Running head: FAMILIAR OTHERS =
IN
JUDGMENTS OF TRAITS
Familiar Others as Reference Points in Judgments of Traits
Szkoła Wyższa Psychologii
Społecznej (Poland)
and
Experiment provided=
evidence
for the recruitment of social exemplars as reference points in judgments of=
traits.
Making a judgment regarding one target-person facilitated making the same
judgment regarding the other. As predicted this was more the case when the =
two
targets were of the same sex and similar age (category-specific priming) th=
en
when they were not (cross-category priming). Both the category-specific and=
the
cross-category priming effects appeared less consistently for self than for
other social exemplars. Also, consistently with our model, both effects were
facilitated by focusing on attribute-centered rather than on person-centered
mental representations, particularly when the target of the priming task was
relatively unfamiliar and the target of the test task was highly familiar. =
Familiar Others as Reference Points in Judgments of Traits
Social judgments ar=
e not
made in a vacuum. When thinking about behaviors and characteristics of a
specific person, we are likely to consider not only what we know about that
person but also what we know about related social objects. Such involvement=
of
social reference points may facilitate judgments in at least two ways. Firstly, they may serve as substit=
utes
for poorly developed representations of relatively unfamiliar targets
(Karylowski, Konarzewski, & Motes, 2000; cf., Beike & Sherman, 1994=
). Secondly, they may function as anc=
hors,
helping in establishing standards or norms for the judgments to be made
(Biernat & Manis, & Nelson, 1991; Dunning & Hayes, 1996). The first type of function, which =
might
be described as judgments by proxy is likely to lead to assimilation
effects. The second, based on anchoring, is more likely to lead to contrast
effects.
What
kinds of social representations do people bring to mind as points of refere=
nce
in thinking about a specific person? Until recently, most theories and rese=
arch
in this area concerned the role of mental representations of socially defin=
ed
classes of persons, particularly stereotypes, e.g., racial, national, or ge=
nder
stereotypes (Hamilton & Sherman, 1994; Kunda & Thagard, 1996; Mackie
& Hamilton, 1993; Macrae, Stangor, & Hewstone, 1996). Nevertheless, the role of exemplar=
s,
that is cognitive representations of specific individuals, as reference poi=
nts
in judgments about people has also been proposed. (Andersen & Baum, 199=
4;
Andersen & Cole, 1990; Andersen, Glassman, Chen, & Cole, 1995; Ande=
rsen, Glassman, & Gold, 1998; Holyoak
& Gordon, 1983; Karylowsk=
i,
1990; Karylowski & Skarzynska, 1992; Karylowski, Konarzewski, & Mot=
es,
2000; Smith & Zarate, 199=
2).
Among
exemplars, self has been postulated to enjoy a privileged position as a point of reference in making judg=
ments
about others (cf., Dunning, 2003; Holyoak & Gordon, 1983; Karylowski, 1=
990;
Karylowski, Konarzewski, & Motes, 2000; Mussweiler & Bodenhausen,
2002). Several lines of research, including attributive projection (Holmes,
1968), attributive contrast (Dunning & Cohen, 1992), false consensus (M=
arks
& Miller, 1987), and attitudinal assimilation-contrast effects (Hovland
& Sherif, 1952) have been interpreted in such terms. In addition, recent
studies provide more direct, experimental, evidence that people do, in fact,
access information about themselves when making judgments about others (Dun=
ning
& Hayes 1996; Karylowski, Konarzewski, & Motes, 2000; Mussweiler &a=
mp;
Bodenhausen, 2002).
However, self may n=
ot be
the only social exemplar that people habitually bring to mind as a reference
point in making judgments about others..Research by Andersen, Karylowski and
their associates (Andersen & Baum, 1994; Andersen & Cole, 1990;
Andersen, et al., 1995; Karylowski, 1999; Karylowski, Konarzewski, & Mo=
tes,
2000) suggests that individuals may activate their representations of highly
familiar persons (e.g., best friends) as points of reference in making judg=
ments
about less familiar targets. Moreover, there is some preliminary evidence
(Karylowski, 1999; Karylowski, Konarzewski, & Motes, 2000) that familia=
r-other
exemplars might be more likely to be used as points of reference not only
compared to social stereotypes but also compared to self. The main purpose =
of
the present experiments was to provide further experimental evidence for the
spontaneous recruitment of familiar-other exemplars as points of reference =
in
social judgments. Our experiments were&nbs=
p;
designed to allow comparisons between effects involving representati=
ons
of familiar others, represent=
ations
of the self , representations of people in general, and gender- and age-
specific stereotypes .
