevailing (power-
based). The negotiation scenario in this study required all group members to agree to the final
solution; other types of group decision making might foster different dynamics. Group size might
also influence the dynamics of shifting strategies. In addition to investigating the theoretical
questions generated by this research, future research might examine whether the relative
proportion of cooperative and individualistic negotiators has the same effect on negotiation
strategies in larger groups. Such research might also consider whether negotiators are more likely
Social Motives in Negotiation 40
to form coalitions with other group members who share their motives as opposed to their views,
or how motive-based coalitions affect shifts in the use of integrative and distributive negotiation
strategy. Future research on group negotiation will be needed to determine how strategy unfolds
over time in these group negotiation situations. This study suggests that the dynamics are more
complex than in dyadic negotiations because the strategy-mix depends on the group’s
composition of cooperators but not individualists.
Examination of cognitive and affective responses to social motive diversity in groups will
shed additional light on the development of group negotiation strategies. For example, a more
cognitive approach could explore the question of how group members discover the orientations
of others. Social motive diversity, in contrast to racial or functional diversity, is not externally
visible and is discovered only through social interaction. Given the not-so-subtle effects of social
motive composition on the group’s negotiation process in our study, we would expect even
stronger effects in groups where the diversity is more salient.
Negotiators can increase the likelihood of an outcome that takes the interests of all the
members into account by engaging in integrative strategy (De Dreu et al., 2000). Yet, in the
group context, a negotiator cannot assume that teammates are homogenous in terms of social
motives. This study shows that whether or not a group is composed of all individualistic
members, all cooperative members, or a mix is likely to have a direct impact on cooperators’
negotiation strategy. Cooperative negotiators sense the mix of motivations in the group and
adjust their negotiation strategies accordingly.
Individualists’ lack of flexibility and responsiveness to cooperation does not reveal much
to be optimistic about. However, it might be possible to motivate individualists to be more
Social Motives in Negotiation 41
attentive to others’ strategies using incentives. In experimental settings incentives have been used
to influence negotiators’ motivational orientations (De Dreu et al., 2000). Perhaps the most
obvious advice is to choose teammates for their motivational orientation. If all group members
are cooperatively-motivated, negotiators can engage in value creation without much risk of being
exploited. The mixed motive situation appears to be considerably more treacherous.
Individualistically-motivated members act strategically to promote their own interests, put
cooperatively-motivated members on the defensive, and keep the group focused on value
claiming. Clearly the question that this research raises and does not answer is how to encourage
individualistically-motivated group members in mixed motive groups to pay attention to the
strategic behavior of others, and to engage in integrative behaviors when others introduce those
strategies. Boosting the use of integrative strategy by individualistic negotiators may provide a
means for stimulating value creation by the group.
Based on a global sample of over 1000 managers, Brett (in press) reports that cooperative
motives are dominant among business negotiators (53%), followed by individualistic motives
(37%). (In comparison, our sample was skewed more toward individualists, perhaps because of
the U.S. context, the business school setting, or because of the relative youth of our participants.)
The global dominance of cooperative negotiators among managers is good news, suggesting a
strong potential for a responsive, flexible, and strategic use of integrative strategy in global
Social Motives in Negotiation 42
Agresti, A. (1990). Categorical data analysis. New York: Wiley.
Beersma, B., & De Dreu, C. K. W. (1999). Negotiation processes and outcomes in prosocially
and egoistically motivated groups. International Journal of Conflict Management, 10,
Beersma, B., & De Dreu, C. K. W. (2002). Integrative and distributive negotiation in small
groups: Effects of task structure, decision rule, and social motive. Organizational
Behavior & Human Decision Processes, 87, 227-252.
Beggs, R. J., Brett, J. M., & Weingart, L. R. (1989). Teaching Notes for Towers Market.
Evanston, IL: Northwestern University, Dispute Resolution Research Center.
Ben-Yoav, O., & Pruitt, D. G. (1984a). Resistance to yielding and expectation of cooperative
future interaction in negotiation. Journal of Experimental Social Psychology, 20, 323-
Ben-Yoav, O., & Pruitt, D. G. (1984b). Accountability to constituents: A two-edged sword.
Organizational Behavior and Human Performance, 34, 283-295.
Brett, J. M. (in press). Negotiating globally (2nd Ed.). San Francisco: Jossey-Bass.
