Ideology is related to basic cognitive processes involved
in attitude formation
Luigi Castelli, Luciana Carraro
To cite this version:
Luigi Castelli, Luciana Carraro. Ideology is related to basic cognitive processes involved in
attitude formation. Journal of Experimental Social Psychology, Elsevier, 2011, .
HAL Id: peer-00995260
https://hal.archives-ouvertes.fr/peer-00995260
Submitted on 23 May 2014
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
Ideology is related to basic cognitive processes involved in attitude formation
Luigi Castelli, Luciana Carraro
PII:
DOI:
Reference:
S0022-1031(11)00081-3
doi: 10.1016/j.jesp.2011.03.016
YJESP 2647
To appear in:
Journal of Experimental Social Psychology
Received date:
Revised date:
Accepted date:
13 December 2010
26 February 2011
19 March 2011
Please cite this article as: Castelli, L. & Carraro, L., Ideology is related to basic cognitive
processes involved in attitude formation, Journal of Experimental Social Psychology (2011),
doi: 10.1016/j.jesp.2011.03.016
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
ACCEPTED MANUSCRIPT
Ideology and attitude formation 1
RI
P
T
Running Head: Ideology and attitude formation
NU
SC
Ideology is related to basic cognitive processes involved in attitude formation.
MA
Luigi Castelli & Luciana Carraro
AC
CE
PT
ED
University of Padova
Please address correspondence to:
Luigi Castelli
Dipartimento di Psicologia dello Sviluppo e della Socializzazione
Via Venezia, 8
35131 Padova- Italy
Email: [email protected]
Keywords: political ideology; illusory correlation; attitude formation; negativity bias.
ACCEPTED MANUSCRIPT
ED
MA
NU
SC
RI
P
T
Ideology and attitude formation 2
Abstract
PT
Ideological orientation shapes the perception of the social world and conservatism is associated to
CE
an increased weighting of negative over positive information. In the present work we explored how
this ideology-based difference is also related to basic cognitive processes involved in attitude
AC
formation. In particular, we hypothesized that conservatives, as compared to liberals, would show
stronger illusory correlation effects when negative information is relatively infrequent. In Study 1
we employed the typical illusory correlation paradigm (Hamilton & Gifford, 1976) and results
confirmed the hypothesis: Conservatives developed more negative impressions toward the minority
group and showed consistent memory biases. In Study 2, positive information represented the
infrequent dimension and in this case no ideology-based difference was observed. Overall, findings
indicate that when exposed to numerically different novel groups and negative behaviors are
infrequent, illusory correlation effects are accentuated among individuals embracing conservative
rather than liberal views of the world. This result may help to understand why conservatives tend to
form more negative attitudes toward social minorities.
ACCEPTED MANUSCRIPT
Ideology and attitude formation 3
AC
CE
PT
ED
MA
NU
SC
RI
P
T
165 words
ACCEPTED MANUSCRIPT
Ideology and attitude formation 4
Stereotypes about social groups often reflect exaggerations of real group differences
T
(Brigham, 1971; Ford & Stangor, 1992), but they may also be formed even in the absence of any
RI
P
real difference between the groups. This has been clearly demonstrated by the pioneering studies by
Hamilton and colleagues on illusory correlation (Hamilton, Dugan, & Trolier, 1985; Hamilton &
SC
Gifford, 1976). In the standard experimental procedure participants are presented with a series of
NU
positive and negative behaviors performed by members of two social groups (e.g., Group A and
Group B). Importantly, one group is larger than the other, and the relative frequency of the two
MA
types of behavior is also manipulated. For instance, Group A could be twice as large as Group B,
but the ratio between positive and negative behaviors within the two groups is held constant. In one
ED
of the original experiments (Hamilton & Gifford, 1976; Study 1), Group A performed 18 positive
and 8 negative behaviors, whereas Group B performed 9 positive and 4 negative behaviors.