&=
nbsp; &nbs=
p; &=
nbsp; &nbs=
p; &=
nbsp; Present
Research
We addressed the
utilization of social reference points using priming methodology. We expected that
answering a question regarding highly familiar Person A (priming target),
should facilitate answering the same question regarding relatively unfamili=
ar Person
B (test target). The exemplar model maintains that highly developed, access=
ible
exemplars are recruited when making judgments about relatively unfamili=
ar
targets (Karylowski, Konarzewski, & Motes, 2000). If so, making the same
judgment regarding a highly familiar person immediately before making a
judgment involving a relatively unfamiliar person should facilitate the
exemplar recruitment, thus speeding up the second judgment. It must be noted, however, that be=
cause
this prediction concerns the effects of direct activation of familiar-person
exemplars during the priming task, the spontaneous nature of exemplar
recruitment would not be established.
A spontaneous chara=
cter
of such a process leads to a
paralel prediction, namely, t=
hat answering
a question regarding a relatively unfamiliar Person B (priming target), sho=
uld
facilitate answering the same question regarding familiar Person A (test
target).This second prediction relies on the following reasoning: if answer=
ing
a question regarding the relatively unfamiliar Person A (priming target),
brings to mind the highly familiar Person B as a reference point (or one of
reference points), then we should be particularly fast at subsequently
answering the same question regarding Person B (test target).
We also hypothesize=
d that
both types of priming effects described above should be stronger if the
target-person of the first (priming) question and the target-person of the
second (test) question are seen as being of the same type (cf., Smith
& Zarate, 1992). We refer to such effects as category-specific priming
effects and distinguish them from cross-category priming effects. In
particular, we predicted that when two individuals belong to the same sex
category (male of female) and the same age category (generation), answering=
a
question about one will facilitate answering the same question regarding the
other more than when they belong to different categories.
The social judgment task employ=
ed in
the experiment consisted of t=
rait
judgments, undoubtedly one of the most widely used experimental tasks in so=
cial
cognition research (and, presumably, one that people often engage in
spontaneously outside of psychological labs, cf., Winter & Uleman, 1984;
Carlston & Skowronski, 1994). We used two types of trait judgment task.
Subjects were asked either to make judgments regarding targets (straight
judgments instruction) or to predict how would those targets judge themselv=
es
with respect to given traits (role taking instruction). We expected to find
evidence of a more predominant use of self as a reference point in the case=
of
the role taking instruction. Understanding how others think about themselves
calls for information about covert processes (thoughts and feelings) of oth=
ers.
Because self is characterized by rich and accessible representations of such
covert processes (Andersen, Glassman & Gold, 1998; Karylowski & Ranieri, 20=
04;
Karylowski & Niewiarowski, 2004), a tendency to use self as a point of
reference in judgments requiring role-taking might be enhanced.
We also predicted that presenti=
ng the
trait a few seconds before the presentation of the target during the priming
task would increase the effects of priming. On the other hand, present=
ing
the target a few seconds before the presentation of the trait should reduce
such effects. This prediction was based on the premise that social knowledg=
e is
cognitively represented in a variety of ways (cf., cognitive diversification
principle, Carlston & E. R. Smith, 1996). In particular, we assumed that
mental representations of social knowledge include both attribute-centered
representations (e.g., extroversion, politeness, or liking for country musi=
c)
and target-centered representations (e.g., my friend John, a man sitting ne=
xt
to me in a movie theater, or people in general). We reasoned that thinking
about a specific attribute-dimension prior to obtaining information identif=
ying
a target should increase the accessibility of attribute-centered, as oppose=
d to
target-centered, representations. Such attribute-centered representations a=
re
not restricted to a specific target and are likely to include examples of h=
ow
different social exemplars are placed along a particular attribute dimensio=
n.
