url:
http://66.218.69.11/search/cache?ei=UTF-8&p=journal+on+personality+psychology&fr=yfp-t-305&fp_ip=PH&u=www.som.yale.edu/faculty/keith.chen/negot.%2520papers/MorrisLarSu_BargandFuAtEr99.pdf&w=journal+journals+personality+psychology&d=CXkYY0LURju9&icp=1&.intl=us
Journal of Personality and Social Psychology
Copyright 1999 by the American Psychological Association, Inc.
Volume 77(1) July 1999 p 52–67
Misperceiving Negotiation Counterparts: When Situationally Determined
Bargaining Behaviors Are Attributed to Personality Traits
[Interpersonal Relations and Group Processes]
Morris, Michael W.
1,2,5
; Larrick, Richard P.
3,6
; Su, Steven K.
4
1
Graduate School of Business, Stanford University
2
Institute for Personality and Social Research, University of California, Berkeley
3
Graduate School of Business, University of Chicago
4
Organizational Behavior Group, INSEAD.
5
Correspondence concerning this article should be addressed to Michael W. Morris, Graduate School of Business,
Stanford University, Stanford, California 94305-5015. Electronic mail may be sent to morris_michael@gsb.stanford.edu.
6
Correspondence concerning this article should be addressed to Richard P. Larrick, Graduate School of Business,
University of Chicago, 1101 East 58th Street, Chicago, Illinois 60637. Electronic mail may be sent to
rick.larrick@gsbpop.uchicago.edu.
We acknowledge helpful comments at various stages of this research from Max Bazerman, Jonathon Bendor, Oliver
John, Roderick Kramer, Keith Murnighan, Margaret Neale, Richard Nisbett, Richard Robins, Lee Ross, and Amos
Tversky. We gratefully acknowledge research funding from the Stanford University Graduate School of Business, the
University of Chicago Graduate School of Business, and the William S. Fishman fellowship.
Received Date: November 30, 1998; Accepted Date: December 1, 1998
Outline
Abstract
Background
Actual Causes of Negotiation Behavior
Perceived Causes of Negotiation Behavior
Hypothesis Development
Experiment 1: Patterns of Personality Misperception in Negotiations
Method
Participants.
Procedure.
Negotiation measures.
Situation perception measures.
Personality perception measures.
Results and Discussion
Negotiation measures.
Situation perception measures.
Personality perception measures.
Experiment 2: Processes Contributing to Misperceptions
Method
Participants.
Procedure.
Results and Discussion
Negotiation measures.
Page 1 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
Situation perception measures.
Perceptions of the counterpart's personality.
Persistence of personality trait ascriptions.
Future behavior toward the perceived.
General Discussion
The Issue of Context Specificity in Personality Dispositions
Implications for Theory
Social psychological processes in negotiation.
Accuracy in social perception.
Dynamics of self-confirming beliefs.
Implications for Practice
References
Graphics
Figure 1
Table 1
Table 2
Figure 2
Table 3
Table 4
Table 5
Figure 3
Table 6
Table 7
Table 8
Table 9
Table 10
Abstract
Several experiments provided evidence that negotiators make systematic errors in personality-
trait attributions for the bargaining behaviors of their counterparts. Although basic negotiation
behavior is highly determined by bargaining positions, negotiators primarily interpret their
counterpart's behavior in terms of the counterpart's personality, such as his or her level of
cooperativeness or agreeableness. Data support a model of 4 processes that contribute to
misperceptions: (a) the primacy of situations in determining bargaining behavior, (b) the primacy
of personality traits in attributions, (c) the lack of sufficient information about the other's
situation to discount personality attributions, and (d) the potentially self-confirming
consequences of personality attributions for subsequent interactions. The authors discuss
implications for research areas such as social cognition in negotiation, accuracy in social
perception, and the dynamics of belief confirmation.
The psychology of resolving mixed-motive conflicts has long been of interest to social
psychologists, sociologists, and political scientists, not only because explicit negotiations are
ubiquitous and consequential phenomena in work and political life, but also because many other
interpersonal relations can be modeled as negotiations of mixed-motive conflicts (
Axelrod, 1984
;
Page 2 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
Kelley & Thibaut, 1954
;
Strauss, 1964
). In the psychology of conflict resolution, an important role is
played by negotiators' perceptions of their counterparts. Social psychologists have argued that
interpersonal conflicts are particularly likely to give rise to attributions of personality traits to
one's counterpart (
Orvis, Kelley, & Butler, 1976
). Conflict theorists have suggested that a negotiator's
attributions for a counterpart's behavior exert a pivotal influence on the negotiator's strategic
decisions (
Schelling, 1960
). A first purpose of the current research is to investigate regularities in
the personal dispositions perceived in negotiation counterparts as a function of structural features
of the conflict. A second purpose is to investigate the consequences of these attributions on
decisions in subsequent interactions with the counterpart.
If attributions in conflict have far-reaching consequences, it is important to know whether
attributions are accurate—whether the perceived causes of negotiation behavior correspond to
the actual causes. To address this, we reviewed the literature on the actual and perceived causes
of negotiation behavior, and then we developed hypotheses about systematic attributional biases
in which effects of specific negotiation situations are misinterpreted in terms of specific
personality traits. Our methods are influenced by recent social perception researchers (e.g.,
McArthur & Baron, 1983
;
Swann, 1984
) who have mounted challenges to traditional evidence that
attributions are biased and have defended more stringent criteria for identifying attribution
errors. In a first experiment we tested hypotheses about two general patterns of misperception. In
a second experiment we tested hypotheses about four component social psychological processes
that contribute to misperceptions in negotiations.
Background
Actual Causes of Negotiation Behavior
A point where the findings of social psychology diverge from the core beliefs of conventional
Western social thought is the degree to which an individual's behavior is shaped by stable
personal characteristics as opposed to temporary social situations.
Milgram's (1963)
famous
obedience experiment, in which pressure from an authority figure led participants to deliver
painful electric shocks, illustrates the extent to which people's behavior is determined by the
pressures and incentives of a situation. Sometimes even minor factors in a person's situation
shape behavior to a greater extent than the most central factors in the person's character. A
classic illustration is
Darley and Batson's (1973)
study of the determinants of altruistic behavior. They
observed whether seminary students heading across campus to deliver a sermonette on the Good
Samaritan stopped to assist a stranger in medical need. The probability of stopping was not at all
influenced by whether the student had indicated “helping others” was a central religious value.
However, stopping was influenced by whether the student had been made to feel that he or she
was running late to deliver the sermon. In sum, acting as a Good Samaritan was determined more
by the superficial constraints of a person's situation than by the person's deepest values.
Empirical research on the determinants of negotiation behavior has yielded similar
conclusions. Many researchers have modeled negotiations by running experiments in which
participants play bargaining games. Recent reviews of this work have concluded that important
components of bargaining behavior, such as the extent to which a negotiator haggles with his or
her counterpart instead of making concessions, are greatly determined by a player's economic
bargaining situation and little determined by personality traits (
Thompson, 1990
). A second research
paradigm has been direct observation of naturally occurring negotiations. Although researchers
in this tradition have generally not been able to take the measurements required for testing causal
hypotheses, their interpretive conclusions about the causes of negotiators' behavior have
paralleled those from bargaining game experiments in their emphasis on the importance of
Page 3 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
situations and unimportance of personality. From one of the more comprehensive studies of
industrial negotiators inside and outside the bargaining room,
Douglas (1962)
concluded that a
negotiator “more nearly resembles his opposite number in the conference room than himself
outside the conference room” (p. 159).
Perceived Causes of Negotiation Behavior
The fascination of social psychologists with attributions for behavior is owed in part to
discoveries of systematic biases, which seem to play a pivotal role in many social interactions.
Ichheiser noted that we have “in everyday life the tendency to interpret and evaluate the behavior
of other people in terms of specific personality characteristics rather than in terms of specific
social situations in which these people are placed” (
Ichheiser, 1943
, p. 151; emphasis added). This
bias was explained by
Heider (1958)
in terms of a gestalt perceptual process and more recently by
Nisbett and Ross (1980)
in terms of judgmental heuristics; it is referred to as the fundamental
attribution error (
Ross, 1977
) or the tendency to make correspondent inferences from behaviors to
traits (
Jones, 1990
). It is most clearly documented in contexts where the actual causes of behavior
are known. For example,
Safer (1980)
found that observers of a film of
Milgram's (1963)
experiment
attributed the actions of obedient participants to sadistic personalities rather than to the
situational pressure to comply with authority. Likewise,
Pietromonaco and Nisbett (1982)
found that
participants who read about
Darley and Batson's (1973)
seminarians attributed the behavior of the
hurried seminarians to their personalities rather than their situation.