PT
Although there was no factual basis for considering one group as better than the other, perceivers
CE
formed more negative impressions about Group B. In addition, memory biases also emerged and
people erroneously overestimated the frequency of the co-occurrence of the two infrequent
AC
information (i.e., being a member of Group B and performing undesirable behaviors). This effect
has been explained by assuming that rare information is particularly distinctive and therefore it
attracts the attention so that to become extremely accessible (Hamilton & Gifford, 1976; Hamilton
et al., 1985; for a review, see Stroessner & Placks, 2001). For instance, it has been shown that
people spend more time processing sentences describing members of the infrequent group
performing infrequent behaviors and this particularly careful processing is responsible of the typical
illusory correlation effects (Stroessner, Hamilton, & Mackie, 1992). Although various alternative
(or complementary) accounts have been proposed (Fiedler, 1991; Rothbart, 1981; Smith, 1991), the
distinctiveness of the information appears to play a key role (Risen, Gilovich, & Dunning, 2007;
Sherman, Kruschke, Sherman, Percy, Petrocelli, & Conrey, 2009; Stroessner & Placks, 2001).
ACCEPTED MANUSCRIPT
Ideology and attitude formation 5
Ideology-based differences in the processing of valenced information. Political ideology is
related to several personality and cognitive styles differences (Jost, Glaser, Kruglanski, & Sulloway,
T
2003), as well as to the reactions to positive and negative information. For instance, conservatives
RI
P
display more extreme physiological reactions (e.g., skin conductance) to threatening information
(Oxley et al., 2008). Shook and Fazio (2009) also recently nicely demonstrated that conservatives,
SC
as compared to liberals, are more likely to display avoidant strategies while exploring the
NU
environment which, in the end, lead them to stronger learning asymmetries, favoring the learning of
negative over positive stimuli. Other research (Carraro, Castelli, & Macchiella, 2011) has shown
MA
that negative information automatically grabs the attention of conservatives. For instance,
conservatives, as compared to liberals, responded more slowly to negative words in a Stroop task
ED
(Carraro et al., 2011, Study 1). Similarly, spatial attention was related to personal ideology and
conservatives were more likely to direct their attention toward spatial locations in which negative
PT
information was presented (Carraro et al., 2011, Study 2), even when other individual differences
CE
(i.e., need for closure and need for cognition) were taken into account (Carraro et al., 2011, Study
3).
AC
Thus, political ideology appears to be associated to asymmetries in the way positive and
negative information is processed, with conservatives being more vigilant toward negative
information. In the current work we explored whether the distinctiveness of rare negative behaviors
when performed by numerical minority groups is even more accentuated for perceivers who
embrace a conservative view of the world. This would be indexed by stronger illusory correlation
effects among conservatives, that is an increased biased perception about numerical minority
groups.
Study 1
Participants. Two-hundred and thirty-four students (194 females) participated in the study1.
ACCEPTED MANUSCRIPT
Ideology and attitude formation 6
Procedure. Participants were shown 39 sentences each describing a behavior performed by
either a member of Group A or B. Group A was larger than Group B (26 vs. 13 exemplars).
T
Sentences were either positive (27 sentences; e.g., “Jim, who belongs to Group A/B, has given way
RI
P
to sit to an elderly lady.”) or negative (12 sentences; e.g., “James, who belongs to Group A/B,
usually tells many lies.”), and the ratio between positive and negative behaviors was identical
SC
within the two groups (i.e., 0.44). Thus, members of Group A performed 18 positive and 8 negative
NU
behaviors, whereas members of Group B performed 9 positive and 4 negative behaviors. Sentences
were presented one after the other in a random order and each of them remained visible for 7
MA
seconds. Participants were then required to evaluate the two groups along 11 traits (funny, irritable,
intelligent, stupid, willing, sociable, brilliant, lazy, happy, unhappy, unpleasant). Responses had to
ED
be provided along 5-point Likert scales from 1 (= not at all) to 5 (= very much). Subsequently,
participants were provided with the full list of behaviors and asked to indicate the group
PT
membership of the actor of each behavior. This cued recall task enabled to assess how many
CE
positive and negative behaviors were attributed to the two groups. Next, a conceptually similar
measure was administered and participants were asked to estimate how many behaviors were
AC
performed by members of Group A and how many of them were negative. Identical questions were
administered in relation to Group B. Finally, political ideology was assessed by asking participants
to report their level of agreement (from 1 = “not at all” to 7 = “very much”) with 5 different topics
(reduction of immigration, medically assisted procreation, homosexual marriage, use of arms for
personal defense, adoption by homosexual couples; after appropriate rescaling assigning higher
values to conservative ideologies, α = .73, M = 3.77, SD = 1.22)2.