Increasing the relative salience of attribute-centered mental representatio=
ns should,
therefore, increase the tendency to utilize social exemplars as points of
reference in making social judgments.
As a result, the effectiveness of both category-specific and
cross-category target priming should increase. The opposite effect should occur w=
hen
the target is presented first, followed by the delay, followed by the
presentation of the trait.
We further predicted that such
modifying effects of the delay manipulation on priming should occur primari=
ly
when the target person of the priming task is relatively unfamiliar (e.g.,
same-sex acquaintance) and the target person of the test task is highly
familiar (e.g., same-sex friend). It is under those circumstances that
people would be particularly =
likely
to spontaneously use the target person of the subsequent testing task as a
reference point (or one of reference points) in making trait judgments about
the priming target. (Note that the delay manipulation occurred for the prim=
ing
targets only). This prediction is theoretically important because it helps =
to
distinguish between two somewhat different processes that are postulated: a
spontaneous recruitment of familiar person exemplars taking place during
priming task involving unfamiliar targets and a prompted recruitment of such
recently primed exemplars taking place during the testing task.
Six targets were used in Experi=
ment 1:
self, most people, two
friends/family members (one college age and one older) and two acquaintances
(again, one college age and one older). The design involved an intentional
confounding of sex and age category memberships, both college-age targets w=
ere
of the same sex as the participant and both older targets were of the oppos=
ite
sex. Each target served as a priming target for some trials and as a test
target for other trials. The experiment included trials in which the priming
and the testing target were identical, thus establishing an ultimate priming
condition.
&=
nbsp; &nbs=
p; &=
nbsp; &nbs=
p; &=
nbsp; Me=
thod
Participants
A total of 144
Procedure and Materials<=
/u>
Participants were tested in
individual rooms.
Preliminary task. Subjects w=
ere
asked to enter to the computer names, initials, or other unique labels
identifying four people: (a) same-sex best friend; (b) opposite-sex parent;=
(c)
same-sex, college-age acquaintance; and (d) opposite-sex, older acquaintanc=
e.
Trait judgments task.
One hundred and fifty =
trait
adjectives[1] were u=
sed
in the trait judgment task. Half of the participants (straight judgments
condition) were simply asked to indicate, on a scale from, 1 (Not at all=
)
to 5 (Very much so), how well the given adjective describes the targ=
et
person. The remaining participants
(role taking condition) were asked to indicate, using the same scale,
how they thought the target person would describe himself/herself with resp=
ect
to the given adjective.
There were 144 experimental tri=
als
divided into 4 blocks, each consisting of 36 trials. The experimental trials
were preceded by six practice trials. Each trial consisted of three
questions. There were two tra=
it
judgment questions using the same trait-adjective followed by a filler ques=
tion
consisting of a simple math problem[2]. The trait-adjectives were presented=
in a
constant random order with a different trait used in each trial. A total of
six targets were used for each
participant: the four target persons named in response to the preliminary t=
ask
questions, yourself, and the control target most people. Targ=
ets
of the two judgments in each block of experimental trials represented 36
possible two-element combinations of those six targets and included six
combinations with the same target appearing twice (thus resulting in trials with the same question appe=
aring
twice). Each of the 36 test combinations appeared once in each of the four
blocks.
For half of the participants in=
each
of the two task conditions, both priming question and test question in each
trial were presented with the name of target person appearing above the tra=
it
(target first condition); for the other half, the name of the target person
appeared below the trait (trait first condition). Also, for half of the
participants in each of the
conditions above, there was a=
4 s
delay between the presentation of target person and the presentation of the
trait in the first (priming) question on each trial (delay condition). For the remaining half of the
participants, target and trait were always presented with no perceptible de=
lay
(no delay condition).
