Social Cognition,
Vol.
28, No. 4,2010, pp. 465-489
BETTER
SAFE
THAN SORRY:
SITUATIONAL CORRECTION IN
INTERPERSONAL COMPETITION
Yan-mei Li
Chinese Academy of
Sdences,
China
Koji Murata
Hitotsubashi University, Japan
Shu Li
Chinese Academy of
Sciences,
China
We conducted two experiments to investigate whether situational cor-
rection occurs when interpersonal competitors makedispositional infer-
ences from their opponent's situationally constrained behavior. In both
experiments, participants who expected interpersonal competition,
judged a target person's disposition after reading information about a
constrained or not constrained behavior of their opponent or non-op-
ponent. As predicted, we found that insufficient situational correction
occurred when the situationally constrained behavior of the opponent
reflected a disposition that implied a high threat level (competitiveness
or high ability) but sufficient correction occurred when the behavior
reflected a low threat-level disposition (cooperativeness or low ability).
Finally, we discussed the relationship between risk-avoidance and
situ-
ational correction in competition.
If you know yourself and your
enemy,
you'll never
lose a
battle.
—Chinese strategist Sun Tzu (6th century B.C.),
The
Art of
War
The reported research was partially supported by National Natural Science Foundation of China
(No.
70971127) and the Knowledge Innovation Project of the Chinese Academy of Sciences (No.
KSCX2-YW-R-130).
We thank
Prof.
Ying-Yi Hong for her indispensable counsel and helpful comments on earlier
versions of this article.
We
also thank He Tian for her contribution to this research and thank Drs.
Edmund
F.
and Rhoda E. Perozzi for extensive English and content editing assistance.
Correspondence concerning this article should be addressed to Yan-mei Li, Institute of Psychology,
Chinese Academy of
Sciences,
4A Datim Road, Chaoyang District, Beijing, China,
100101.
E-mail:
liym@psych.ac.cn.
465
466 LI
ET
AL.
In competition, people may benefit from knowing their opponent well be-
cause they can appraise their opponent's strengths, weaknesses, and strategies
(Ruscher
&
Fiske, 1990). How and how well do people know their opponent?
In this study we investigated people's tendencies to judge their opponent's
disposition in a situation in which they were expecting an interpersonal com-
petition.
To
gain understanding about others, people often observe the other person's
behavior and infer his or her disposition from their behavior. Although be-
haviors are often determined or partially determined by situational factors,
many studies have showed that people have a tendency to draw dispositional
inferences even when another's behavior is situationally determined or con-
strained (e.g., Gilbert
&
Jones, 1986; Jones & Harris, 1967). This tendency to-
ward insufficient situational discounting has been variously described as
"fundamental attribution error" (Ross, 1977), "correspondence bias" (Gilbert
& Malone,
1995;
Jones, 1979,1990), or "dispositional bias" (Trope, 1998; Bier-
brauer, 1979). Throughout this article, we will use the term
dispositional
bias.
Will
people make
a
situational correction if they infer their opponent's disposi-
tion when expecting an interpersonal competition? Our research investigates
whether and how people adjust their dispositional judgment when making an
inference from their opponent's situationally constrained behavior.
In Gilbert, PeUiam, and KruU's (1988) dual-process model, three stages—^be-
havioral categorization, dispositional characterization, and situational correc-
tion—follow each other sequentially to complete the process of dispositional
inference. The first two of these stages are relatively automatic processes that
need little effort or conscious attention, but situational correction is a deliber-
ate and relatively controlled process that requires the perceiver's effort and
cognitive resources (Gawronski,
2003;
also see Bargh,
1994).
Dispositional bias
occurs when the perceiver is xmable (e.g., does not have enough cognitive re-
sources) or unwilling to make a situational correction about the dispositional
inferences that he/she draws relatively automatically.
Some studies suggest that interdependence can lead people to allocate more
cognitive resources and make m.ore effort to understand a person on whom
their outcomes depend and to improve the accuracy of their judgment, at least
under some conditions (Neuberg
&
Fiske, 1987).