Brett, J. M., Shapiro, D. L., & Lytle, A. L. (1998). Breaking the bonds of reciprocity in
negotiations. Academy of Management Journal, 41, 410-424.
Brett, J. M., Weingart, L. R., & Olekalns, M. (2004). Baubles, bangles and beads: Modeling the
evolution of negotiating groups over time. In M. A. Neale, E. A. Mannix, & S. Blount
(Eds.), Research in Managing Groups and Teams: Time in Groups (Vol. 6, pp. 39-64).
New York: Elsevier Science.
Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models: Applications and data
Social Motives in Negotiation 43
analysis methods. Thousand Oaks, CA: Sage.
Cai, D. A., Wilson, S. R., & Drake, L. E. (2000). Culture in the context of intercultural
negotiation: Individualism-collectivism and paths to integrative agreements. Human
Communication Research, 26, 591-671.
Dawes, R. M., McTavish, J., & Shaklee, H. (1977). Behavior, communication, and assumptions
about other people’s behavior in a commons dilemma situation. Journal of Personality
and Social Psychology, 35, 1-11.
De Dreu, C. K. W., & Van Lange, P. A. M. (1995). The impact of social value orientation on
negotiator cognition and behavior. Personality and Social Psychology Bulletin, 21, 1178-
De Dreu, C. K. W., Weingart, L. R. & Kwon, S. (2000). Influence of social motives on
integrative negotiations: A meta-analytic review and test of two theories. Journal of
Personality and Social Psychology, 78, 889-905.
Deutsch, M. (1973). The resolution of conflict: Constructive and destructive processes. New
Haven: Yale University Press.
Donohue, W. A. (1981). Development of a model of rule use in negotiation interaction.
Communication Monographs, 48, 106-120.
Greenacre, M. J. (1993). Correspondence analysis in practice. San Diego: Academic Press.
Kelley, H. H., & Stahelski, A. J. (1970a). Errors in perception of intentions in a mixed motive
game. Journal of Experimental Social Psychology, 16, 411-438.
Kelley, H. H., & Stahelski, A. J. (1970b). Social interaction basis of cooperators and competitors
beliefs about others. Journal of Personality and Social Psychology, 16, 66-91.
Kenny, D., Kashy, D., & Bolger, N. (1998). Data analysis in social psychology. In D.T. Gilbert &
Social Motives in Negotiation 44
S.T. Fiske (Eds.), The Handbook of Social Psychology (Vol 2, 4th ed.), New York,
McGraw Hill, pp 233-265.
Klein, K. J. & Kozlowski, S. W. (2000). From micro to meso: Critical steps in conceptualizing
and conducting multilevel research. Organizational Research Methods, 3, 211-236.
Komorita, S. S., & Parks, C. D. (1995). Interpersonal relations: Mixed motive interaction.
Annual Review of Psychology, 46, 183-207.
Kramer, R. M., McClintock, C. G., & Messick, D. M. (1986). Social values and cooperative
response to a simulated resource conservation crisis. Journal of Personality, 54, 576-591.
Kuhlman, D. M., Camac, C., & Cunha, D. A. (1986). Individual differences in social orientation.
In H. Wilke, D. Messick, and C. Rutte, (Eds.), Experimental social dilemmas (pp. 151-
176). New York: Verlag Peter Lang.
Kuhlman, D. M., & Marshello, A. (1975). Individual differences in game motivation as
moderators of preprogrammed strategy effects in Prisoner’s Dilemma. Journal of
Personality and Social Psychology, 32, 922-931.
Kuhlman, D. M., & Wimberley, D. L. (1976). Expectations of choice behavior held by
cooperators, competitors, and individualists across four classes of experimental game.
Journal of Personality and Social Psychology, 34, 69-81.
Lax, D. A. & Sebenius, J. K. (1986). The manager as negotiator: Bargaining for cooperation
and competitive gain. New York: Free Press.
Liebrand, W. B. G. & Van Run, G. (1985). The effects of social motives across two cultures on
behavior in social dilemmas. Journal of Experimental Social Psychology, 21, 86-102.
Liebrand, W. B. G., Wilke, H. A. M., Vogel, R., & Wolters, F. J. M. (1986). Value orientation
and conformity. Journal of Conflict Resolution, 30, 77-97.