The overall design of the trait
judgments task was 2 x 2 x 2 x 6 x 6 x 6 x 4 with the last three factors as
within-subject. The between-subject portion of the design consisted of task (straight judgments vs. role taking), order of present=
ation
(trait-target vs. target-trait), delay (delay vs. no delay), and counterbalancing (6 versions of the
counterbalancing order for priming targets). The 144 experimental trials
constituted a completely balanced 6 (target of the first question) x 6 (tar=
get
of the second question) x 4 (replication) within-subject design. Response
latencies to test questions constituted the main measure of interest.
Results and Discussion
Latencies below 500 ms (.1 % of
responses to priming questions and 1.0 % of responses to test questions) and a=
ll
judgments associated with those latencies&=
nbsp;
were excluded. Also, latencies that were more than 3.5 standard
deviations above the subject’s mean for the matching category of
questions (1.1 % of responses to priming questions and 1.1 % of responses to
test questions) were truncated to that value. All of the latencies were then
converted to logarithms and collapsed across replications. All of the judgm=
ents
data were also collapsed. To facilitate presentation, mean latencies presented in text a=
nd in
figures are converted back to the metric scale. Because our sample incl=
uded
relatively few males and because preliminary analysis revealed no interpret=
able
effects involving subjects’ sex, the subjects’ sex was not incl=
uded
in the final analysis.
Exploratory Analysis=
On average, participants took 2=
.494 s
to make a priming judgment compared to 1.560 s to make a test judgment, =
t(143)
=3D 32.48, p < .001. A 6 (priming target) x 6 (test target) x 2
(presentation order) x 2 (delay) x 2 (judgment type) x 6 (counterbalancing
order) mixed-model ANOVA performed on test latencies revealed that, overall,
subjects who were given role-taking instructions required more time to make
test judgments (M =3D 1.624) than subjects who were given standard
instructions (M =3D 1.501), <=
/span>F(1,
96) =3D 5.06, p <
&n=
bsp; The
interaction between priming target and test target was also significant,
Effects of Exemplar Priming on
Judgments about Most People
A separate 5 (priming target) x=
2
(presentation order) x 2 (delay) x 2 (judgment type) x 6 (counterbalancing
order) mixed-model ANOVA performed on the latencies of responses to test
We see this pattern of results =
as an
indication that the self is clearly not a preferred point of reference for
making judgments about people in general. The data suggest that other social
exemplars play a more promine=
nt
role. Furthermore, this seems to be the case both for straight trait judgme=
nts
and for predicting others’ self-judgments. However an alternative
interpretation of our findings must be considered. This interpretation assu=
mes
that subjects were spontaneously accessing their representations of the most
people category when making judgments during the priming phase of each tria=
l.
Applying this interpretation to our findings would lead to the conclusion t=
hat
the most people category was =
more
likely to be accessed when making judgments regarding acquaintances than wh=
en
making judgments regarding self or regarding opposite-sex parent. However, =
the
finding that judgments about individuals (including friends and acquaintanc=
es)
were facilitated to a much greater degree by judgments regarding specific
exemplars, both matched and mismatched, than by judgments regarding most pe=
ople, makes such alternative interpretat=
ion
unlikely.
Category-Specific and Cross-Cat=
egory
Effects of Exemplar Priming
In order to provide a more dire=
ct test
of our predictions regarding category-specific and cross-category effects of
target priming, we standardized test latencies within each test target
(excluding trials with prime and test targets identical) and then computed
average standardized test latencies for trials with matched primes, mismatc=
hed
primes, and most people primes. A 3 (prime type) x 2 (presentation order) x=
2
(delay) x 2 (judgment type) x 6 (counterbalancing order) mixed-model ANOVA
performed on test latencies revealed the main effect of prime type, F(2, 19=
2) =3D
107.26, p < .001. Follow-up tests confirmed that response latencies for =
test
judgments preceded by matched primes were faster (M =3D -.228) than response
latencies for test judgments preceded by either mismatched primes (M =3D -.=
025), F(1,
96) =3D 33.21, p < .001, or most people primes, (M =
=3D
.426), F(1, 96) =3D 152.81, p < .001. Moreover, the advant=
age of
mismatched primes over most people primes was also highly significan=
t, F(1,
96) =3D 91.33, p < .001, indicating that judgments about individu=
als
were facilitated more by priming person-exemplars, even exemplars that did =
not
match judgment targets with respect to sex and age, than by priming the =
most
people (control) target. This finding provides further evidence for
cross-category (generalized) effects of target priming. It constitutes a
conceptual replication of a similar finding obtained by Karylowski, Konarze=
wski
& Motes (2000).