Results
Illusory correlation. After appropriate rescaling (i.e., high scores indicate more positive
evaluations), the mean evaluation of Group A (α = .81) and Group B (α = .85) were separately
calculated. A t-test showed more negative attitudes toward Group B (M = 3.12, SD = .56) than
ACCEPTED MANUSCRIPT
Ideology and attitude formation 7
Group A (M = 3.65, SD = .48), t(232) = 9.73, p < .001. As for the cued recall task, a phi coefficient
correlation was computed for each participant3 (see Hamilton & Gifford, 1976) in such a way that
T
positive values indicated an illusory association between Group B and negative behaviors. The
RI
P
observed value was indeed positive (M = .21, SD = .40) and significantly higher than zero, t(216) =
7.43, p <.001. As for the estimation of the frequency of negative behaviors within each group, we
SC
calculated the perceived proportion of negative behaviors given the overall number of behaviors
NU
attributed to the group. As expected the proportion was higher in relation to Group B (M = .59, SD
= .21) than Group A (M = .38, SD = .19), t(223) = 9.74, p < .001. Overall, strong illusory
MA
correlation biases emerged.
Illusory correlation as a function of ideology. From the evaluative ratings toward the two
ED
groups we calculated the difference between the perception of Group A and Group B. A regression
analysis showed that the score on the political ideology scale was a significant predictor of the
PT
differential evaluation of the two groups, β = .17, t(232) = 2.55, p = .011, so that increased levels of
CE
social conservatism were associated to more negative evaluations of Group B as compared to Group
A. The phi correlation coefficient derived from the cued recall task was also significantly associated
AC
to political ideology, β = .22, t(216) = 3.31, p < .001, and, as predicted, the illusory correlation
between Group B and negativity was accentuated among conservatives. As for the third index, we
calculated the difference between the proportion of negative behaviors attributed to Group B and
Group A. In line with previous findings, political ideology was again a significant predictor, β =
.20, t(223) = 3.11, p = .002, further confirming the stronger tendency among conservatives to incur
into biased perceptions of the minority group. Although the gender composition of our sample was
unbalanced, exploratory analyses were carried out including participant gender and its interaction
with ideology as predictors. In no case, significant effects involving gender emerged (all ps > .13)
suggesting that ideology is related to illusory correlation independently of participant gender.
ACCEPTED MANUSCRIPT
Ideology and attitude formation 8
Future research with a larger sample of males, however, should more specifically focus on this
T
issue.
RI
P
Study 2
In Study 1 illusory correlation was consistently stronger among conservatives and we
SC
hypothesized this was due to their increased attention toward negative information (Carraro,
NU
Castelli, & Macchiella, 2011). However, alternative explanations could be advanced. Indeed, one
possibility is that conservatives, as compared to liberals, are more heavily influenced by the
MA
numerical composition of the two groups and provide overall more negative evaluations toward
minorities independently from the valence of the behaviors. In addition, conservatives’ need for
ED
structure may foster an accentuation of the (perceived) difference between the two groups (see Jost
et al., 2003), and thus exaggerate illusory correlation. In this case, political ideology should predict
PT
the strength of illusory correlation even when desirable behaviors are relatively less frequent than
CE
undesirable behaviors. In contrast, if political ideology impacts onto illusory correlation modulating
the salience of the infrequent negative information, no relation between political ideology and
AC
illusory correlation should emerge when desirable behaviors are infrequent.
Participants. One-hundred and thirty-four students (124 females) participated in the study.
Procedure. The structure of the study was very similar to the one employed in Study 1. One
key difference was in the valence of the presented behaviors. Indeed, positive behaviors were less
frequent than negative behaviors. Members of Group A performed 18 negative and 8 positive
behaviors, whereas members of Group B performed 9 negative and 4 positive behaviors. The
evaluation of the two groups and cued recall were assessed as in Study 1. As for the frequency
estimation task, we asked participants to indicate the overall number of behaviors performed by
members of Group A and how many of them were positive. The same questions were then asked in
relation to Group B. Political ideology was assessed through the same scale used in Study 1 (after
ACCEPTED MANUSCRIPT
Ideology and attitude formation 9
appropriate rescaling assigning higher values to conservative ideologies, α = .64, M = 3.96, SD =
1.08).