Social Motives in Negotiation 45
McClintock, C. G., & Liebrand, W. B. (1988). Role of interdependence structure, individual
value orientation, and another's strategy in social decision making: A transformational
analysis. Journal of Personality and Social Psychology, 55, 396-409.
McGinn, K. L., & Keros, A. (2002). Improvisation and the logic of exchange in embedded
negotiations. Administrative Science Quarterly, 47, 442-473.
McGrath, J. E. (1984). Groups: Interaction and performance. Englewood Cliffs, NJ: Prentice
Messick, D. M., & McClintock, C. G. (1968). Motivational basis of choice in experimental
games. Journal of Experimental Social Psychology, 4, 1-25.
Miller, D. T. & Holmes, J. G. (1975). The role of situational restrictiveness on self-fulfilling
prophecies: A theoretical and empirical extension of Kelley and Stahelski’s Triangle
Hypothesis. Journal of Personality and Social Psychology, 31, 661-673.
Olekalns, M., & Smith, P. L. (1999). Social value orientations and strategy choices in
competitive negotiations. Personality and Social Psychology Bulletin, 25, 657-668.
Olekalns, M., & Smith, P.L. (2000). Negotiating optimal outcomes: The role of strategic
sequences in competitive negotiations. Human Communication Research, 24, 528-560.
Olekalns, M., & Smith, P.L. (2003a). Testing the relationships among negotiators' motivational
orientations, strategy choices, and outcomes. Journal of Experimental Social Psychology.
Olekalns, M, & Smith, P. L. (2003b). Social motives in negotiation: The relationships between
dyad composition, negotiation processes, and outcomes. International Journal of Conflict
Management, 14, 233-254.
Parks, C. D. (1994). The predictive ability of social values in resource dilemmas and public
Social Motives in Negotiation 46
goods games. Personality and Social Psychology Bulletin, 20, 431-438.
Pruitt, D. G. (1981). Negotiation behavior. New York: Academic Press.
Pruitt, D. G. (1983). Achieving integrative agreement. In M. H. Bazerman and R. J. Lewicki
(Eds.), Negotiating in organizations (pp. 35-50). Beverly Hills, CA: Sage.
Pruitt, D. G., & Carnevale, P. J. (1982). The development of integrative agreements. In V. J.
Derlega and J. Grzelak (Eds.), Cooperation and helping behavior: Theories and research.
New York: Academic Press.
Pruitt, D. G., Carnevale, P. J., Ben-Yoav, O., Nochajski, T. H., & Van Slyk, M. R. (1983).
Incentives for cooperation in integrative bargaining. In R. Teitz (Ed.), Aspiration levels in
bargaining and economic decision making (Vol. 213, pp. 22-34). Berlin: Springer-
Pruitt, D. G., & Lewis, S. (1975). Development of integrative solutions in bilateral negotiation.
Journal of Personality and Social Psychology, 31, 621-633.
Putnam, L. L. (1990). Reframing integrative and distributive bargaining: A process perspective.
In B. H. Sheppard, M. H. Bazerman, and R. J. Lewicki (Eds.), Research on negotiation in
organizations (Vol. 2, pp. 3 – 30). Greenwich, CT: JAI.
Putnam, L. L., & Jones, T. S. (1982). Reciprocity in negotiations: An analysis of bargaining
interaction. Communication Monographs, 49, 171 - 191.
Putnam, L. L., & Wilson, S. R. (1989). Argumentation and bargaining strategies as
discriminators of integrative outcomes. In M. A. Rahim (Ed.), Managing Conflict: An
interdisciplinary approach (pp. 121-144). New York: Praeger.
Schei, V. (2004). Cat among the Pigeons? Cooperators’ response to individualists in negotiating
groups. Paper presented at the 17th Annual Conference of the International Association
Social Motives in Negotiation 47
for Conflict Management, Pittsburgh, PA.
Schei, V., & Rognes, J. K. (2001). Can individualists and cooperators play together? Mixed
motivational orientations in negotiations. Working paper, Norwegian School of
Economics and Business Administration, Bergen, Norway.
Schei, V. & Rognes, J. K. (2003). Knowing me, knowing you: Own orientation and information
about the opponent’s orientation in negotiation. International Journal of Conflict
Management, 14, 43-59.
Schei, V., & Rognes, J. K. (2005). Small group negotiation: When members differ in
motivational orientation. Small Group Research, 36, 289-320.