Effects of Presentation Order a=
nd
Delay on Priming
The predicted interaction betwe=
en
order of presentation, delay, and matching was significant, F(2, 192=
) =3D
8.24, p < .001. In addition, all interactions between order, dela=
y,
and single df contrasts involving matched, mismatched and most pe=
ople
primes were also significant, (F(1, 96) =3D 12.11, p <
.001, F(1, 96) =3D 4.1=
8, p
< .05, and F(1, 96) =3D 5.64, p < .02, for matched vs. =
most
people, matched vs. mismatched, and mismatched vs. most people
primes, respectively). This indicates that, as expected, both category-spec=
ific
effects and cross-category effects of target priming were qualified by the
order of presentation and delay. As is shown in Figure 2, introducing a del=
ay
after the trait presentation but before the target presentation resulted in
increasing the effects of target priming (F(2, 48) =3D 3.53, p <
.05, for the interaction between delay and matched vs. mismatched vs. mo=
st
people primes and F(1,48)
=3D 5.62, p < .05, =
F(1,48)
=3D 1.97, p < .17, and F(1,48)
=3D 2.08, p < .16 for interactions between delay and single df=
contrasts involving matched vs. most people, matched vs. mismatched,=
and
mismatched vs. most people primes, respectively). On the other hand, introducing del=
ay
after the target presentation and before the trait presentation resulted in
reduced effectiveness of target priming (F(2, 48) =3D 4.85, p < .01, for=
the
interaction between delay and matched vs. mismatched vs. most people
primes and F(1, 48) =
=3D 6.52, p
< .02, F(1, 48) =3D 3.88, p < .06, and F(1, 48) =
=3D
2.24, p < .15, for the interactions between delay and single d=
f
contrasts involving matched vs. most people, matched vs. mismatched,=
and
mismatched vs. most people primes, respectively). This pattern of
results is consistent with our reasoning that the recruitment of social
exemplars as reference points is facilitated by focusing on attribute-cente=
red,
as opposed to person-centered, mental representations. Finally, consistently
with our predictions, the interaction between order of presentation, delay,=
and
matching was significant for trials in which judgments about an unfamiliar
target were followed by judgments about a familiar target, F(1, 96) =3D 6.48, p=
<
.02, but not for trials involving familiar/unfamiliar and familiar/familiar
target combinations, both F’s=
< 1.[3]
Controlling for Similarity of
Responses to Priming and Test Items
Parallel analysis was conducted=
on standardized
absolute differences between the numerical responses to priming and test
questions. Those revealed a
significant main effect of prime type, F(2, 192) =3D 14.11, p=
<
.001. Follow-up tests showed that numerical responses were less different in
the case of matched trials, (M =3D -.182[4])
than in the case of mismatched trials (M =3D .076), F(1, 96) =3D 32.96, =
p
< .001, and in the case of trials with most people primes, (M<=
/u>
=3D .110), F(1, 96) =3D19.92, p < .001. However, in contra=
st to
the latency data, the means for the last two categories did not differ
significantly, F(1, 96) < 1. Similarly, the crucial three-way
interaction between order of presentation, delay, and matching did not appr=
oach
significance, F(2, 192) < 1, providing some evidence of the (at =
least
partially) independent nature of the latency effects.
Despite of such evidence, some =
basic similarities
of results obtained for latencies and for judgments raise the possibility t=
hat
the category-specific priming effects obtained for latencies were simply
due to subjects' tendency to =
make
more similar judgments in the case of trials involving matched targets than=
in
the case of trials involving =
mismatched
targets. To exclud=
e such
possibility we performed anal=
ysis
of covariance on latencies wi=
th
absolute response differences as covariates. This analysis confirmed the
significant effect of prime type, F(2,
191) =3D 91.65, p < .001.