The attribution of negotiation behavior to personality traits rather than situational constraints
(although familiar to those who have been privy to the thoughts of disputants in a contentious
strike or divorce) has been the subject of relatively few experimental studies. Although
researchers have run many experiments using detailed simulations of negotiation procedures to
search for actual effects of personality in negotiations, research on perceptions of opponents has
primarily involved abstract conflict games and perceptions of strategy or intent. Nonetheless,
studies have found that an opponent's situational constraints are often misunderstood.
Dorris,
Gentry, and Kelley (1966)
created a bidding game in which a buyer and a seller were each given a
limit on the price on which they could settle and still make a profit. When sellers had a high
limit, deals were less likely. Interestingly, although participants knew their counterparts were
constrained by an externally imposed limit, they nonetheless attributed their counterparts'
behavior to personal intent: sellers in the high-limit condition were perceived by counterparts to
have a competitive intent. In similar studies of buyer–seller games, participants have
systematically underestimated their counterpart's limit (
Kelley et al., 1970
;
Pruitt & Drews, 1969
). When
participants had the option of attributing a behavior to both personal and situational factors, they
primarily attributed it to personal factors (
Kelley et al., 1967
). These early studies of perceptions of
conflict counterparts did not specify what sort of inferences are drawn. Evidence from other
domains suggested that perceivers draw mistaken inferences about enduring personality traits
(
Nisbett & Ross, 1980
).
Although systematic errors in social perception were widely accepted a decade ago, recently
critics have challenged the evidence. Some challenges focus on the contexts in which errors have
been demonstrated.
McArthur and Baron (1983)
argued that attributions drawn about a target person
on the basis of a vignette description or on abstract information about moves in a game are
responses to impoverished social stimuli; hence, errors may occur that would not occur in the
ecology of real-world social interactions, where responses are based on a richer flow of
information (for a review of related issues, see
Funder, 1995
). Another point is that social
perception accuracy increases in contexts where perceivers are motivated to perceive another
person accurately, such as when the perceiver's and the target person's outcomes are
Page 4 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
interdependent (
Neuberg & Fiske, 1987
;
Ruscher & Fiske, 1990
). Other challenges focus on how
researchers model lay conceptions of personality.
Swann (1984)
pointed out that instructions in
some studies may tap perceptions of a target's disposition within a bounded social situation, and
these perceptions may indeed have circumscribed accuracy. Other researchers have objected to
the contrived trait lists on which perceivers' impressions have been measured. Studies with more
comprehensive measures have identified distinctive types of bias on particular personality
dimensions (
Robins, Spranca, & Mendelsohn, 1996
).
In part because of concerns about the methods used to establish attribution errors, another
tradition of research on social perception in conflicts has eschewed experiments in favor of
retrospective surveys of naturally occurring conflicts. Surveys have revealed general biases in
attribution, such as the difference between the attributions made for behavior by oneself and for
behavior by the other person (i.e., the actor–observer difference;
Jones & Nisbett, 1971
). For
example, members of romantic couples in conflict ascribed negative traits (e.g., selfishness) more
often for behaviors of their partner than for their own behaviors (
Orvis, Kelley, & Butler, 1976
).
Likewise, college students ascribed uncooperativeness more to their roommates (
Sillars, 1980
) and
executives ascribed competitiveness more to their rivals (
Thomas & Pondy, 1977
) than they did to
themselves. It is tempting to integrate these actor–observer differences in natural conflicts with
the findings of
Kelley et al. (1967)
that competitive intent is overattributed to opponents in
bargaining games. However, actor–observer differences can reflect either error in the perceiver's
self-attribution or error in his or her attributions about the counterpart, and it is not clear which of
these underlies the findings. A second complication in integrating actor–observer findings from
field studies of conflict is that some of the identified personality traits are semantically unrelated
to uncooperativeness. For instance, negotiators have been observed to overattribute to opponents
characteristics such as anxiety (
Douglas, 1962
) and insincerity (
Baron, 1984, 1988
). This suggests that
there may be more than one dimension on which negotiators misperceive their counterparts'
personalities.
Hypothesis Development
Our general hypothesis is that certain bargaining behaviors are primarily determined by the
negotiator's situational constraints but are attributed by their opponents to personality traits.
Through a review of literature on the dimensions of bargaining situations, we will work toward
more specific hypotheses about the content of particular misperceptions. A negotiator's
bargaining situation can be described in terms his or her alternative option, the option he or she
will take if no agreement is reached in the negotiation (
Raiffa, 1982
;
White & Neale, 1991
). A
negotiator's alternative is like the ground from which he or she negotiates; it is the basic
situational factor that determines his or her bargaining style and behavior. Negotiation
alternatives vary on two basic dimensions: value and risk.
Value is the amount of a negotiator's alternative. For instance, a job recruit's bargaining
position depends greatly on whether his or her offer from an alternative employer is relatively
low or high in value.
Figure 1
illustrates the bargaining position of a recruit occupying four
different bargaining positions varying in value and risk. The top row of the matrix shows that a
recruit with an alternative offer that is relatively high in value is like a seller with a high
minimum price: In the negotiation with the recruiter, this recruit has a relatively small zone of
possible agreements (i.e., where the recruiter pays no more and the recruit earns no less than with
their respective alternatives). To achieve an agreement appreciably better than his or her
alternative, the recruit will likely have to engage in hard bargaining, or haggling, with the
recruiter—it will not be easy to find a mutually pleasing outcome. By contrast, when the recruit's
Page 5 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
alternative is relatively low and the zone is relatively large (as shown in the bottom row of
Figure 1
), then the recruit will not have to haggle much to reach an acceptable agreement
(
Thompson, 1997
). Although negotiators know that their own haggling depends on the situation, we
suspect that they nonetheless misattribute their counterpart's level of haggling to his or her
personality. Hence, the first error pattern we predict is that negotiators misattribute the trait of
uncooperativeness to counterparts who have a relatively high-value alternative in the negotiation
where they meet.
Figure 1. An illustration of four different bargaining situations created by crossing the value (high vs. low) and risk (certain vs.
risky) of the recruit's alternative. In the right column, the dashed line indicates that the recruit's alternative is an expected value
corresponding to the mean of the market distribution.
A second aspect of a negotiator's alternative option is its level of risk. In some cases, a recruit
may have a definite offer of a specific salary from another employer as his or her alternative. In
other cases, a recruit's alternative may be uncertain or risky—a risky alternative is an expected
value based on knowing that one will draw an offer from a distribution, such as a market (for
further definition see
Larrick & Boles, 1995
). In
Figure 1
, the recruits in the left column are standing
on the secure ground of a definite offer from another company; recruits in the right column are in
the more precarious position of not being certain about the value of their alternative, having only
an expectation based on the distribution from which the offer will be drawn. Having a risky, as
opposed to a certain, alternative affects one's bargaining in several ways. A recruit without a
definite, concrete alternative will have more difficulty setting a consistent lower limit (or
reservation price) in the negotiation, because this limit must take into account not only the
expected value of the alternative but also the value that he or she places on certainty. A recruit
without a firm alternative offer is also more likely to be vague about his or her alternatives
during the negotiation. In sum, a certain alternative option makes it more likely that negotiators
will be consistent and steady, whereas a risky alternative makes it more likely that negotiators
will waffle in the limits they set and in the justifications they offer for these limits. Although we
expect that waffling is induced by the situation of uncertainty, we expect that it will be primarily
interpreted by perceivers in terms of personality traits such as inconsistency or insincerity.
Hence, our second prediction is that negotiators will misattribute dishonesty or insincerity to
Page 6 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
counterparts who have relatively risky alternatives.
In a pilot experiment, a test of these predictions was conducted by having master of business
administration (MBA) students participate in a negotiation exercise that closely simulated a job
compensation negotiation typical of the ones they were currently involved in at the time. One
hundred twenty students were randomly assigned to roles of recruiter and recruit for a scenario in
which they were negotiating the compensation of a possible position for the recruit. The recruit
role varied in the value and certainty of the alternative, resulting in the four conditions illustrated
by
Figure 1
. After negotiating, participants rated themselves and their opponent on an ad hoc list
of traits relevant to negotiation. Recruits in different experimental conditions did not differ in
self-ratings, yet they differed in how they were rated by their opponents. The pattern of
differences fit our predictions. Compared with recruits with a low-value alternative, those with a
high- value alternative were perceived as more stubborn (4.60 vs. 3.73 on a 7-point scale ranging
from 1 = disagree to 7 = agree), F(1, 57) = 3.29, p < .10, and more competitive (5.30 vs. 4.40), F
(1, 57) = 6.21, p < .05. Recruits with a risky, as opposed to a certain, alternative were rated less
sincere (4.85 vs. 5.38), F(1, 57) = 3.0, p < .10. However, differences on other trait dimensions,
such as that recruits were rated lower on the friendly dimension when they had a high-value
alternative (4.89 vs. 5.54), F(1, 57) = 3.69, p < .10, and when they had a risky alternative (4.81
vs. 5.58), F(1, 57) = 4.82, p < .05, suggested that misperceptions involve more general
personality constructs than specific trait terms.