It was found that people in interdependent (task dependent) situations pay
increased attention to each other's attributes and to expectancy-inconsistent
information (Erber & Fiske, 1984; Neuberg & Fiske, 1987; Ruscher & Fiske,
1990;
Stevens & Fiske, 2000). When interdependence is symmetrical, people
individuate others on whom their outcomes depend in both cooperative and
competitive situations (Neuberg
&
Fiske,
1987;
Ruscher
&
Fiske, 1990). When
the dependence is asynunetrical, people whose outcomes depend more on
others
(e.g.,
powerless) individuate the others (e.g., power-holder) and do not
discount information which is inconsistent with expectancy (Stevens
&
Fiske,
2000).
Because of its role in undercutting expectancy-based impressions and
reducing Stereotypie thoughts and evaluations, interdependence may some-
what improve the accuracy of judgments.
SITUATIONAL CORRECTION IN INTERPERSONAL COMPETITION 467
However, studies which did not involve expectancy have suggested that in-
sufficient situational correction may still occur in the inferences of interdepen-
dent individuals.
People in interdependent situations, such as dating or the prisoner's dilem-
ma (PD)
game,
have been found
to
make more extreme dispositional inferences
(Berscheid, Graziano, Monson, & Dermer, 1976; Miller, Norman, & Wright,
1978) because they want to make future outcomes predictable and control-
lable by knowing the dispositions of the others (Heider, 1944,1958) on whom
their outcomes depend. Recent studies have also found that, although basic
negotiation behavior is highly determined by bargaining positions, negotia-
tors primarily interpret their counterpart's behavior in terms of the counter-
part's personality, such as his or her level of cooperativeness or agreeableness
(Morris, Larrick,
&
Su, 1999). These findings imply that people in interdepen-
dent situations may not make sufficient situational corrections when making
dispositional inferences based on their opponent's situationally constrained
behavior because interdependence increases their desire to know their oppo-
nent's disposition. However, the above studies did not directly manipulate
situational constraints.
COMPETITION AND SITUATIONAL CORRECTION
Interpersonal competition can be considered
to
be a
t3rpe
of interdependence
(e.g., Ruscher
&
Fiske,
1990;
Vonk, 1998). However, interpersonal competition
distinguishes itself from other types of interdependence in several
ways.
In in-
terpersonal competition the interdependent outcomes of competitors are
nega-
tively
correlated,
such that the success of one individual necessitates the failure
of the other (Deutsch, 1960,1973; Ruscher
&
Fiske, 1990). When a situation is
structured competitively, individuals seek an outcome that is personally ben-
eficial but detrimental to all others in the situation (Deutsch,
1962;
Johnson &
Johnson, 1989), Nevertheless, each individual's goal may be thwarted by the
capabilities of the other. These features of interpersonal competition are apt to
become potential causes of insufficient situational correction.
The thwarted goals of competitors evoke hostility toward their opponent
(Deutsch, 1960,1973), but such hostility can lead perceivers to derogate their
opponents' dispositions (Czopp, Monteith,
&
Mark,
2006;
Ross
&
Ward, 1995).
Klein and Kunda (1992) found that participants who expected competition
rated their opponent's ability more negatively. Similarly, Sabatier, Hunter, and
McLaughlin (1987) found that people impugn the motives and reasonableness
of their opponents. However, these studies also did not directly manipulate
situational constraints. In addition, the competitions in these studies were not
interpersonal. Nevertheless, it is quite conceivable that interpersonal competi-
tors may derogate their opponent to enhance themselves. If so, insufficient
situational correction may be more likely to appear when their opponent's
situationally constrained behavior appears to indicate a negative disposition.
468 LI
ET
AL.
Vonk's study (1999) provided direct evidence to support this prediction. She
found that people in a PD game took longer to process information about the
other player when the other player's behavior was situationally constrained.
She also jfound that the situational correction was significant when the other
player's situationally constrained behavior was cooperative but that the situ-
ational correction was not significant when the other player's situationally
constrained behavior was competitive. She argued that the reason that dispo-
sitional bias reduction (sufficient situational correction) did not appear when
the behavior was individualistic was probably that the potential costs of an
inaccurate judgment were higher in inferences relevant to socially negative
dispositions (i.e., individualistic, dishonest), which are potentially damaging
(Vonk, 1999).
However, for competitors the potentially damaging dispositions are not only
such socially negative
ones,
but also any high-level abilities that are relevant to
the competition task. Will people make a situational correction for a behavior
of their opponent that reflects high ability?