Moreover, all three single df contrasts for this effect remai=
ned
significant, all p’s < .01. In addition, the interaction
between order of presentation, delay, and matching also remained significan=
t, F(2,
191) =3D 9.14, p < .001, and the interaction between order of
presentation, delay, and the matched primes versus mismatched primes contra=
st
was marginally significant, F(1, 95) =3D 2.97, p < .088.
Consistently with the results of the ANOVA, this last interaction was
significant for trials in which judgment about an unfamiliar target was
followed by a judgment about a familiar target, F(1, 96) =3D 6.13, <=
u>p
< .02, but not for trials involving familiar/unfamiliar and
familiar/familiar target combinations, both F’s < 1.
“Default” Targets in
Semantic Processing of Trait Labels
A
separate but related issue concerns “default” targets: social
exemplars that people are particularly likely to bring to mind when thinking
about traits. We examined this question by comparing the trait-first/no del=
ay
condition and the trait-first/delay condition with respect to the time requ=
ired
to make the first (priming) judgment for each of the six targets. If during
semantic processing of trait labels people spontaneously access representat=
ions
of specific social targets (exemplars and/or stereotypes), then introducing=
a
delay between the presentation of the trait label and the presentation of t=
he
target should result in a relatively greater reduction in the time needed to
make trait judgments regarding those specific targets. This effect would
manifest itself as an interaction between target of the first judgment and
delay. Because this reasoning does not apply to target-first condition, only
trait-first condition subjects were included in the subsequent analysis.
A 2 (delay) x 2 (task) x 6 (tar=
get of
first question) x 6 (counterbalancing order) mixed model ANOVA revealed a
significant main effects for delay, F(1, 48) =3D 4.51, p <=
.05,
and target, F(5, 240) =3D 6.13, p < .001. Mean latencies
presented in Figure 3 indicate that, in general, subjects in the delay
condition required less time to make judgments than subjects in the no delay
condition, which simply indicates that knowing the trait in advance made the
judgments faster. Also, follow-up tests revealed that, overall, judgments regarding most people=
were significantly slower than judgments regarding the self, best friend,
opposite-sex parent, or opposite-sex older acquaintance (ps < .05)
and marginally slower than judgments regarding same-sex acquaintance (p<=
/u>
< .06). In addition, judgments regarding the self were significantly fas=
ter
than judgments regarding same-sex friend, same-sex young acquaintance,
opposite-sex older acquaintance, (all ps < .05) and marginally fa=
ster
than judgments regarding opposite-sex parent (p < .11). This last set of findin=
gs
replicates the results of numerous earlier studies which found that trait
judgments regarding the self are typically faster than judgments regarding
other social targets (e.g., Kuiper, 1981; Lord, 1980; Mueller, Thompson, &a=
mp;
Dugan, 1986; Rogers, 1974; Rogers, Kuiper, & Kirker, 1977). Most importantly, the critical Del=
ay x
Target interaction was also significant, F(5, 240) =3D 2.26, p <
.05. Follow-up tests and examination of the means presented in Figure 3
indicate that it were the self-judgments and, to a lesser extent judgments
about most people, that benefitted most from the delay. Interactions=
between
delay and single df contrasts involving pairs of targets were
significant or nearly significant for self versus best friend, F(1, =
48)
=3D 3.09, p < .09, self versus same-sex acquaintance, F(1,=
48) =3D
4.44, p < .05 self versus opposite-sex parent, F(1, 48) =3D
2.80, p < .11, self versus older acquaintance, F(1, 48) =3D
11.66, p < .001, and most people versus older acquaintance=
, F(1,
48) =3D 4.39, p < .05. Hence, even though we found no indication =
of the
privileged status of self as a reference point (and in fact, we found evide=
nce
to the opposite), our data indicate that self does occupy a privileged posi=
tion
as a default referent in the semantic processing of trait labels.