Experiment 1: Patterns of Personality Misperception in Negotiations
To frame hypotheses at the most general level that conveys the content of misperceptions, we
drew on models of the major dimensions of personality differences and personality perceptions
(
Goldberg, 1990
). Four major factors that reemerge in analyses of trait terms in different languages
(
Yik & Bond, 1993
) are usually labeled as follows: (a) Extraversion, (b) Agreeableness, (c)
Conscientiousness, and (d) Emotional Instability or Neuroticism. In numerous studies,
researchers using reliable indexes of these general dimensions rather than ratings of isolated trait
terms have found a considerable degree of consensus in perceivers' impressions of a target
person's personality (
John & Robins, 1994
). Scales measuring these dimensions seem to capture
content-general constructs in lay perceptions of personality (
McCrae & John, 1992
). For instance, a
target person rated high in Agreeableness is expected to be more cooperative than the average
person in the wide range of domains for which this trait is relevant. If biases in negotiators'
personality perceptions affect their core impressions of others, then it should be possible to
specify their biases in terms of basic personality dimensions. We propose that the traits to which
people should attribute haggling and waffling behavior (if they are making correspondent
inferences) map closely onto two of these factors (to the extent that uncooperativeness is an
element of low Agreeableness and to the extent that insincerity and inconsistency are related to
Emotional Instability). Hence, we can frame our predictions in terms of these more general
constructs as follows: (a) Negotiators with high-value alternatives will haggle more and hence
evoke perceptions of low Agreeableness, and (b) negotiators with risky alternatives will waffle
more and hence evoke perceptions of Emotional Instability.
Our goal was to test these predictions with methods that meet the criteria for error in social
perception emphasized in recent scholarship. That is, in our research personality perception is
measured in an ecologically valid context, in which errors have personal consequences for
participants and competition motivates accurate perception. Also, our instructions emphasized to
participants that they were to indicate their inferences about enduring characteristics of the other
person, not merely describe the other's behavioral patterns. Specifically, we asked them to
Page 7 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
indicate characteristics that they would expect to see again in a future interaction with the
other person. Also, they rated this person on a comprehensive set of trait descriptors.
Method
Participants.
Participants were 376 MBA students from four sections of a course on negotiation at Stanford
University (1 year after the pilot study) and from three sections of a course on negotiation at the
University of Chicago. To make the motivation to succeed in one's outcome more salient, one
randomly selected participant in each role was publicly awarded a cash payment in direct
proportion to the value of their settlement (the payment ranged from $0 to $24) in class after the
negotiation simulation.
Procedure.
The data were collected during negotiations in a simulation exercise conducted during the 2nd
week of a course on negotiation. The negotiation exercise was conducted between two students
acting as, respectively, a recruiter from Alpha Inc., a consulting firm, and an MBA recruit. It
concerned only one issue: the salary the firm would offer the recruit (all other terms, such as the
signing bonus, had been settled). It was clearly stated that the two individuals would never be
working together in the future, so the relationship between the two players was not an issue in
the negotiation.
Negotiators in both roles were given private information about their alternatives. On the one
side, recruiters were told that if they did not reach a satisfactory agreement with this recruit then
they would definitely sign an alternative candidate. The alternative candidate had asked only that
Alpha Inc. match the salary offer being sent in the mail by another firm; because this figure was
as yet unknown, the recruiter's alternative was risky. Recruiters were told that their expectation
of what they would have to pay the alternative candidate came from an office survey that found
salaries of recruits from the top five schools in this industry fell on a normal distribution with a
mean of $90,000 and with most offers falling in a range from $75,000 to $105,000.
On the other side, recruits were told that in addition to their offer from Alpha Inc., they had an
offer from another consulting firm, Lambda. If the negotiation with Alpha Inc. led to a
satisfactory salary offer, they would sign immediately and not pursue the offer from Lambda.
However, if the negotiation did not lead to a satisfactory offer, they would accept the offer from
Lambda, which would be a first and final offer (this firm does not negotiate salary). The recruits'
expectation about their salary with Lambda varied across four conditions, which corresponded to
the conditions in
Figure 1
:
1. In the risky, high-value condition, recruits were uncertain of the amount Lambda would offer
but expected a value around $85,000, an estimate based on a market survey in Fortune of the
offers to graduates of the top five business schools that showed essentially a normal distribution
with a mean of $85,000 and with most offers falling in the range from $75,000 to $95,000.
2. In the certain, high-value condition, the role differed only in that recruits knew for certain that
the Lambda salary offer would be $85,000.
3. In the risky, low-value condition, recruits were uncertain of the amount Lambda would offer
but expected a value around $70,000 on the basis of a market survey that showed essentially a
normal distribution with a mean of $70,000 and with most offers falling in the range from
$60,000 to $80,000.
Page 8 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
4. In the certain, low-value condition, the role differed from Condition 3 only in that recruits
knew for certain that the Lambda salary offer would be $70,000.
Negotiation measures.
Participants were given 20 min to study their role. They then answered several prenegotiation
questions about their strategy and plans for the negotiation. One of these asked participants to
state their reservation price—the lowest salary amount that they would accept in the negotiation
instead of resorting to their alternative option. After completing the prenegotiation questions,
each participant was randomly assigned to a counterpart. The pairs of participants were directed
to a private area for a 0.5-hr-long negotiation session. They were then separated to answer
postnegotiation questions about whether they reached an agreement and, if so, the amount of the
salary contract.
Situation perception measures.
Next, we asked recruiters for their perceptions of the recruit's bargaining situation. Participants
were familiar from negotiation class with the concepts of the value and risk of an alternative
option and had been taught guidelines for estimating these from an opponent's negotiation
behavior (see
Lewicki & Litterer, 1985
). Perceptions of value were measured with the following
question: “What do you think is the amount of the recruit's best alternative to a negotiated
agreement? The amount the recruit expected to receive from an alternative firm if the negotiation
with you ended in an impasse is $__.” Perceptions of risk were measured with
the following question: “How certain was the recruit's alternative offer?” Responses were taken
on a 5-point scale with one end (1) labeled known with certainty, the midpoint labeled estimated
precisely, and the other end (5) labeled estimated loosely.
Personality perception measures.
Finally, we asked recruiters about their perceptions of the recruit's personality traits. We
requested that recruiters share their perceptions of stable, global characteristics by asking
participants to indicate “personality characteristics that you would expect to see again in future
interactions with this person” on a 5-point scale (1 = disagree strongly, 5 = agree strongly). We
emphasized that the question applied to cross-temporal, cross-situational traits so that
participants would not merely select characteristics that described the other's behavior in this one
bounded situation. We then presented 28 randomly ordered trait descriptors that marked the first
four factors of personality perception (O. P. John, personal communication, January 30, 1995).
These items and the scale reliability statistics may be seen in
Table 1
.
Page 9 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
Table 1 Trait Descriptors for Four Factors of Personality Perception
Results and Discussion
Negotiation measures.
Recruiters set reservation prices (M = $94.95K) above the level of their alternative option.
1
Recruits in the low-value conditions set reservation prices (M = $73.55K) slightly above their
alternative option, which created the intended large settlement zones in this condition. Recruits
in the high-value conditions set their reservation prices much higher (M = $85.83K), F(1, 175) =
169.65, p < .001, which created the intended smaller settlement zones. There was a slight
tendency for the recruit's reservation prices to be higher in the risky (M = $80.83K) than in the
certain conditions (M = $77.95K), F(1, 175) = 5.24, p < .05. There was no interaction of risk and
value manipulations in determining reservation prices.
Not surprisingly, the outcome of negotiations was highly affected by the manipulation of the
value of a recruit's alternative. As may be seen in
Table 2
, when the value of the recruit's
alternative was high rather than low, the settlement rate was lower (86% vs. 97%), [chi]
2
(1, N =
188) = 25.36, p < .001, and the average settlement amount was higher ($89.21K vs. $84.67K), F
(1, 170) = 20.50, p < .001. Negotiation outcomes were not significantly affected by the recruit's
level of risk nor by the interaction of the risk and value manipulations.
Page 10 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
Table 2 Negotiation Outcomes and Perceptions of the Recruit's Personality as a Function of the Recruit's Situation (Experiment
1)
Situation perception measures.
The results suggest that participants were partially successful in estimating their counterparts'
bargaining position. Recruiters ascribed an alternative to the recruit that was higher in the high-
value conditions (M = $89.10K) than in the low-value conditions (M = $84.04K), F(1, 185) =
13.01, p < .001. Not surprisingly, perceptions of the value of the recruit's alternative did not vary
as a function of the risk manipulation. There was an unanticipated interaction effect, F(1, 185) =
6.57, p < .05, which reflected a greater difference between perceptions of recruits' alternatives in
the low-value versus high-value conditions when the alternative was risky (Ms = $82.33K vs.