T
Results
RI
P
Illusory correlation. The mean evaluation of Group A (α = .85) and Group B (α = .83) were
separately calculated as in Study 1. A t-test showed more positive attitudes toward Group B (M =
SC
3.27; SD = .49) than Group A (M = 3.12; SD = .54), t(133) = 2.04, p < .05. A phi coefficient
NU
correlation was computed from responses in the cued recall task, in such a way that positive values
indicated an overassociation of Group B with positive behaviors. Although the value was positive
MA
(M = .02, SD = .43), it was not different from zero, t(129) = .59, ns. Finally, the proportion of
positive behaviors attributed to Group A (M = .48; SD = .17) and Group B (M = .57; SD = .22) was
ED
separately calculated, and, as expected, the perceived proportion was higher in the case of Group B,
t(131) = 2.93, p < .005.
PT
Illusory correlation as a function of ideology. Regression analyses were performed as in Study
CE
1. In no case political ideology was predictive of illusory correlation biases (all ps >.32). Next, we
divided the sample into quartiles and t-tests were performed comparing the responses of the
AC
participants with the most extreme liberal and conservative attitudes. In no case, political ideology
was related to the measures of illusory correlation, all ps > .394.
General discussion
Data revealed that although the objective information participants were exposed to did not
legitimate any differential perception toward the two groups, participants overestimated the
negativity (Study1) or positivity (Study 2) associated to the less numerous group (Hamilton et al.,
1985; Hamilton & Gifford, 1976). Thus, observers overestimated the co-occurrence of statistically
infrequent events, and, confirming previous findings (see Mullen & Johnson, 1990), the effect was
somehow stronger when the infrequent behaviors were undesirable.
ACCEPTED MANUSCRIPT
Ideology and attitude formation 10
Most notably, illusory correlation effects involving negative (but not positive) information
were significantly accentuated among people embracing conservative ideologies instead of more
T
liberal ideologies. Therefore, the present data demonstrate a link between individual differences in
RI
P
political ideology and a basic cognitive process involved in the formation of attitudes toward social
groups. In general, conservatives appear to be more responsive to environmental threats and be
SC
more affected by them (Carraro et al., 2011; Oxley et al., 2008; Shook & Fazio, 2009). In the
NU
present work we showed that social conservatism influences the memory for the negative behaviors
performed by the minority group as well as the evaluation toward such group. Ideology-based
MA
differences thus modulate the distinctiveness of negative information, and we here identified one
relevant consequence on social perception, namely more negative impressions toward the minority
ED
group. Importantly, ideology was not related to illusory correlation phenomena when positive
information was infrequent. Therefore, alternative explanations based on an overall more negative
PT
perception of minorities or on conservatives’ stronger need to accentuate the difference between the
CE
groups seem unlikely. In contrast, ideology appears to shape illusory correlation only when negative
behaviors represent the infrequent dimension. Distinctiveness-based illusory correlation contributes
AC
to the formation of prejudiced attitudes even in absence of any real difference between groups and
political ideology may indeed impact onto the distinctiveness of negative information further
accentuating such prejudiced attitudes. It has to be remarked that the items included in the ideology
scale were related to sensitive social issues which are typically opposed by conservatives (e.g., gay
marriage, arms control) and, therefore, in the current studies conservatism partially overlapped with
avoidance motivations. Future research will have to ascertain whether responses concerning other
policies which are typically favored by conservatives (e.g., privatization and deregulation) are also
related to the observed illusory correlation effects.
Indubitably, people embracing conservatives views of the world often hold more prejudiced
attitudes (see Duckitt, Wagner, du Plessis, & Birum, 2005) and are more likely to justify unequal
ACCEPTED MANUSCRIPT
Ideology and attitude formation 11
treatments toward social groups (Sidanius & Pratto, 1999). In the present work we showed that
ideology-based differences are associated to a key basic cognitive mechanism involved in the
T
formation of social attitudes, and findings may help to illuminate one of the routes that contribute to
AC
CE
PT
ED
MA
NU
SC
RI
P
lead conservatives to develop more negative perceptions about minorities.
ACCEPTED MANUSCRIPT
Ideology and attitude formation 12
T
References
RI
P
Brigham, J. C. (1971). Ethnic stereotypes. Psychological Bulletin, 76, 15-38.
Carraro, L., Castelli, L., & Macchiella, C. (2011). The automatic conservative: Ideology-
SC
based attentional asymmetries in the processing of valenced information. Manuscript submitted for
NU
publication.
Duckitt, J., Wagner, C., du Plessis, I., & Birum, I. (2002). The psychological bases of
MA
ideology and prejudice: Testing a dual-process model. Journal of Personality and Social
Psychology, 83, 75-93.