Conclusions
Overall, the results provide st=
rong support
for the recruitment of social exemplars as reference points in judgments ab=
out
relatively unfamiliar targets (acquaintances and people in general). As
predicted, the effects were stronger when target and reference point were
matched with respect to sex and age than when they were mismatched. Moreove=
r,
consistently with our hypotheses, focusing on attribute-centered rather than
person-centered representations increased the effects. Consist=
ently
with the notion that not only similar but also dissimilar members of a cate=
gory
may be used as anchors in making judgments about other members of the same
category, priming effects remained statistically significant with similarity
controlled. As in previous experiments (Karylowski, Konarzewski, &am=
p;
Motes, 2000), no evidence was found for a privileged position of self as a
reference point in judgments about others and there was some evidence to the
opposite. This was the case even for the role-taking task. However, our res=
ults
indicate that self does occupy a privileged position as a default referent =
in
the semantic processing of trait labels.
Andersen, S. M. & Baum, A. (1994). Transference in interpersonal rela=
tions:
Inferences and affect based on significant-other representations. Journal
of Personality, 62, 459-497.
Andersen, S. M. & Cole, S. W. (1990). Do I know you?:=
The
role of significant oth=
ers in
general social perception. Journal =
of
Personality & Social
Psychology, 59, 384-399.
Andersen, S. M., Glassman, N. S., Chen, S., & Cole, S=
. W.
(1995). Transference in social
perception: The role of chronic accessibility in significant-other
representations. Journal of Persona=
lity
& Social Psychology, 69, 41-57.
Andersen, S. M., Glassman, N. S., & Gold, D. A. (1998=
).
Mental representations of the self, significant others, and nonsignificant
others: Structure and processing of private and public aspects. Journal of Personality & Social
Psychology, 75, 845-861.
Anderson, N. H. (1968).&=
nbsp;
Likeableness ratings of 555 personality-trait words. Journal
of Personality & Social Psychology, 9, 272-279.
Beike, D. R. & Sherman, S. J. (1994). Social inferenc=
e:
Inductions, deductions, and analogies. In R. S. Wyer, Jr. & T. K. Srull
(Eds.), Handbook of social cognition (s=
econd
edition).
Biernat, M., Manis, M., & Nelson, T. E. (1991). Stereotypes and standards of
judgment. Journal of Personality & Social Psychology, 60, 485‑499.
Carlston, D. E. & Skowronski, J. J. (1994). Savings in
the relearning of trait information as evidence for spontaneous inference
generation. Journal of Personality =
&
Social Psychology, 66, 840-856.
Carlston, D. E. & Smith, E. R. (1996). Principles of
mental representation. In E. T. Higgins & A. W. Kruglanski (Eds.) Social psychology: Handbook of basic
principles.
Dunning, D. (2003). The relation of self to social
perception. In: M. R. Leary &=
; J.
P. Tangney (Eds.), Handbook of self and identity (pp.421-441).
Dunning, D., & Cohen, G. L. (1992). Egocentric
definitions of traits and abilities in social judgment. Journal of
Personality & Social Psychology, 63, 341-355.
Dunning, D. & Hayes, A. F. (1996). Evidence for
egocentric comparison in social judgment.&=
nbsp;
Journal of Personality &
Social Psychology, 71, 213-229.
Hamilton, D. L. & Sherman, S. J. (1994). Stereotypes.=
In
R. S. Wyer, Jr. & T. K. Srull (Eds.), Handbook
of social cognition (2nd ed.).
Holmes, D. S. (1968). Dimensions of projection. Psychological Bulletin, 69, 248-26=
8
Holyoak, K. J. & Gordon, P. C. (1983). Social referen=
ce
points. Journal of Personality & Social
Psychology, 44, 881-887.
Hovland, C.
Karylowski, J. J. (1990). Trait prototypicality and the
accessibility of trait related
information in self‑other similarity judgments. Journal of Personality & Social
Karylowski, J. J. (1999). Rola reprezentacji Ja i repreze=
ntacji
osob bliskich w formulowaniu ocen dotyczacych osob malo znanych: Nie takie =
Ja
wszechmozne jak je maluja.[The role of self and significant others in judgm=
ents
regarding unfamiliar persons: The self not so powerful after all]. <=
i>=
Studia
Psychologiczne, 37, 77-92.
Karylowski, J. J.,
Konarzewski, K., & Motes, M. A. (2000). Recruitment of exemplars as reference points =
in
social judgments. Journal of Experi=
mental
Social Psychology, 36, 275-303.