$91.43K) than when it was certain ($85.46K vs. $87.19K). However, participants were not
successful in perceiving the recruit's level of risk. Perceptions of the recruit's level of risk did not
differ at all between recruits who had a certain versus a risky alternative (Ms = 3.66 vs. 3.57).
2
Not surprisingly, perceptions of the recruit's risk did not differ as a function of the value
manipulation or as a function of the interaction.
Personality perception measures.
Summary scores for perceived personality factors were created by reverse coding the negative
items and then averaging the seven ratings for each factor. Perceptions of Agreeableness and
Emotional Instability were predicted to vary as a function of the manipulated factors, whereas
perceptions of Extraversion and Conscientiousness were not. As may be seen in
Table 2
, the
predicted differences but no other differences were obtained. Recruiters ascribed a more
agreeable character to recruits who had a low-value (M = 3.65) rather than a high-value (M =
Page 11 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
3.38) alternative offer, F(1, 179) = 7.75, p < .01. Recruiters ascribed a more emotionally
unstable character to recruits who had a risky (M = 2.25) rather than a certain (M = 2.04)
alternative offer, F(1, 178) = 5.31, p < .05. In sum, recruits placed in a situation that induced or
required haggling were taken to be disagreeable persons, and recruits placed in a situation that
induced waffling were taken to be emotionally unstable persons.
A question that the reader may ask is, Why did participants in the recruiter role ascribe traits
even though their ascriptions of the counterpart's situational constraints did in fact correspond to
the counterpart's actual situation? Classical models portray personality and situational
attributions as hydraulically related (
Heider, 1958
), yet our results indicate that situational and
personality ascriptions were both elevated. Our interpretation is based on a model of attributional
judgment (
Morris & Larrick, 1995
), which holds that after observing a behavioral effect, a perceiver
raises his or her confidence in each of its multiple possible causes to a degree depending on the
perceived extremity of effect. If the behavior is perceived as extreme (i.e., a high degree of
haggling or waffling in the role-play procedure), then a strong situational attribution does not
discount a strong personality attribution (
Morris, Smith, & Turner, 1998
). In our next study, we
measured the perceiver's impression of the extremity of the bargaining behavior to test this
interpretation and more incisive hypotheses about the roots of social misperceptions.
Experiment 2: Processes Contributing to Misperceptions
Heretofore we have tested predictions at the level of overall patterns of misperception. At this
point, we refine our analysis to four specific hypotheses about processes that contribute to
misperceptions of personality. Misperceptions originate from the combination of the primacy of
situations in the actual causes of negotiation behavior (Hypothesis 1) and the primacy of
personality in the perceived causes of negotiation behavior (Hypothesis 2). These two
hypotheses are expressed in
Figure 2
as the first two links in a sequential process that leads
negotiators to misperceive their counterparts. In Experiment 2 we tested each of these links
separately by measuring the recruits' actual personalities in addition to manipulating recruits'
situation and by measuring recruiters' perceptions of bargaining behavior, situation, and
personality.
Figure 2. Illustration of four hypothesized processes that contribute to misperceptions of negotiation counterparts: Primacy of
situational rather than personality causes of bargaining behavior (Hypothesis 1, or H1); primacy of personality rather than
situation in ascriptions for bargaining behavior (Hypothesis 2, or H2); personality ascriptions that persist rather than being
discounted by the uncertain situational information a negotiator receives (Hypothesis 3, or H3); and self-confirming
consequences of personality ascriptions (Hypothesis 4, or H4). As the bold arrows indicate, we expect that perceivers misjudge
their counterpart's personality, even though they pick up some cues to the counterpart's bargaining situation.
Page 12 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
Misperceptions of personality persist in negotiations because the information a perceiver
receives about the counterpart's situation is too uncertain to discount his or her personality
attributions (Hypothesis 3). Although one can estimate a counterpart's degree of situational
constraint from his or her words and actions in the negotiation, the information is rarely clear and
trustworthy enough for one to be certain of the presence of the situational factor. Both in
normative analyses and in empirical studies, little discounting of personality attributions follows
from information about the mere possibility that a situational factor is present (see
Morris & Larrick,
1995
).
A final process that perpetuates personality misperceptions is the effect of personality
ascriptions on subsequent interactions with the perceived. Personality ascriptions shape decisions
that affect the other person and can induce behavior consistent with the ascribed personality.
There is considerable evidence that an impression of an opponent as uncooperative leads one to
choose uncooperative strategies in future rounds of conflicts. This has been observed in studies
of conflict games, such as the multiple-round prisoner's dilemma (
Radlow & Weidner, 1966
), and in
studies of naturally occurring multiple-round conflicts, such as those between roommates (
Sillars,
1980
) and nations (
Jervis, 1976
;
Bar-Tal & Geva, 1986
). Because uncooperative strategies beget
uncooperative responses from one's counterpart (
Kelley & Stahelski, 1970
), they prevent the
disconfirmation of one's original impression and even create spurious confirmatory evidence for
one's initial impression. In organizational settings, there are several ways in which a perceiver,
such as a recruiter, makes decisions that place others in situations that lead them to act
consistently with the perceiver's impression of their personality (Hypothesis 4).
A first organizational mechanism for belief confirmation is the choice of how to resolve a
future conflict. One can either proceed through direct negotiation or through the arbitration of a
third party, such as an ombudsperson or mutual supervisor (
Ury, Brett, & Goldberg, 1989
). We predict
that employees who perceive their counterpart to be low in Agreeableness will avoid negotiation
because they expect that the other person will not exhibit the cooperativeness and flexibility
needed for negotiation to proceed effectively. Instead they will choose arbitration, which does
not require active cooperation from the parties. Yet, the structure of arbitration limits flexibility
and evokes competitive behaviors (
Lewicki & Litterer, 1985
). In sum, having the impression that the
other's personality is low in Agreeableness begets conflict resolution through arbitration, which,
in turn, induces uncooperative, inflexible behavior consistent with one's initial impression.
A second organizational mechanism for belief confirmation is the process of recommending
others for role assignments. For example, a new employee who comes across as having a low-
Agreeableness personality would be recommended for roles that require inflexible behavior (e.g.,
the role of handling external bargaining over one-time purchases, such as real estate, which
requires a demanding, intransigent stance). This new employee would not likely be
recommended for roles that require high-Agreeableness behaviors (e.g., the role of managing
relationships with important long-term clients, which requires cooperativeness and flexibility).
As with conflict resolution decisions, role assignment decisions that are influenced by
perceptions of Agreeableness will work to confirm that personality perception.
Method
Participants.
Participants were 210 MBA students from two sections of a course on negotiation at Stanford
University and from two sections of a course on negotiation at the University of Chicago (1 year
Page 13 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
after Experiment 1). A week before the negotiation exercise, participants were asked to fill out
the Big Five Index (BFI), a reliable and valid self-report inventory for the basic factors of
personality (see
John & Donahue, 1994
). Respondents were presented with 44 short phrases that
completed the sentence “I see myself as someone who …” and rated each on a 5-point scale (1 =
disagree strongly, 5 = agree strongly). Examples of phrases that load on the five factors are as
follows: for Agreeableness, “Is considerate and kind to almost anyone”; for Emotional
Instability, “Gets nervous easily”; for Extraversion, “Is outgoing, social”; for Conscientiousness,
“Does a thorough job”; and for Openness to Experience, “Is curious about many different
things.” Students were given this instrument along with several other questionnaires in a packet.
This provided our measure of actual personality.
Procedure.
The procedure was almost identical to that followed in Experiment 1 except for the addition of
several measures of the recruiter's perceptions of the recruit's bargaining behaviors and of his or
her organizational decisions affecting the recruit. The only change to the basic manipulation was
that the value of the recruit's alternative in the high-value conditions was increased by $10,000.
This was because recruiters in Experiment 1 had been willing to pay more than their alternative
was worth, and the average salary on the job market had increased. Immediately after
negotiating, participants indicated the extent to which their counterpart had engaged in a number
of specific tactical behaviors on a 5-point agreement scale (1 = disagree, 5 = agree). Some items
on the list were designed as measures of haggling, some as measures of waffling, and some as
filler. Some items designed to capture perceived haggling were the following: “‘haggled’ (argued
over the number) a lot before settling,” “made extreme requests,” “moved quickly to a
reasonable position” (reverse scored), and “responded to my requests for concessions” (reverse
scored). Some items designed to capture perceived waffling were the following:
“‘waffled’ (changed his or her mind) about his or her limit,” “was unsure or uncertain at
moments,” “stated a limit and stuck to it consistently” (reverse scored), “remained confident in
his or her bargaining stance” (reverse scored).