Schlenker, B. R., & Goldman, H. J. (1978). Cooperators and competitors in conflict: A test of the
“Triangle Model.” Journal of Conflict Resolution, 22, 393-410.
Smith, P.L., Olekalns, M., & Weingart, L. (2005). Markov chain analyses of communication
processes in negotiation. International Negotiation, 10, 97-113.
Steinel, W., & De Dreu, C. K. W. (2004). Social motives and strategic misrepresentation in
social decision making. Journal of Personality and Social Psychology, 86, 419-434.
Ten Velden, F. S., Beersma, B., & De Dreu, C. K. W. (2004). Heterogeneous social motives in
negotiating groups: The moderating effects of decision rule and interest position. Paper
presented at the 17th Annual Conference of the International Association for Conflict
Management, Pittsburgh, PA.
Thompson, L. L. (1991). Information exchange in negotiation. Journal of Experimental Social
Psychology, 27, 161-179.
Van Lange, P. A. (1992). Confidence in expectations: A test of the triangle hypothesis. European
Journal of Personality, 6(5), 371-379.
Social Motives in Negotiation 48
Van Lange, P. A. (1999). The pursuit of joint outcomes and equality in outcomes: An integrative
model of social value. Journal of Personality and Social Psychology, 77(2), 337-349.
Van Lange, P. A., & Visser, K. (1999). Locomotion in social dilemmas: How we adapt to
cooperative, Tit-for-tat, and noncooperative partners. Journal of Personality and Social
Psychology, 77, 762-773.
Walton, R. E., & McKersie, R. B. (1965). A behavioral theory of labor negotiation: An analysis
of a social interaction system. New York: McGraw-Hill.
Weingart, L. R., Bennett, R. J., & Brett, J. M. (1993). The impact of consideration of issues and
motivational orientation on group negotiation process and outcome. Journal of Applied
Psychology, 78, 504-517.
Weingart, L. R., Hyder, E. B., & Prietula, M. J. (1996). Knowledge matters: The effect of tactical
descriptions on negotiation behavior and outcome. Journal of Personality and Social
Psychology, 70, 1205-1217.
Weingart, L., Olekalns, M., & Smith, P.L. (2004). Quantitative coding of negotiation processes.
International Negotiation, 9, 441-456.
Weingart, L. R., Prietula, M. J., Hyder, E., & Genovese, C. (1999). Knowledge and the
sequential processes of negotiation: A Markov chain analysis of response-in-kind.
Journal of Experimental Social Psychology, 35, 366-393.
Weingart, L. R., Thompson, L. L., Bazerman, M. H., & Carroll, J. S. (1990). Tactical behavior
and negotiation outcomes. The International Journal of Conflict Management, 1, 7-31.
Social Motives in Negotiation 49
Towers Market Summary of Points
Temperature c 20
Advertising c 10
e 60 100
Clerks b 20
Maintenance c 0
Position b 40
e 0 100
Maximum Total Points: 260 280
Note: Issue options represented by letters a through e were presented to the participants as substantive
options within the context of the scenario.
bakery liquor florist grocery
Social Motives in Negotiation 50
1 G2 is the likelihood ratio, a measure of overall model fit. ∆G2 represents the change to
model fit that results from adding, in this case, effects of group composition on the use of
strategy sequences. Because of the sparseness of the contingency table, ∆G2 is a more
informative statistic than G2.
2 The log-linear models control for frequency effects [s1][s2][s3] (i.e., the relative
frequency with which each strategy [s] was used), first-order sequential dependencies [s1,s2]
[s2,s3] (the probability that a cooperator (individualist) would engage a given strategy, e.g. [s2], in
response to a preceding strategy engaged by another member of the group, e.g. [s1]), and second-
order sequential dependencies [s1,s2,,s3] (the probability that a cooperator (individualist) would
engage a given strategy [s3] in response to two preceding strategies engaged by members of the
3 These comparisons are informative because, as a new term is added to the model, the
marginal totals associated with that term in the contingency table are constrained so that observed
and expected frequencies are the same. For example, when we add the term [nc,s1,s2] to build
Model 3, we constrain the marginal totals associated with that term. However, the same
marginal total associated with same term are unconstrained in Model 2. Consequently, large
deviations from fit in Model 2 help us understand which specific strategy sequences contributed
to improvements in model fit when we added the term [nc,s1,s2] to the model.