ED
Fiedler, K. (1991). The tricky nature of skewed frequency tables: An information loss account
of distinctiveness-based illusory correlations. Journal of Personality and Social Psychology, 60, 24-
PT
36.
CE
Ford, T. E., & Stangor, C. (1992). The role of diagnosticity in stereotype formation:
Perceiving group means and variance. Journal of Personality and Social Psychology, 63, 356-367.
AC
Hamilton, D. L., Dugan, P. M., & Trolier, T. K. (1985). The formation of stereotypic beliefs:
Further evidence for distinctiveness-based illusory correlation. Journal of Personality and Social
Psychology, 48, 5-17.
Hamilton, D. L., & Gifford, R. K. (1976). Illusory correlation in intergroup perception: A
cognitive basis of stereotypic judgments. Journal of Experimental Social Psychology, 12, 392-407.
Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway, F. J. (2003). Political conservatism as
motivated social cognition. Psychological Bulletin, 129, 339-375.
Mullen, B. & Johnson, C. (1990). Distinctiveness-based illusory correlations and
stereotyping: A meta-analytic integration. British Journal of Social Psychology, 29, 11-28.
ACCEPTED MANUSCRIPT
Ideology and attitude formation 13
Oxley, D. R., Smith, K. B., Alford, J. R., Hibbing, M. V., Miller, J. L., Scalora, M., Hatemi P.
K., & Hibbing, J. R. (2008). Political attitudes vary with physiological traits. Science, 321, 1667-
T
1670.
RI
P
Risen, J. L., Gilovich, T., & Dunning, D. (2007). One-shot illusory correlations and
stereotype formation. Personality and Social Psychology Bulletin, 33, 1492-1502.
SC
Rothbart, M. (1981). Memory and social beliefs. In D. L. Hamilton (Ed.), Cognitive processes
NU
in stereotyping and intergroup relations (pp. 145-181). Hillsdale, NJ: Erlbaum.
Sherman, J. W., Kruschke, J. K., Sherman, S. J., Percy, E. J., Petrocelli, J. V., & Conrey, F. R.
MA
(2009). Attentional processes in stereotype formation: A common model for category accentuation
and illusory correlation. Journal of Personality and Social Psychology, 96, 305-323.
ED
Shook, N. J., & Fazio, R. H. (2009). Political ideology, exploration of novel stimuli, and
attitude formation. Journal of Experimental Social Psychology, 45, 995-998.
PT
Sidanius, J. & Pratto, F. (1999). Social Dominance: An Intergroup Theory of Social
CE
Hierarchy and Oppression. New York: Cambridge University Press.
Smith, E. R. (1991). Illusory correlation in a simulated exemplar-based memory. Journal of
AC
Experimental Social Psychology, 27, 107-123.
Stroessner, S. J., Hamilton, D. L., & Mackie, D. M. (1992). Affect and stereotyping: The
effect of induced mood on distinctiveness-based illusory correlations. Journal of Personality and
Social Psychology, 62, 564-576.
Stroessner, S. J. & Placks, J. E. (2001). Illusory correlation and stereotype formation: Tracing
the arc of research over a quarter century. In G. B. Moskowitz (Ed.), Cognitive social psychology:
The Princeton symposium on the legacy and future of social cognition (pp. 247-259). Mahwah, NJ:
Erlbaum.
ACCEPTED MANUSCRIPT
Ideology and attitude formation 14
Footnotes
In some cases participants failed to respond to one or more questions and therefore the
T
1
2
RI
P
degrees of freedom may change accordingly.
A different sample of participants drawn from the same student population (N = 40)
SC
responded to the ideology scale and also indicated their political affiliation along a 10 cm
NU
continuum (from left-wing to right-wing). The two variables emerged to be positively correlated,
r(40) = .62, p < .001, suggesting that responses to the ideology scale are indeed related to political
3
Phi coefficients were always converted to Fisher’s z scores before analyzing them.
Three additional regression analyses were performed combining the data from the two
ED
4
MA
self-designation.
PT
studies. In addition to ideology, the study and the interaction term were entered as predictors of the
three indexes of illusory correlation. The interaction term was significant in the case of the memory
CE
measures (both ps <.05) but fell short of significance in the case of group evaluation (p = .06).
Overall, these results indicate that ideology was indeed differently related to illusory correlation in
AC
the two studies.