After completing the situation perception and personality perception questions, recruiters were
asked to make two decisions involving the recruit. First, recruiters were asked to imagine that in
the future they found themselves working in the same business organization with the person who
was their counterpart in this negotiation. Recruiters were asked to decide how to respond if a
conflict arose between themselves and the recruit. They were then to describe their intentions
and expectancies by rating on a 5-point scale (1 = disagree, 5 = agree) the following four
statements: “I would be willing to negotiate the issue with the recruit,” “I would be willing to let
a third party settle the issue,” “Negotiating the issue would prevent tension or animosity,” and
“Letting a third party settle the issue would prevent tension or animosity.”
In the final measure, recruiters were asked to make recommendations about which role in the
organization the recruit should be assigned to. The following four roles were described:
“relationship manager role, which involves contact with important, long-term clients and
requires positive interpersonal skills”; “external bargaining role, which involves handling one-
time transactions such as real estate purchases and which requires assertiveness and
competitiveness”; “scheduling and crisis management role, which requires precision in planning
and firmness in decisions”; and “innovation and research role, which requires the ability to
perform unstructured tasks.” The descriptions of these roles made it clear that an employee
would be required to display particular sorts of behavior in the performance of each role. The
first two were designed to be roles that require and thereby induce high-agreeableness and low-
agreeableness behaviors, respectively. The second two were an attempt to create roles that
Page 14 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
induce high-stability and low-stability behaviors.
Results and Discussion
Negotiation measures.
Recruiters set reservation prices (M = $96.24K) above the value of their alternative, as in
Experiment 1. Recruits with a low-value alternative set relatively low reservation prices (M =
$73.65K), which allowed a broad settlement zone, whereas those recruits with a high-value
alternative set substantially higher reservation prices (M = $98.02K), F(1, 97) = 261.02, p < .001.
Reservation prices in the risky conditions (M = $85.80K) were not significantly different than
those in the certain conditions (M = $85.11K), F(1, 97) = 0.49, ns.
As in the previous study, the value of a recruit's alternative greatly affected the outcomes that
were obtained. As may be seen in
Table 3
, moving from the low to the high condition, the
settlement rate decreases (98% vs. 54%), [chi]
2
(1, N = 94) = 35.43, p < .001, but the average
settlement amount increases ($84.93K vs. $94.24K), F(1, 70) = 19.26, p < .001. The rate and
amount of the settlement were not affected by the certainty of the recruit's alternative or by the
interaction of the manipulated factors.
Table 3 Negotiation Outcomes and Perceptions of the Recruit's Personality as a Function of the Recruit's Situation (Experiment
2)
Measures of recruits' tactical bargaining behaviors were important for testing hypotheses about
the determinants of personality perceptions. To derive measures of haggling and waffling
Page 15 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
behaviors by recruits, we performed a principal components factor analysis across ratings of
behaviors on a list of 16 specific tactics. Two factors emerged that accounted for 26% and 16%
of the total variance, respectively. The pattern of items loading on these corresponded to the a
priori haggling and waffling categories.
3
Scales for haggling ([alpha] = .85) and for waffling
([alpha] = .71) were constructed, using the items that loaded onto each of the two factors, by
reverse coding the negatively loading items and then averaging. Haggling behavior was much
more marked among recruits who had a high-value alternative (M = 3.36) than among those who
had a low-value alternative (M = 2.35), F(1, 94) = 53.10, p < .001. By regressing a recruit's
haggling score on the relevant personality factor (the recruit's actual score on the Agreeableness
dimension of the BFI) and the relevant situational factor (the value of the recruit's alternative),
we can test the hypothesis that situations are primary in the actual causes of bargaining behavior
(Hypothesis 1). As may be seen in
Table 4
, results support the hypothesis that bargaining behavior
is primarily driven by situational rather than personality factors.
4
Table 4 Recruit's Haggling Behavior Regressed on the Actual Value of the Recruit's Alternative and the Recruit's Actual
Agreeableness Score
Whereas results suggest that the haggling scale captured the dimension of behavior perception
that mediates the pattern of misperceived disagreeableness that we have observed (in Experiment
1 and the pilot experiment), the results do not suggest that the waffling scale captures the
bargaining behavior dimension underlying the pattern of misperceived emotional instability that
we have observed. Contrary to predictions, recruits' waffling behavior was not significantly
higher among recruits who had a risky (M = 2.73) as opposed to a certain alternative (M = 2.62),
F < 1. Although, as expected, perceived waffling was not correlated with the relevant personality
factor (Emotional Instability), the lack of an effect of the situational factor means a failure to
support the first hypothesis. This failure may reflect that the waffling scale missed the target—
the bargaining behavior affected by risk.
Situation perception measures.
Page 16 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
As in Study 1, participants were somewhat successful at perceiving their counterpart's
bargaining situation. Recruiters' perceptions of the value of the recruits' alternative varied as a
function of its actual value ($83.51K vs. $95.89K), F(1, 92) = 33.72, p < .001. Yet, as in
Experiment 1, recruiters' perceptions of risk were not responsive to the actual level of risk.
Oddly, perceptions of risk were responsive to the value of the recruits' alternative: Recruits who
actually had a higher value alternative were perceived to have a more certain offer. Perhaps
participants in the recruiter role assumed their opponents were averse to risk, so an opponent
with a higher apparent reservation price was inferred to have a more certain alternative.
However, given the inconsistent results with the certainty-perception variable, conclusions
cannot be drawn until further studies using more fine-grained measures are conducted.
Perceptions of the counterpart's personality.
Measures of the four trait factors reached acceptable levels of reliability (see
Table 1
).
Personality perceptions across conditions of the experiment may be seen in
Table 3
. As in Study 1,
the value of the alternative offer had a strong effect on the recruit's perceived agreeableness.
Perceptions of agreeableness were greater for recruits with low-value versus high-value
alternative offers (3.81 vs. 3.16), F(1, 94) = 26.63, p < .001. To test that this effect of the
situational factor was mediated by the recruiter's perception of the recruit's haggling behavior,
we regressed perceptions of disagreeableness on all of the variables that are prior to it in the
proposed path model (
Figure 2
). We obtained evidence for a mediation relationship, according to
the criteria of
Baron and Kenny (1986)
: The two independent variables (actual situation and actual
personality) individually predicted perceived disagreeableness, as did the proposed mediator,
haggling (Equations 1, 2, and 3 in
Table 5
). When haggling was included with the independent
variables, haggling still significantly predicted the dependent variable, whereas the recruit's
actual situation did not (Equation 6), which indicates that the effect of the recruit's situation on
the attribution of a disagreeable disposition to the recruit is indeed mediated by the recruit's level
of haggling.
5
As can be seen in
Figure 3
, there was an effect of the recruit's actual personality on
perceived agreeableness that was not mediated by the recruit's level of haggling. An
interpretation (for which we thank a reviewer) is that agreeableness affects other behaviors
besides haggling, and the recruiters had some ability to read these behaviors.
Table 5 Recruiter's Perception of the Recruit's Disagreeableness Regressed on Predictor Variables
Page 17 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
Figure 3. Observed relationships in Experiment 2 between actual causes of haggling behavior, perceived causes, and decisions
about future interactions. Only the significant paths from the full regression models in
Table 4
(Equation 3),
Table 5
(Equation 7),
Table 6
(Equation 7), and
Table 9
(Equation 7) are shown.
Table 6 Recruiter's Perception of the Value of the Recruit's Alternative Regressed on Predictor Variables
Page 18 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
Table 9 Recruiter's Preference for Arbitration (vs. Negotiation) Regressed on Predictor Variables
An isomorphic regression model of antecedents of recruiters' ascription of situational factors
was also run. Haggling, although primarily caused by the recruit's situation, was not primarily
attributed to the situation. As Equation 7 reveals, haggling predicted ascription of a high-value
alternative to the recruit only marginally ([beta] = .23, p < .10,
Table 6
), not nearly as well as it
predicted ascription of a disagreeable personality ([beta] = .55, p < .001,
Table 5
). As illustrated in
Figure 3
, the one significant predictor of perceptions of the value of the recruit's situation was its
actual value. This influence, which was not mediated by haggling, reflects that some other cues
to the value of the recruit's alternative were available.
Counter to our prediction and to the pattern observed in Experiment 1, perceptions of
emotional instability were not affected by the manipulation of the risk in the recruit's alternative,
as may be seen in
Table 3
. One possible explanation for the failure to replicate the pattern of
misperception is that the introduction of the waffling perception questions inhibited the trait
attribution on this dimension. That is, the pejorative connotations of waffling may have led
participants to be reticent in their trait ratings. Consistent with this interpretation, ratings of
emotional instability in Experiment 2 were generally lower than those in in Experiment 1, and
this depression of ratings was specific to emotional instability rather than being a general pattern.