Karylowski, J. J.
& Niewiarowski, J. (2004). Unpublished data.
Karylowski J. J. & Ranieri (2004). Self a= s a point of reference in thinking about about traits. Unpublished manuscript.<= o:p>
Karylowski, J. J. & Skarzynska, K. (1992).
Asymmetry effect in self-other
similarity judgments depends on priming. Social Cognition, 10, 235-254.
Kuiper, N. A. (1981). Convergent ev=
idence
for the self as a prototype: The "inverted U RT effect" for self =
and
other judgments. Personality & =
Social
Psychology Bulletin, 7, 438-443.
Kunda, Z. & Thagard, P. (1996). Forming
impressions from stereotypes, traits, and behaviors: A parallel
constraint-satisfaction theory. Psy=
chological
Review, 103, 284-308.
Lord,
C.G. (1980). Schemas and images as memory aids: Two modes of processing soc=
ial
information. Journal of Personality=
&
Social Psychology, 38, 257-269.
Mackie, D. & Hamilton, D. L. (1993). Affect,
cognition, and stereotyping: Interactive processes in group perception.=
Macrae, C. N., Stangor, C., & Hewstone, M.
(1996). Stereotypes and stereotypin=
g.
Marks, G., & Miller, N. (1987). Ten years=
of
research on the false consensus effect: An empirical and theoretical review=
. Psychological Bulletin, 102, 72-90=
.
Mueller, J. H., Thompson, W.B., &am=
p;
Dugan, K. (1986). Trait distinctiveness and accessibility in the self-schem=
ata.
Personality & Social Psychology
Bulletin, 12, 81-89.
Mussweiler, T. & Bodenhausen, G., V. (200=
2). Journal of Personality & Social
Psychology, 82, 19 – 32.
Rogers, T. B. (1974). An analysis of two cent=
ral
stages underlying the process of responding to personality items: The
self-referent decision and response selection. Journal of Research in Personality, 8, 128-138.
Rogers, T. B., Kuiper, N. A., & Kirker, W=
. S.
(1977). Self-reference and the encoding of personal information. Journal of Personality & Social
Psychology, 35, 677-688.
Smith, E. R. &
Zarate, M. A. (1992). Exemplar-based model of social judgment. Psychological Review, 99, 3-21.
Winter, L. & Uleman, J. S. (1984). When a=
re
social judgments made? Evidence for the spontaneousness of trait inferences=
. Journal of Personality & Social
Psychology, 47, 237-252.
&=
nbsp; &nbs=
p; &=
nbsp; &nbs=
p; &=
nbsp; Author
Note
Correspondence rega=
rding
this article should be addressed to Jerzy J. Karyłowski, Szko
Figure Captions
Figure 1. Latencies of responses to test questions as a
function of priming target and test target
Figure 2. Standardized latencies of responses to test quest=
ions
as a function of presentation order, delay, and matching between priming target an=
d test
target
Figure 3. Latencies of responses to priming questions in th=
e target-first
condition as a function of ta=
rget
and delay
=
[1]The adjectives were selected fr=
om the
[2] Subtraction problems involving two= -digit numbers with correct answers between 1 and 5 were used as filler questions, separating consecutive trials.
=
[3] We also conducted a 3 x 2 x 2 =
x 2 x
6 mixed model ANOVA with familiarity (unfamiliar priming target/ familiar t=
est
target vs. familiar priming target/ unfamiliar test target vs. familiar pri=
ming
target/ familiar test target) and matching (targets matched vs. targets
mismatched) as within-subject factors and presentation order, delay, judgme=
nt
type, and counterbalancing order as between-subject factors. This analysis
confirmed the main effect of matching and the Matching x Presentation Order=
x
Delay interaction. However, the four-way interaction between familiarity,
matching, presentation order, and delay did not approach significance.
Figure 1
![](3D"Karylowskirevppbmarch2005_files/image002.gif")
![](3D"Karylowskirevppbmarch2005_files/image004.gif")
![](3D"Karylowskirevppbmarch2005_files/image006.gif")
Figure 3