A group by any other name

29 Pages·1.693 MB·English

by  Hamilton et al.

This article was downloaded by:[NEICON Consortium]
On:22 October 2007
Access Details:[subscription number 781557152]
Publisher:Psychology Press
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
European Review of Social
Psychology
Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713684724
A Group By Any Other Name - The Role of Entitativity in
Group Perception
David L. Hamiltona; Steven J. Shermanb; Luigi Castellic
aUniversity of California, Santa Barbara, USA
bIndiana University, USA
cUniversity ofPadova, Italy
First Published on:01 January 2002
To cite this Article:Hamilton, David L., Sherman, Steven J. and Castelli, Luigi
(2002) 'A Group By Any Other Name - The Role of Entitativity in Group Perception',
European Review of Social Psychology, 12:1, 139 - 166
To link to this article: DOI:10.1080/14792772143000049
URL:http://dx.doi.org/10.1080/14792772143000049
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use:http://www.informaworld.com/terms-and-conditions-of-access.pdf
Thisarticlemaybeusedforresearch,teachingandprivatestudypurposes.Anysubstantialorsystematicreproduction,
re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly
forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents will be
complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be
independentlyverifiedwithprimarysources.Thepublishershallnotbeliableforanyloss,actions,claims,proceedings,
demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or
arising out of the use of this material.
7
0
0
2
er
b
o
ct
O
2
2
1 Chapter 5
4
1:
1
At:
m]
u A Group By Any Other Name-The
orti
s
n
Co Role of Entitativity in Group
N
O
EIC Perception
N
y: [
B
ed David L. Hamilton
d
a
o University of California at Santa Barbara, USA
nl
w Steven J. Sherman
o
D
Indiana University, USA
Luigi CasteUi
University of Padova, Italy
ABSTRACT
The social world consists of numerous and diverse groupings of people into
meaningful and important collectives. As perceivers, we routinely encounter
aggregates of people, some of which we endow with the property of groupness,
and others we do not. Moreover, the variety of groups is enormous, yet per-
ceivers differentiate among them and understand their properties. This chapter
discusses how and why perceivers “see” an aggregate of persons as a group, the
distinctions among different types of groups that perceivers detect, the variation
among groups in their perceived groupness or entitativity, and the consequences
that follow from perceiving a group as an entitative unit. The results of our
research program addressing these issues are summarized, and implications for
remaining unanswered questions are discussed.
Correspondence regarding this chapter should be addressed to David L. Hamilton, Department
of Psychology, University of California, Santa Barbara, CA 93106, USA.
European Review of Sociul Psychology, Volume 12. Edited by Wolfgang Stroebe and Miles Hewstone.
0 2002 John Wiley & Sons Ltd.
14 0 DAVID L. HAMILTON, STEVEN J. SHERMAN AND LUIGl CASTELLI
Committee-a group of men who individually can do nothing but as a group
decide that nothing can be done (Fred Allen).
7
0
20 Land of our birth, we pledge to thee
er Our love and toil in the years to be;
b
cto When we are grown and take our place,
2 O As men and women with our race (Rudyard Kipling).
2
41 To the family-that dear octopus from whose tentacles we never quite escape,
1:
1 nor, in our inmost hearts, ever quite wish to (Dodie Smith).
At:
m]
u All of these quotes refer to groups with which we are very familiar. All of us
sorti originate from a family context; we all have some national and ethnic identi-
n
o
C fication, and (for better or worse) we have all served on committees. To a
N
O considerable extent our social life is organized according to group-like struc-
C
EI tures, and we all are members of numerous groups. Yet, as the quotations
N
y: [ above illustrate, the range and variety of those groups is impressively broad.
B
d Indeed, the term “group” encompasses an enormous array of social struc-
e
ad tures. Not only are we members of, and participants in, groups of various
o
wnl kinds, but also we, individually and collectively, perceive and interact with
o
D groups continually in everyday life. As perceivers, we recognize that others
belong to and function in groups, and our cognizance of that fact influences
both our perceptions of and behavior toward those persons.
This chapter is concerned with the perception of groups by perceivers. A
group, at its minimum, is a collection of individuals. However, not all collec-
tions of individuals are groups. The fact that perceivers can somehow dif-
ferentiate between collections of persons that are and are not groups, and that
they can further differentiate among groups that are families, committees, and
nationalities, is perhaps trivially obvious. It is also fundamental to successful
adaptation in social life. And, for the social psychologist interested in group
perception, it also poses a number of intriguing and challenging questions, the
answers to which are not so obvious. In this chapter we raise some of those
questions and attempt to address them.