Furthermore, in subsequent studies that have omitted the behavior perception measures, the
effect of the recruit's level of risk on the recruit's perceived personality was observed once again.
Although in Experiment 2 we have no evidence for a misperception of personality and hence
obviously cannot test mediating variables, there was support for the notion that a negotiator's
attribution on this personality dimension was predicted more by his or her counterpart's waffling
behavior than by the counterpart's actual level of emotional instability, as the regression model in
Table 7
reveals. Finally, perceptions of the Extraversion and Conscientiousness dimensions were
not affected by the manipulations.
Page 19 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
Table 7 Recruiter's Perception of Recruit's Emotional Instability Regressed on the Recruit's Actual Emotional Instability Score
and the Recruit's Waffling Behavior
Persistence of personality trait ascriptions.
Regression analyses also allowed us to test our third hypothesis that ascription of personality
traits persists in the face of information about possible situational causes for behavior. We
predicted that people's confidence in ascribing a personality factor would not be appreciably
discounted by cues to a situational factor that fall short of definite information about a sufficient
alternative (
Morris & Larrick, 1995
). The observed haggling behavior should lend credence to both
unobserved possible causes: a disagreeable personality and an attractive alternative. Because
more extreme levels of perceived haggling should lend relatively more credence to both
attributions, we have interpreted the positive association between personality and situational
attributions observed in Experiment 1 as arising from variation in the extremity of perceived
haggling behavior across dyads. Accordingly, we expected to once again observe a positive
bivariate relationship between personality and situational ascriptions, but we predicted that this
association would disappear once variation in perceived haggling was controlled (in
Figure 2
we
predict no direct association between the ascription of situational and personality factors). In
Tables 5 and 6
we can see that that the bivariate relationship is strongly positive (Equation 4), but
this relationship disappears when haggling is controlled (Equation 7).
Future behavior toward the perceived.
The fourth hypothesis is that erroneous perceptions of disagreeableness lead to decisions
regarding one's counterpart that induce behavior confirming one's misperceptions. A first test of
this hypothesis drew on questions about how the recruiter would choose to resolve a dispute with
the recruit if they ended up working in the same organization in the future: Would he or she
resolve the conflictual issues through direct discussion and negotiation or let a third party
arbitrate the matter? As may be seen in the means presented in
Table 8
, the experimental
manipulation of the value of the recruit's alternative produced differences in recruiters'
expectancies and preference for arbitration versus negotiation. To test that the difference across
Page 20 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
conditions of the experiment was mediated by personality perceptions (as illustrated in
Figure
2
), we performed a set of linear regressions using a summary variable that summarized
preference for arbitration. The composite variable was created from the four future conflict
variables through a principal component factor analysis, which indicated that the four variables
loaded onto a single dimension that explained 52% of the variance in the variables. Factor scores
for the four variables were approximately equal, and a composite variable was created by
summingthe two variables that tapped preference for arbitration and subtracting the sum of the
two variables that tapped preference for negotiation. The regression results presented in
Table 9
reveal that willingness to use arbitration is strongly predicted by ascription to the recruit of a
disagreeable personality. Importantly, as hypothesized, effects of more remote variables, such as
the recruit's alternative (see Equation 1) and the recruit's level of haggling (see Equation 3), are
mediated by the ascription variable (see Equation 7). Moreover, a parallel analysis revealed that
the effect of a high ascribed alternative was also mediated by ascribed disagreeableness.
Table 8 Recruiter's Decisions About Future Interaction as a Function of the Recruit's Situation
A second belief-confirmatory dynamic was tested by asking recruiters to recommend the
recruit for organizational roles. They considered roles that demand different kinds of behavior
from the people that fill them. Of interest are the two roles that require, respectively, high- and
low-agreeableness behavior.
6
As predicted, the manipulation of the recruit's alternative value led
to differences in the extent to which the recruit was later recommended for the relationship
manager and external bargaining roles—the relevant means may be seen in
Table 8
. Recruits in
the high-value alternative condition were recommended more for the external bargaining role
(3.67 vs. 2.58), F(1, 89) = 22.54, p < .001, and were deemed less appropriate for the relationship
manager role (3.51 vs. 3.82), although this latter difference was nonsignificant, F(1, 89) = 1.69,
ns. Recommendation decisions were submitted to regression analyses, isomorphic to those
described for the conflict resolution decision. As summarized in
Figure 3
, we again observed that
Page 21 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
ascribed disagreeableness mediated the effect of more remote variables. In sum, results of
Experiment 2 supported the hypothesis that the personality misperceptions that arise in
negotiations have consequences for decision making in organizations. False attributions of a
disagreeable personality led to organizational decisions regarding the new recruit that make him
or her unlikely to disconfirm the false impression and likely to reinforce it.
General Discussion
The current research greatly clarifies the dynamics of attribution errors and their consequences
in conflicts. In a pilot experiment we found support for two predicted patterns in which
bargaining behaviors arising from basic aspects of a counterpart's situation are erroneously
attributed to corresponding traits in the counterpart's personality. In Experiment 1 we found that
these misperceptions affect major dimensions of personality perception. Counterparts who had a
highly attractive alternative offer and hence required a higher salary to accept the job were
perceived to be lower on the Agreeableness dimension of personality. Also, counterparts who
had a less certain alternative offer and hence were less able to refer openly and consistently to
their alternative were perceived to be higher on the Emotional Instability dimension of
personality.
In Experiment 2 we tested hypotheses about specific processes that contribute to and sustain
misperceptions of the personalities of conflict counterparts. Misperceptions originate from the
combination of the primacy of situations in the actual causes of bargaining behavior (Hypothesis
1) and the primacy of personality in attributions (Hypothesis 2). Misperceptions are sustained by
the lack of discounting of personality ascriptions based on uncertain information about
situational causes (Hypothesis 3) and by the self-confirming effects of personality impressions
on subsequent interactions with the other person (Hypothesis 4). All four hypotheses were
clearly supported with regard to the pattern of perceived disagreeableness. Perceptions of
disagreeableness arose from the fact that a recruit's level of haggling depended primarily on the
situational factor (the value of his or her alternative), not on the corresponding personality factor
(level of Agreeableness), yet haggling was attributed by the recruiter almost exclusively to the
personality factor rather than to the situational factor. As expected, although recruiters' estimates
of the recruits' alternative did respond to the clues they received about its actual level, recruiters
did not significantly discount their personality ascriptions on the basis of their (uncertain)
estimates of the situational factor. And, finally, impressions of the recruit as being a disagreeable
type influenced future decisions to place the recruit in situations that induce further low-
agreeableness behavior: (a) the decision to resolve a future conflict with the recruit through the
adversarial procedure of arbitration rather than the collaborative procedure of negotiation and (b)
the decision to assign the recruit to an organizational role that requires demandingness and
intransigence rather than to a role that requires cooperativeness and flexibility. In sum, results of
Experiment 2 unequivocally reveal a systematic misperception in which negotiators in a
commanding economic position are attributed demanding, uncooperative personalities.
Our data are not inconsistent with recent findings that Agreeableness has a weak impact on
competitive bargaining in some situations (
Barry & Friedman, 1998
), but our emphasis is that the
impact of personality is small in comparison with the impact of the economic bargaining
situation and in comparison with the role that lay perceivers accord personality.
With regard to perceptions of Emotional Instability, it was not possible to test hypotheses
about component processes because the basic effect of the certainty condition on emotional
instability perceptions (seen in the pilot experiment and Experiment 1) failed to replicate in
Page 22 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
Experiment 2. Various features of the current findings and subsequent findings in our research
suggest that this misattribution vanished because the derogatory connotation of the waffling
items reduced their endorsement, which carried over to the emotional instability items. In
ongoing work, we have found that the misperception of emotional instability reappears when the
waffling measure is removed. Also, the effect occurs more strongly when the recruit's alternative
is high in value, probably because the alternative is more salient when it is attractive. Hence,
although our findings in general point to a pattern of misperception in which negotiators in an
uncertain, insecure economic position are attributed unstable, insecure personalities, we are less
confident (than with regard to the first pattern) that our measures fully capture this pattern or that
the pattern is very robust. The behavioral effects of a risky alternative may not be captured by the
construct of waffling, and they may be attributed to a personality disposition that does not fall
squarely on the Emotional Instability dimension. For example, the trait of insincerity (which we
used successfully in the pilot experiment) may be a correspondent inference evoked by the
behavior of inconsistent statements in bargaining, but not a trait that loads on the dimension of
Emotional Instability. Although the findings about Emotional Instability did not allow a test of
specific hypotheses, results were consistent with the general claim that major dimensions of
personality perceived in negotiation counterparts are mirages. Perceptions of a counterpart's
personality on the Agreeableness and Emotional Instability dimensions had little to do with the
counterpart's actual personality.