COMPETITORS' ESTIMATION OF THEIR OPPONENT'S ABILITIES:
"BETTER SAEE THAN SORRY"?
Findings by Chan and Ybarra (2002) showed that participants who expected
interpersonal competition believed that it was more costly to underestimate
than to overestimate how able one's opponent is and that their chances of
success would be higher if they overestimated their opponents. Bond's study
(1979) implies that this belief may influence people's judgment about their
opponent's ability. He found that a future opponent was judged to possess
higher levels of competition-relevant skills (i.e., quickness, intelligence) than
a non-opponent. Vonk's research (1998) suggests that competition produces
caution, careful processing, and conservative judgment as evidenced by the
observation that participants in a competitive outcome-dependent situation
did not derogate the task-relevant social skills of their opponent. They took
more time to read the relevant information, and their judgments were moder-
ate and relatively unaffected by initial expectancies (Vonk, 1998). The above
findings imply that people who expect interpersonal competition may not
derogate their opponent's task-relevant
abilities;
rather they may overestimate
these abilities in an illustration of the phrase, "better to be safe than sorry." So
it
is
possible that sufflcient situational correction may not occur in an inference
about an opponent's situationally constrained behavior which reflects high
ability. Since situational correction will lower the estimation of these abilities
and increase the risk of underestimating one's opponent; inferences like this
are contrary to a competitors' usual caution and conservativeness.
THE PRESENT RESEARCH
To summarize, few studies exist about how people who expect interpersonal
competition infer their opponent's dispositions from his/her situationally con-
SITUATIONAL CORRECTION IN INTERPERSONAL COMPETITION 469
Strained behaviors and whether they make a situational correction, although
we can find everyday indications of this in the real world. Even fewer stud-
ies have investigated situational correction from the perspective of perceived
threat.
We
conducted the following studies to explore these issues.
We reasoned that cautiousness may influence whether a situational correc-
tion will be sufficient. When an opponent's behavior reflects a disposition that
implies a high threat level (i.e., competitiveness or high ability), a sufficient
situational correction will lower the threat level of the disposition and may
increase the risk of underestimating the opponent. The potential cost will be
high if the inference is wrong. However, insufficient situational correction will
maintain a high threat level and may keep competitors alert. In this case the
potential cost will be low if the inference is wrong. However, when an oppo-
nent's behavior reflects a low threat-level disposition (i.e., cooperativeness or
low ability), a sufficient situational correction will increase the threat level and
may lower the risk of underestimating the opponent. Thus the potential cost
will be low if the inference is wrong. Insufficient situational correction, on the
contrary, will keep the threat level low, so the potential cost will be high if the
inference is
wrong.
Therefore, we predicted that insufficient situational correc-
tion is more likely to occur when an opponent's situationally constrained be-
havior reflects a disposition that implies a high threat level than when the be-
havior reflects a low threat-level disposition. We conducted two experiments
to test this prediction.
In competition, people may be quite concerned about their opponent's co-
operativeness and their task-relevant ability because cooperativeness deter-
mines their choice of strategy and ability determines task performance. In
Experiment 1, we focused on cooperativeness, and predicted that situation-
al correction is more likely to occur when an opponent's situationally con-
strained behavior is cooperative (reflects low threat-level disposition) than
that when the behavior is competitive (reflects high threat-level disposition;
Hypothesis
1). However, this predicted result could be interpreted as an ef-
fect of negativity bias (Skowronski & Carlston, 1989; Taylor, 1991) because
competitive or cooperative behaviors can be considered to be behaviors with
negative or positive valences. Alternatively, this predicted result could be in-
terpreted as an effect of a desire to derogate an opponent (Klein & Kunda,
1992;
Sabatier et al., 1987) because a sufficient situational correction about
cooperative behavior will decrease the positive evaluation (i.e., cooperative)
about an opponent and an insufficient situational correction about competi-
tive behavior will maintain the negative evaluation (i.e., competitive) about
an opponent.
To
rule out these alternative explanations, we focused on ability
in Experiment
2,
and predicted that situational correction is more likely to oc-
cur when an opponent's situationally constrained behavior reflects low abil-
ity (low threat-level disposition) than when the behavior reflects high ability
(high threat-level disposition;
Hypothesis
2).