Since its beginnings, social psychology has developed massive literatures
concerned with both the nature of groups at various levels of analysis and how
they function (Levine & Moreland, 1998; Pruitt, 1998) and the perception of
groups by observers and how those perceptions guide behavior (Brewer &
Brown, 1998; Crocker, Major, & Steele, 1998; Fiske, 1998). Despite the rich
and extensive history of research on these topics, questions of how perceivers
“see” a collection of persons as a group have received little systematic
investigation.
In our work, we have focused on one important aspect of group perception,
the perceived “entitativity” or “groupness” of the group. Although the term
“entitativity” was coined by Campbell (1958) over 40 years ago, it is only in
recent years that researchers have attempted to understand how that percep-
ROLE OF ENTITATIVITY IN GROUP PERCEITION 141
tion is achieved and what differences it makes. However, this recent work has
already revealed the varieties and complexities inherent in our conception of
7
00 groups. As we shall see, differences in perceived entitativity have important
2
er consequences for various perceptions, inferences, and evaluations of the
b
cto group and its members, as well as for behaviors toward, and interactions with,
O
2 group members.
2
1
4
1:
1
At: ENTITATIVITY: THE CONCEPT
m]
u
sorti To begin our analysis, it is useful to consider the meaning of the concept of
n
Co entitativity. When he introduced the term, Campbell (1958) simply offered (in
N
O a footnote) the following: “Entitativity: the degree of being entitative. The
C
EI degree of having the nature of an entity” (p. 17). Campbell’s discussion made
N
y: [ clear that the term referred to perceiving the “groupness” of groups, and his
B
d article was effective in delineating some of the cues that perceivers might use
e
ad in detecting that an aggregate of individuals possesses that quality of group-
o
nl ness. Nevertheless, the definitional imprecision and the variety of cues on
w
o
D which such perceptions might be based has resulted in considerable latitude in
the way the term has been defined, manipulated, and measured in research.
Some authors (e.g., Brewer, Weber, & Carini, 1995; McGarty, Haslam,
Hutchinson, & Grace, 1995; Yzerbyt, Rogier, & Fiske, 1998) have focused on
perceived similarity or variability among group members as a cue to group-
ness. Others (e.g., Haslam, 1998; Yzerbyt, Rocher, & Schadron, 1997) have
emphasized a relation between group entitativity and the concept of essential-
ism in the perception of social categories. Another view (Brewer & Harasty,
1996) has related perceived entitativity to the extent to which the perceiver
has a clear prototype representing the group. Finally, some writers (e.g.,
Gaertner & Schopler, 1998 Hamilton, Sherman, & Lickel, 1998) have
stressed the interdependence, interconnectedness, and organization among
group members as an important basis for perceived entitativity. These various
approaches often reflect different, and sometimes unspecified, assumptions
about the nature of entitativity. To avoid such assumptions and any ideologi-
cal constraints they may entail, we choose to employ a neutral, generic defini-
tion of the term. We will use the term “entitativity” to refer to the perception
that an aggregate of individuals is bonded together in some way to constitute a
group.
The variety inherent in current conceptions of entitativity has brought both
advantages and disadvantages to progress in research inquiry. The disadvan-
tage, of course, is that there remains an unsatisfying lack of clarity in the
meaning of this term and especially in the limits on its referents. On the other
hand, and particularly in the early stages of a new line of inquiry, this concep-
tual liberalism is perhaps beneficial in that it can permit researchers to
142 DAVID L. HAMILTON, STEVEN J. SHERMAN AND LUIGl CASTELLI
consider and to explore numerous implications of the concept. We believe
that this has, in fact, been the case in this new research on entitativity, and in
7
00 the remainder of this chapter we hope to convey the richness and diversity of
2
er questions and issues that have been examined as a consequence. Neverthe-
b
cto less, it will soon become important that researchers come to a common under-
O
2 standing of what is and is not meant by this term.
2
1 As Campbell (1958) noted, an analysis of entitativity is not simply a matter
4
11: that some collections of groups “have it” and others do not; groups vary
At: considerably in the extent to which we perceive them as possessing this
m]
u quality. Moreover, as suggested at the outset, the term “group” itself can have
sorti a variety of meanings; it can refer to entities that vary considerably in their
n
Co properties. Do perceivers meaningfully differentiate among different types of
N
O groups, and if so, on what basis? If perceivers do recognize groups as varying
C
EI in entitativity, what difference does it make in their perceptions of those
N
y: [ groups? Finally, do different types of groups differ in the extent to which they
B
d are perceived as entitative? In this chapter, we discuss the findings of recent
e
ad research on perceiving group entitativity and attempt to use those findings to
o
nl assess our understanding of these questions.
w
o
D
ANTECEDENTS TO PERCEPTIONS OF ENTITATIVITY
Cues to Entitativity
Campbell’s (1958) discussion of entitativity was focused on the question of
how perceivers use available informational cues to infer that an aggregate of
individuals has the quality of being a group. His analysis, based on a Gestalt
approach, highlighted certain properties that, if present, might suggest that an
aggregate is a group. For example, if the individual persons are in close
proximity to each other, if they appear similar to each other, if they move
together in coordinated action and appear to have common fate (e.g., shared
goals and common outcomes), then it is more likely that they would be
perceived as an entitative unit. Although Campbell reported no research
evidence in support of his analysis, these ideas have provided useful guides to
subsequent investigators.
Following Campbell’s conceptual analysis, our research has investigated the
role of different types of stimulus information one might have about groups as
bases for perceivers’ inferences about the entitativity of groups. In a series of
studies (Lickel, Hamilton, Wieczorkowska, Lewis, Sherman, & Uhles, 2000),
we sought to determine the extent to which several stimulus properties are
predictive of perceptions of a group’s entitativity. Parallel studies were con-
ducted at the University of California, Santa Barbara (UCSB) and the Uni-
versity of Warsaw, Poland, Participants rated a diverse array of 40 groups
(e.g., a family, a committee, women, an orchestra, people at a bus stop) on a
ROLE OF ENTITATIVITY IN GROUP PERCEPTION 143
variety of rating scales. Instructions indicated that some groups seem to be
“groups” to a greater extent than others are (without suggesting why), and
7
00 participants were asked to rate the extent to which each one of the 40 groups
2
er “qualified as a group”, which we used as a measure of perceived group en-
b
o
ct titativity. Participants then rated each of those same groups on several group
O
2 properties that previous authors had suggested as cues to entitativity. These
2
41 included the extent to which members interact, the importance of the group to
1:
1 its members, the similarity among group members, the extent to which mem-
m] At: bers share common goals and common fate, the size of the group, the length
u of its history, and whether joining and leaving the group is easy or difficult.
sorti Among these group attributes, interaction, importance, common goals, com-
n
o
C mon fate, and similarity were most strongly correlated with perceived group
N
O entitativity. In contrast, group size, length of group history, and the per-
C
EI meability of group boundaries were essentially uncorrelated with entitativity.
N
y: [ These studies provided among the first empirical tests of the relative import-
B
d ance of a variety of group properties as indicators of group entitativity across
e
ad a broad spectrum of groups. Importantly, the results of these analyses, based
o
wnl on data from the UCSB and Warsaw samples, were remarkably similar (see
o
D Lickel et al., 2000, for details).
Types of Groups
Our next step was to use these rated group properties to discern whether
distinct types of groups could be identified. That is, the 40 stimulus groups
rated by participants intentionally included a very diverse array of groups,
which differed in many respects. Therefore, we considered the ratings on
these eight group characteristics to comprise a profile for each group, and
we then used these profiles to statistically identify several group types, each
of which had a distinct pattern of features. For example, intimacy groups
(e.g., family, close friends, local street gang) are small groups with high
levels of interaction; membership is important to their members, the groups
have long histories, and changing membership is not easy. Task-orienred
groups (e.g., a committee, a jury, cast of a play, people working in a
restaurant) are also relatively small and highly interactive and their mem-
bers have shared goals and common outcomes, but compared to intimacy
groups, these groups are less centrally important to their members, are of
shorter duration, and joining and leaving such groups is not difficult. In
marked contrast to these first two types, social categories (e.g., women,
Blacks, Jews, Americans) are very large groups with long histories and
relatively impermeable boundaries, and lower levels of interaction among
members. Loose associations (e.g., students at the same university, people
living in the same neighborhood, people who like classical music) are also
large groups but have very low levels of interaction, have shorter histories,
144 DAVID L. HAMILTON, STEVEN J. SHERMAN AND LUlGl CASTELLI
are not centrally important to their members, and changing membership is
relatively easy. Finally, transitory groups (people at a bus stop, people in line
7
00 at a bank) were low on all properties except that joining and leaving the
2
er groups is very easy. We then determined the average entitativity rating of
b
cto the groups classified into each group type. We consistently found that inti-
O
2 macy groups were rated highest in entitativity, followed by task-oriented
2
1 groups, social categories, and loose associations, in that order. In most
4
11: analyses, the differences between group types in perceived entitativity were
At: all statistically significant. The group types obtained in one of these analyses,
m]
u and the mean ratings of each group type on the various attributes, are shown
sorti in Table 5.1.
n
o
C
N
O Table 5.1 Average ratings of five group types on entitativity and group properties
C
EI
N Property Intimacy Task groups Social Weak social Transitory
y: [ groups categories groups groups
B
d
e
ad Entitativity 6.79 6.34 4.47 4.66 2.31
o
nl Interaction 8.22 6.57 4.21 4.64 1.92
w
o Importance 8.24 6.53 5.70 4.68 1.80
D
Goals 7.16 7.21 3.87 4.80 5.03
Outcomes 5.64 5.83 2.52 3.28 3.48
Similarity 5.56 4.84 3.35 3.83 2.25
Duration 6.22 4.34 7.94 5.06 1.48
Permeability 2.87 4.64 3.72 6.01 8.41
Size 1.98 3.57 8.02 5.05 4.35
Nore. Ratings were made on nine-point scales.
From Lickel ef al., 2000, with permission.
Participants in both of these studies also completed a sorting task in which
they were asked to sort the 40 groups into categories that reflected the same
“type” of group. Instructions were fairly non-specific and permitted sorting
on whatever criteria participants chose to use. These sorting data were then
used as a second empirical basis for identifying types of groups, as perceived
by participants. The results of these analyses, for both the UCSB and Warsaw
samples, revealed types of groups that largely paralleled those derived from
participants’ ratings of the groups on attribute dimensions. In fact, a series of
analyses empirically comparing the ratings-based and the sortings-based
structures established a very high degree of comparability between (a) the
two sets of analyses and (b) the results obtained from the two samples (Lickel
et af., 2000).
In a third study, we extended these analyses to the important case where
the participant is a member of the groups being considered. We sought to
determine whether, and how, group membership influences the perception of
group entitativity. Participants at Indiana University were asked to list 12
ROLE OF ENTITATIVITY IN GROUP PERCEPTION 145
groups to which they belonged, and subsequently were asked to rate each of
these groups on a series of attributes. These rating scales assessed the same
7
00 variables rated in the previous studies, as well as one additional attribute-the
2
er importance of membership in the group to the participants themselves. Cor-
b
cto relational analyses of these variables with ratings of groupness (entitativity)
O
2 paralleled those relations found in the earlier studies: interaction, importance
2
1 of membership, common goals, common fate, and similarity again were most
4
11: strongly related to perceived group entitativity. In this case, however, ratings
At: of the importance of one’s own membership in the group were most strongly
m]
u correlated with perceived entitativity. When the group profiles based on these
sorti ratings were analyzed to determine types of groups, the results were again
n
Co similar to the earlier analyses, with intimacy groups, task groups, social cate-
N
O gories, and loose associations evident in these results.
C
EI In sum, in this series of studies we have investigated some of the antecedent
N
y: [ cues upon which perceptions of group entitativity may be based. These studies
B
d have shown that certain group features or properties-high interaction among
e
ad group members, importance of group membership, sharing of common goals
o
nl and common outcomes, similarity among members-onsistently were most
w
o
D highly correlated with ratings of group entitativity across a broad spectrum of
groups. Moreover, structural analyses indicated that perceivers differentiate
several distinct types of groups, including intimacy groups, task groups, social
categories, and loose associations. Finally, these types of groups differed sig-
nificantly in the extent to which they were perceived as possessing entitativity.
All of these findings consistently emerged in similar analyses: (a) in three
separate studies (conducted in Santa Barbara, CA, USA; Warsaw, Poland;
Bloomington, IN, USA); (b) in analyses based on ratings of groups on de-
scriptive scales and in structural analyses based on participants’ sorting of the
groups into categories; and (c) using groups provided by the researchers and
also using groups generated by the participants to which they belonged (see
Lickel et al., 2000, for details). Thus, our findings provide evidence of gener-
ality in the correlates of perceived entitativity, and also suggest that these
types of groups reflect some fundamental distinctions in the ways people think
about social groups.
Spontaneous Differentiation of Group Types
All of the analyses reported by Lickel et al. involved tasks in which the
participants engaged in thoughtful, deliberative procedures. One might plau-
sibly wonder: do people spontaneously make these distinctions among types
of groups in their everyday perceptions, thoughts, and categorizations, even at
times and in contexts when they are not asked to rate or sort various groups
on various attributes? The answer appears to be “Yes”. We (Sherman,
Castelli, & Hamilton, in press) have recently obtained evidence that people
146 DAVID L. HAMILTON, STEVEN J. SHERMAN AND LUIGl CASTELLI
spontaneously distinguish among these group types in their normal cognitive
7 functioning.
0
0 If the group types that emerged in the Lickel et al. studies are indeed
2
er psychologically meaningful, then those types would represent cognitive struc-
b
o
ct tures that would influence the way information about social groups is spon-
O
2 taneously processed and stored in memory. We focused on how information
2
41 about individuals who are presented as members of different social groups is
1:
At: 1 organized in memory. If the defining characteristics of a group type are spon-
m] taneously extracted and used to categorize groups into different typologies,
ortiu specific information about two groups that share similar defining characteris-
s tics should be more likely to be confused, even when this information is
n
o
C unrelated to the distinct defining features.
N
O In the first experiment (Sherman er al., 2000), we adopted a recognition
C
EI memory paradigm that represents a useful, unobtrusive measure of how
N
y: [ information is encoded and stored in memory (Taylor, Fiske, Etcoff, &
B
d Ruderman, 1978). We presented participants with pictures of 60 male faces
e
d
a one time each for 3 seconds, along with a label indicating the group mem-
o
nl
w bership of the person portrayed in the picture. Six different group labels
o
D were used (10 members of each group), two for each of the three major
types of groups identified in the Lickel ef al. (2000) studies (i.e., family
members and friends for intimacy groups, co-workers and jury members for
task-oriented groups, and French and Presbyterians for social-categories).
Participants simply looked at the faces and read aloud the group identifica-
tion labels. In a subsequent recognition test, only the pictures of the 60 faces
were presented, and participants were asked to recall the original group
membership of the person.
Errors and response latencies constituted the main dependent variables.
We predicted that the highest number of errors would occur from a given
group to a group of the same type (e.g., misremembering a French person as a
Presbyterian) rather than to a group of a different type (e.g., misremembering
a French person as a jury member). In other words, “within-group-type er-
rors” should be more frequent than “between-group-type errors”. In addi-
tion, response latencies should be faster for within-group-type errors than for
between-group-type errors, because participants will be more certain of the
(mis)identifications in the former case. The accessibility of the group-type
label will make all groups of that type appear to be familiar.
Results showed significantly more within-group-type errors than between-
group-type errors. Figure 5.1 shows the findings from Experiment 1. Thus, if a
face had been labeled as a Frenchman (a social category), then participants
were far more likely to misidentify the face as a Presbyterian (another social
category) than as a family member (an intimacy group) or a jury member (a
task-oriented group). In addition, the within-group-type errors were signifi-
cantly faster than the between-group-type-errors.
ROLE OF ENTITATIVITY IN GROUP PERCEPTION 147
3.4
7
0
0
2 3.2
er
b
o
ct 3
O
2 identification
2
1 errors 2.8
4
1:
1
At: 2.6
m]
ortiu 2.4
s
n
o 2.2
C
N
O
C 2
EI Within Between Within Between Within Between
N
y: [ intimacy intimacy task task soccat soccat
B
d
de Within- vs. between group-type errors for each type of group
a
o
wnl Figure 5.1 Within-group and between-group errors as a function of group type.
o
D Reproduced with permission
A second experiment served to replicate these results, using behaviors of
different group members (as opposed to photographs) as stimuli. In the recog-
nition phase, the 60 behaviors were presented for group identification. In a
replication of the results of the first experiment, behaviors performed by
members of one group were more likely to be misremembered as having been
performed by members of a different group of the same type than as having
been performed by members of a different group of a different general type.
A third experiment investigated the organization of information in memory
by employing free recall. In particular, the sequence of free recall was used to
reveal this memory organization. If participants utilized the distinctions
among group types in the representation of information in memory, the prob-
ability of sequentially retrieving information about two groups of the same
type should be greater than the probability of sequentially retrieving informa-
tion about two groups of different types. To investigate this possibility, 60
behaviors (again, 10 behaviors for each of six groups; two groups of each
general group type) were presented, and a surprise recall task was then ad-
ministered. As predicted, recall sequences involving transitions within the
same general group type were significantly more common than recall se-
quences involving transitions from one type of group to another type. Finally,
a fourth experiment investigated the effects of our group typology on the
speed and accuracy of category membership verification. It was hypothesized
that, after individuals have learned the correct category membership of a
social target (e.g., a specific person is a Pentecostal, a social category), it will