The Issue of Context Specificity in Personality Dispositions
We have conceived of personality in terms of the five-factor model of context-general traits,
such as Agreeableness. Although this is currently the leading approach to the study of
dispositions in psychology generally (
McCrae & John, 1992
) and in negotiation research specifically
(
Barry & Friedman, 1998
), it is not the only fruitful approach. An increasingly influential alternative
is defended by
Mischel and Shoda (1995)
, who have argued that personality can be conceptualized in
terms of behavioral dispositions specific to particular situational contexts. In extensive studies of
domains involving sustained social interaction, such as a summer camp, they have found that
tailor-made, context-specific measures do better than context-general measures in predicting
behavior. This can be understood in terms of the general notion that increased specificity in
constructs brings gains in fidelity (predictive validity) albeit losses in bandwidth (generality and
parsimony). The magnitude of these gains and losses will, of course, vary from domain to
domain. Proponents of context-general constructs such as Agreeableness argue that in many
domains, the gains in fidelity from narrowing one's constructs are small. To assess this we
compared the performance of the Agreeableness scale with a similar instrument that taps the
disposition to be competitive specifically in workplace conflicts (
Rahim, 1983
). We found (see
Footnote
4
) that the context-specific measure does allow slightly better prediction of haggling
behavior, but the effect remains small relative to that of the situational variable and relative to
the role accorded personality by perceivers.
A second, and independent, claim about context specificity is that lay perceivers form
impressions of others in terms of context-specific traits (
Mischel & Shoda, 1995
;
Shoda & Mischel, 1993
).
Undoubtedly person perception is this nuanced in some social settings, for instance, the
impressions formed by perceptive camp counselors. Yet, there are features of the domain we
studied—one-round negotiations between strangers—that impede such impressions. First, the
perceiver has information about the target person in only one situation (negotiating in the context
of one specific bargaining position) and hence cannot observe situation–behavior covariation.
Second, the perceiver cannot easily detect the target person's situation; negotiations create strong
situational pressure for actors to hide their true situation (strategically, negotiators strive to create
the impression that they have a favorable alternative). Hence, our expectation is that perceivers
Page 23 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
interpret bargaining behavior in terms of context-general dispositions. We investigated this in
an exploratory study using a distributive negotiation similar to the narrow-bargaining-zone
conditions of Experiments 1 and 2. Two weeks prior to the study, all participants were asked to
rate themselves on a series of eight descriptors of negotiation behavior in one of two contexts:
Four synonymous questions asked about their behavior in an advantageous bargaining situation
(having a favorable alternative), and four asked about their behavior in a disadvantageous
bargaining situation (having an unfavorable alternative). For example, they were asked to rate
their agreement with the statement “If I am in a strong bargaining position, I negotiate
aggressively” and were also asked to rate their behavior in the converse situation: “If I am in a
weak bargaining position, I negotiate aggressively.” Following the negotiation exercise,
negotiation opponents were asked to rate their counterpart on six similar if … then items. For
instance, they were asked to rate their opponent on descriptors such as “would negotiate
aggressively if in a strong bargaining position” and “would negotiate aggressively if in a weak
bargaining position.”
Relations between these actor ratings and observer ratings on the situation-contingent
dispositions are shown in
Table 10
. As may be seen in
Table 10
, actors believed that their own
behavior is only weakly related across strong and weak negotiation situations, r(1, 2) = .19,
whereas perceivers expected actors to show a great deal of consistency, r(5, 6) = .50. This fits the
expectation that perceivers, having only one observation of the other person, do not have
information about situational variability, unlike actors who can recall their own behavior across
different kinds of conflict situations. It may also be seen in
Table 10
that actor self-ratings in
particular contexts did not correlate highly with the corresponding perceiver ratings, as would be
expected if actual personality and perceived personality existed at this context-specific level
(these correlations were no higher than for context-general agreeableness perceptions in Study
2). Indeed, the correlation is highest between actors' self-ratings in strong situations and
perceivers' ratings of the actor in weak situations, r(1, 6) = .29, suggesting that perceivers
encoded behavior not as context-specific but in terms of a context-general impression of the
actor's competitiveness. In support of this, a regression of perceived behavior in strong situations
(Variable 5) on actor's self-rating in strong situations (Variable 1) and on perceived behavior in
weak situations (Variable 6) showed that the relationship to actor's self-rating disappeared ([beta]
= -.09), p > .30, whereas the relationship to the perceiver rating remained high ([beta] = .46), p
< .001. Finally,
Table 10
also shows that self-rated context-specific measures predicted perceived
haggling poorly, r(1, 4) = .15 and r(2, 4) = .07, ns, but perceived haggling predicted context-
specific perceptions very well, r(4, 5) = .41 and r(4, 6) = .45, p < .001. It seems that perceivers
were using negotiation behavior to form impressions of the target in strong and weak situations,
and they were doing so equally. Once again, this result suggests that perceivers, in this domain,
worked with a context-general impression of cooperativeness. To test this, we conducted a
partial correlation between perceived haggling and each of the context-specific if … then
measures controlling for perceptions of global Agreeableness (Variable 3). The previous
correlations of .41 and .45 decreased to .13 (ns) and .19 (p < .05), indicating that the global
impression largely mediated the relationship between perceived behavior and context-specific
perceptions. In sum, the empirical findings converge with our conceptual arguments to support
the assumption that first impressions of negotiation counterparts are formed in terms of context-
general constructs, such as Agreeableness.
Page 24 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
Table 10 Correlations Between Context-Specific Measures of Personality and Personality Perceptions
Implications for Theory
Social psychological processes in negotiation.
Early social psychological research on negotiation focused on personality dispositions that
affect negotiation outcomes (e.g.,
Rubin & Brown, 1975
) and was not, on the whole, successful in
identifying strong, reliable effects (
Thompson, 1990
). The social psychological study of negotiation
dwindled in the 1980s, at the same time that the study of cognitive biases in negotiation, led by
Neale and Bazerman (1991)
, flourished and redefined the field. Of late, researchers of social cognition
and perception in negotiation (
Thompson, 1997
) have carved out a new territory for social
psychological studies of negotiation. Most of the studies in this stream of research have explored
social psychological factors that distort a negotiator's perceptions of an opponent's preferences
(
Thompson & Hastie, 1990
) or of a third party's preferences (
Thompson & Loewenstein, 1992
). The current
research extends this approach to perceptions of an opponent's personality. More generally, we
draw on the traditional social psychological themes that situations primarily determine the
behaviors one sees, and yet one's subjective attributions for such behaviors, which guide one's
response, often miss the role of situational factors (
Ross & Nisbett, 1991
). We apply this situationist,
subjectivist analysis to elucidate how negotiation counterparts are misperceived.
Accuracy in social perception.
As we have reviewed, the portrait of the error-prone social perceiver painted during the 1970s
and early 1980s (e.g.,
Nisbett & Ross, 1980
;
Ross, 1977
) has been questioned by recent theorists who
have put forth more stringent criteria for errors. These criteria are met by the current studies in
the negotiation context. The ecological validity (
McArthur & Baron, 1983
) is high because the
interaction was natural and in a context that is familiar and motivating to participants. Although
the ecological validity of a role-play method might be questionable in a study of interactions that
are spontaneous or involve deep personal feelings, it is less problematic in a study of negotiation,
which, as a task, requires conscious stage-management of impressions (
Goffman, 1969
).
Furthermore, in the current studies we have examined biases with summary measures of the
major dimensions of personality perception rather than mere ad hoc lists of specific traits. We
have hypothesized and found that patterns of error are specific to major content dimensions such
as Agreeableness, as in other work that has comprehensively measured trait perception (
Robins et
Page 25 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
al., 1996
). Finally, we have established that the misattributions are made to stable dispositions
and hence affect decisions about future interactions; they are not just descriptions of the other
person's behavior within the bounds of one interaction (
Swann, 1984
). In sum, evidence that meets
the recently proffered criteria suggests that the fundamental attribution error (
Ross, 1977
) or
correspondence bias (
Jones, 1990
) is alive and well and living at the bargaining table.
Dynamics of self-confirming beliefs.
The notion that prophecies are self-fulfilling—that beliefs result in actions that evoke
confirmatory responses from others—was first developed by the sociologist
Merton (1940)
to
account for phenomena he observed in organizations. However, the concept has been refined
primarily by social psychologists, particularly those interested in stereotypes (e.g.,
Rosenthal &
Rubin, 1978
). Psychologists have provided detailed evidence for several distinct ways that beliefs
work toward their own confirmation. The simplest process, which almost certainly occurs in
interpersonal conflicts, is that beliefs guide a perceiver's attention to details of the environment
that are confirmatory (
Jones, 1990
). In face-to-face conversation, this process could occur through
a negotiator selectively seeking, attending to, and remembering those actions of the counterpart
that confirm prior impressions (
Snyder & Swann, 1978
). A more complex process, which has been
studied in interpersonal conflicts, is that beliefs support actions that constrain the opponent's
behavior and evoke behavior by the other person that is confirmatory of the beliefs. A classic
demonstration is
Kelley and Stahelski's (1970)
experiment with the repeated-trial prisoner's dilemma
game in which a competitive player's attribution of competitiveness to a counterpart supported
the competitive player's behavior of defecting; the expectation of competitiveness was then
confirmed when this defection induced defection from the counterpart. Also, within the context
of a more informationally rich negotiation conversation, there is a resonance whereby a
recruiter's negative expectancies about a recruit manifest themselves in negative nonverbal
behaviors that evoke confirming responses (
Word, Zanna, & Cooper, 1974
).
