Marie Gustafsson1, Sverker Sikström2 & Torun Lindholm1
1Department
of Psychology, Stockholm University, Sweden,
Ingroup favoritism is often explained as a result of group belonging
and a salient group identity (Brewer, 1991; Tajfel & Turner, 1979;
Mullen, Brown & Smith, 1992).
We argue, in contrast, that ingroup favoritism is a cause rather than
an effect of group belonging that has evolved in selection processes
(Gustafsson & Sikström, submitted). In group formation, it is important
groups
p and to p
promote belonging
g g to the own
to discriminate between g
group. This can efficiently be done in communication by associating
ingroup pronouns to positive context, such that sentences with I and
We are combined with positive words, whereas sentences with He,
She and They arecombined with negative words (for example “I am
clever”, and “They are ugly”).
2
Department of Psychology, Lund University
In three studies, participants were instructed to generate three-wordsentences that contained 1) a personal pronoun 2) a verb and 3) an
evaluative adjective. Participants chose words from boxes as
presented in Figure 2. The presentation of the words was counterbalanced.
The pronouns were coded either as ingroup pronouns (I & We) or
g pp
pronouns ((He,, She &They).
y) The adjectives
j
were coded
Outgroup
either as positive +1, or negative -1.
Study 1 a & b tested whether ingroup favoritism is expressed
when participants generate sentences individually.
Study 2 tested whether a group condition increases the
evaluation of We in relation to I.
Study 1a – English language
Participants (N = 34, Data Collection: https://www.mturk.com)
generated sentences mirroring ingroup favoritism, such that sentences
that started with ingroup pronouns were followed by positive adjectives,
whereas outgroup pronouns were followed by negative adjectives,
F (1,33) = 23.9, p =. 001 ŋ2 = .42.
am
Study 1b – Swedish language
Also in Swedish, participants (N=34 Data Collection: Dept Psychology)
expressed ingroup favoritism in a sentence generating task, F(1,33) =
7.70, p < .01, ŋ2 = .19.
Table 1 - means and standard errors for ingroup and outgroup pronouns
Ingroup
M (SE)
.32 (.08)
.07 (.08)
Study 1a
Study 1b
Outgroup
M (SE)
-.19 (.06)
-.23 (.06)
walk
I
He
We They
She
References: Brewer, M. B. (1991). The social self - on being the same and different at the same time. Personality and Social Psychology Bulletin, 17(5),
475-482. Gustafsson, M., & Sikström, S. (submitted). Ingroup allocation model - redistributing resources to the ingroup. Mullen, B., Brown, R., & Smith, C.
(1992). Ingroup bias as a function of salience, relevance, and status - an integration. European Journal of Social Psychology, 22(2), 103-122. Tajfel, H., &
Turner, J. C. (1979). An integrative theory of intergroup conflict The social psychology of intergroup relations (pp. 33-47). Monterey, CA: Brooks-Cole.
Study 2 – favoring I or We
work
rread
ad
speak
Ingroup favoritism was found in all studies, in English (study 1a) and
in Swedish (study 1b & 2) language. Participants chose outgroup
pronouns together with negative adjectives and ingroup pronouns
together with positive adjectives. Furthermore, participants generating
the sentences individually combined I with more positive adjectives
than We (study 1 and 2), whereas participants in groups combined
positive adjectives
j
than I ((study
y 2).
)
We with more p
These findings suggest that ingroup favoritism is expressed
spontaneously in participants choice of words. A group context
increases the favoritism of the ingroup collective level (We) as
compared to the ingroup individual level (I).
act
clever
ugly terribly
good nice
bad great
Participants (N = 77, Data Collection; Central Station Stockholm)
generated sentences either individually or in groups (dyads). Also in this
study, participants generated sentences mirroring ingroup favoritism,
F (1,75) = 8.95, p = 0.04, ŋ2 = .11.
Table 1 – means and standard errors for ingroup and outgroup pronouns
Ingroup
M (SE)
.24 (.06)
Study 2
Outgroup
M (SE)
-.04 (.05)
In addition, there was a significant interaction effect, F (1,76) = 3.83,
p = .05, ŋ2 = .06, such that participants who generated sentences
individually favored I over We, whereas participants in groups favored
We over I.
.
Figure 3: We was favored over I in the group condition, whereas I was
favored over We in the individual condition.
0.45
Figure 1. The valence of pronouns in three-word-sentences,
0.4
0.5
English
Swedish
0.4
0.3
Mean valence
0.2
0.1
0
‐0.1
‐0.2
Figure 2: The task was to generate sentences from words above
words above
Mean Valence
0.35
0.3
0.25
Alone
Group
0.2
0.15
01
0.1
0.05
I
We
HeShe
They
0
‐0.05
‐0.3
‐0.4
Contact: Marie Gustafsson [email protected], +46 707 644 144.
I
We