The contribution of the current research, however, lies in identifying longer-term
organizational processes through which specific misperceptions born in negotiation work toward
their own confirmation. In the spirit of Merton's examples, we examined two kinds of
organizational decisions that routinely occur following a negotiation and that clearly influence
what will be observed in the counterpart's subsequent behavior. We found that erroneous
personality perceptions affected (a) decisions about the procedure in which one will meet the
other person to resolve a future conflict and (b) decisions about work roles to which the other
person will be assigned. These decisions place the other person in situations that constrain
behavior; hence, they shape the other person's behavior and, as a result, the behavior the decision
maker will observe in the future. Notice also that these organizationally mediated self-
confirming dynamics not only perpetuate the decision maker's mistaken personality impression
but also potentially spread this false personality impression to many other observers because they
shape a person's long-term public behavior.
Implications for Practice
The current findings about systematic negotiator misperceptions suggest a few speculative
recommendations for people negotiating everyday conflicts. Our findings highlight that
negotiations have not only an economic outcome but also a social outcome. Acting to maximize
one's economic outcome by obtaining a valuable alternative can involve the cost of appearing to
be a person low in Agreeableness. However, the trade-off is not inevitable—the misperception
can be avoided. Research in progress (
Larrick, Morris, & Su, 1999
) suggests that if, after the
settlement, negotiators are provided with trustworthy, certain information about their
Page 26 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
counterpart's alternative, then negotiators can discount their ascription of disagreeableness. A
recommendation about avoiding misperceived Emotional Instability is easier to make because
there does not seem to be a trade-off between the economic and the social outcomes. Both
outcomes were better if negotiators had a certain rather than risky alternative option. Even when
the certain alternative was of only average value, negotiators with certain alternatives achieved
economic outcomes just as high as those with risky alternatives and came across as having more
stable personalities. Hence, certain alternatives should be obtained in preparing for negotiations
in which social outcomes matter.
References
Axelrod, R. (1984). The evolution of cooperation. New York: Basic Books.
[Context Link]
Baron, R. A. (1984). Reducing organizational conflict: An incompatible response approach. Journal of Applied
Psychology, 69, 272–279.
[Context Link]
Baron, R. A. (1988). Attributions and organizational conflict: The mediating role of apparent sincerity. Organizational
Behavior and Human Decision Processes, 41, 111–127.
[Context Link]
Baron, R., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research:
Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182.
[Context Link]
Barry, B., & Friedman, R. A. (1998). Bargainer characteristics in distributive and integrative negotiation. Journal of
Personality and Social Psychology, 74, 345–359.
[Context Link]
Bar-Tal, D., & Geva, N. (1986). A cognitive basis of international conflicts. In S. Worchel & W. G. Austin (Eds.),
Psychology of intergroup relations (pp. 118–133). Chicago: Nelson-Hall.
[Context Link]
Darley, J. M., & Batson, C. D. (1973). From Jerusalem to Jericho: A study of situational and dispositional variables in
helping behavior. Journal of Personality and Social Psychology, 27, 100–108.
[Context Link]
Dorris, W., Gentry, G., & Kelley, H. H. (1966). Unpublished raw data, University of California, Los Angeles.
[Context
Link]
Douglas, A. (1962). Industrial peacemaking. New York: Columbia University Press.
[Context Link]
Funder, D. C. (1995). On the accuracy of personality judgment: A realistic approach. Psychological Review, 102, 652–
666.
Ovid Full Text
Bibliographic Links
[Context Link]
Goffman, E. (1969). Strategic interaction. Philadelphia: University of Pennsylvania Press.
[Context Link]
Goldberg, L. R. (1990). An alternative “description of personality”: The Big-Five Factor structure. Journal of Personality
and Social Psychology, 52, 409–418.
[Context Link]
Heider, F. (1958). The psychology of interpersonal relations. New York: Wiley.
[Context Link]
Ichheiser, G. (1943). Misinterpretations of personality in everyday life and the psychologists's frame of reference.
Character and Personality, 12, 145–160.
[Context Link]
Jervis, R. (1976). Perception and misperception in international politics. Princeton, NJ: Princeton University Press.
[Context Link]
John, O. P., & Donahue, E. M. (1994). The Big Five inventory: Technical report of the 44-item version. Berkeley, CA:
Institute of Personality and Social Research.
[Context Link]
Page 27 of 31
Ovid: Morris: J Pers Soc Psychol, Volume 77(1).July 1999.52–67
1/8/2004
http://gateway1.ma.ovid.com/ovidweb.cgi
John, O. P., & Robins, R. W. (1994). Determinants of interjudge agreement on personality traits: The Big Five domains,
observability, evaluativeness, and the unique perspective of the self. Journal of Personality, 61, 521–551.
[Context Link]
Jones, E. E. (1990). Interpersonal perception. New York: Freeman.
[Context Link]
Jones, E. E., & Nisbett, R. E. (1971). The actor and the observer: Divergent perceptions of the causes of behavior. In E.
E. Jones, D. E. Kanouse, H. H. Kelley, R. E. Nisbett, S. Valins, & B. Weiner (Eds.), Attribution: Perceiving the causes of
behavior (pp. 79–94). Morristown, NJ: General Learning Press.
[Context Link]
Kelley, H. H., Helus, Z., Lamm, H., Servais, M., Veen, P., Zaleska, M., & Wouters, K. (1967). Direction of change in
problem difficulty and perceived source of change as independent variables in a negotiation game. Unpublished
manuscript, Training Center, Catholic University of Louvain.
[Context Link]
Kelley, H. H., Shure, G. H., Deutsch, M., Faucheux, C., Lanzetta, J. L., Moscovici, S., Nuttin, J. M., Rabbie, J. M., &
Thibaut, J. W. (1970). A comparative experimental study of negotiation behavior. Journal of Personality and Social
Psychology, 16, 411–438.
[Context Link]
Kelley, H. H., & Stahelski, A. J. (1970). Social interaction basis of cooperators' and competitors' beliefs about others.
Journal of Personality and Social Psychology, 16, 66–91.
[Context Link]
Kelley, H. H., & Thibaut, J. W. (1954). Experimental studies of group problem-solving and process. In G. Lindzey (Ed.),
Handbook of social psychology (Vol. 2, pp. 1–101). Cambridge, MA: Addison-Wesley.
[Context Link]
Larrick, R. P. (1998). [Predictors of ultimatum game play]. Unpublished raw data, Graduate School of Business,
University of Chicago.
[Context Link]
Larrick, R. P., & Boles, T. L. (1995). Avoiding regret in decisions with feedback: A negotiation example. Organizational
Behavior and Human Decision Processes, 63, 87–97.
[Context Link]
Larrick, R. P., Morris, M. W., & Su, S. K. (1999). Agreeing to disagree: From limiting conditions on negotiator
misperceptions to an intervention strategy. Unpublished manuscript, Graduate School of Business, University of
Chicago.
[Context Link]
Lewicki, R. J., & Litterer, J. A. (1985). Negotiation. Homewood, IL: Irwin.
[Context Link]
Loewenstein, G., Thompson, L., & Bazerman, M. H. (1989). Social utility and decision making in interpersonal contexts.
Journal of Personality and Social Psychology, 57, 426–441.
McArthur, L. Z., & Baron, R. M. (1983). Toward an ecological theory of social perception. Psychological Review, 90,
215–238.
Bibliographic Links
[Context Link]
McCrae, R. R., & John, O. P. (1992). An introduction to the five-factor model and its applications. Journal of
Personality, 60, 175–219.
Bibliographic Links
[Context Link]
Merton, R. K. (1940). Bureaucratic structure and personality. Social Forces, 17, 560–568.
Bibliographic Links
[Context Link]
Milgram, S. (1963). Behavioral study of obedience. Journal of Abnormal and Social Psychology, 46, 961–978.
[Context
Link]
Mischel, W., & Shoda, Y. (1995). A cognitive–affective system theory of personality: Reconceptualizing situations,
dispositions, dynamics, and invariance in personality structure. Psychological Review, 102, 246–268.
Ovid Full Text
Bibliographic Links
");
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment