11 Festinger (1954)
A Theory of Social Comparison Processes
they had done worse than similar persons who were not attracted to the group. Those who were attracted
to the group and had scored equal to the others felt that they had done better than did similar persons
who were not attracted to the group. Thus the results of the experiment corroborate the corollary for
abilities.
Corollary VII B: The greater the relevance of the opinion or ability to the group, the stronger will be the pressure
toward uniformity concerning that opinion or ability.
The conceptual definition of relevance of an opinion or an ability to a group is not completely
clear. There are, however, some things one can state. Where the opinion or ability involved is necessary
or important for the life of the group or for the attainment of the satisfactions that push the members into
the group, the need for evaluation in that group will be strong. Groups will thus differ on what one may
call their “realm of relevance”. A group of men who meet every Friday night to play poker, and do only
this together, will probably have a narrow “realm of relevance”. The abilities and opinions for which this
group serves as a comparison will be very restricted. The members of a college fraternity, on the other
hand, where the group satisfies a wider variety of the members’ needs will have a wider “realm of
relevance”.
In spite of the conceptual unclarity which is involved it is possible to create differences in
relevance of an issue to a group which are clear and ‘unambiguous. Thus Schachter (22) created high
and low relevance conditions in the following manner. Groups which were to discuss an issue relevant
to the group were recruited specifically for that purpose. Other groups were recruited ostensibly for very
different kinds of things and on a pretext were asked to discuss the particular issue in question. They
were promised this would never happen again in the life of the group thus making this issue of low
relevance to that particular group. Schachter found, confirming Corollary VII B, that the tendency to
reject deviates was stronger in the high relevance condition than in the low relevance condition.
No other evidence bearing on Corollary VII B has been located.
Thus far we have discussed only factors which, in affecting the pressure toward uniformity,
affect all three manifestations of this pressure in the same direction. There are also factors which affect
the manifestations of pressure toward uniformity differentially. We will discuss two such factors.
Hypothesis VIII:
If persons who are very divergent from one’s own opinion or ability are perceived as different from
oneself on attributes consistent with the divergence, the tendency to narrow the range of comparability becomes
stronger.
There is evidence supporting this hypothesis with respect to both abilities and opinions. In the
previously mentioned experiment by Hoffman,
Festinger, and Lawrence (19) half the groups were told
that the three persons in the group had been selected to take the test together because, as far as could be
determined, they were about equal in intelligence. The other groups were told that one of the three was
very superior to the others. This was reported in a manner which made it impossible for either of the
subjects to suppose that he himself was the superior one. In the “homogeneous” condition the subjects
continued to compete against the paid participant who was scoring considerably above them. In the
condition where they thought one of the others was clearly superior they competed considerably less
with the paid participant and tended to compete with each other. In other words, when there was the
perception of a difference consistent with the fact that the paid participant was scoring above them, they
ceased comparison with him.
There is additional evidence on this point from level of aspiration experiments. Festinger (6)
reports
an experiment where, on an intellectual task, subjects (college students) were told they were
scoring considerably above another group which they ordinarily considered inferior to themselves (high
school students) or were told they were scoring considerably below a group which they considered
superior to themselves (graduate students). In these circumstances there is practically no effect on the
level of aspiration. Thus, the knowledge of this other group’s being divergent in a direction consistent
with the label of the group had no effect on their evaluation. It is interesting to note in this same
A Theory of Social Comparison Processes
12
experiment that if the reported direction of difference is inconsistent with the level of the group this
destroys the incomparability and the effect on the level of aspiration is very great.
The evidence concerning opinions relating to Hypothesis VIII comes from experiments reported
by Gerard (14) and Festinger and Thibaut (u). In both of these experiments
discussions were carried on
in a group of persons with a considerable range of opinion on the issue in question. In each experiment,
half of the groups were given the impression that the group was homogeneous. All the members of the
group had about equal interest in and know-‘ledge about the issue. The other half of the groups were
given the impression that they were heterogeneously composed. There was considerable variation
among them in interest in and knowledge about the problem. In both experiments there was less
communication directed toward those holding extremely divergent opinions in the heterogeneous than in
the homogeneous condition. In other words, the perception of heterogeneity on matters related to the
issue enabled the members of the groups to narrow their range within which they actively compared
themselves with others.
It is interesting, at this point, to look at the data from these two experiments in relation to
Hypothesis III which stated that the tendency to compare oneself with others decreased as the
divergence in opinion or ability increased. In both the Gerard experiment (14) and the Festinger and
Thibaut experiment (9) it was found that most communication was directed toward those whose
opinions were most different from the others. Since we have just interpreted a reduction in
communication to indicate a reduction in comparison with others, it is necessary to explain the over-all
tendency to communicate most with those holding divergent opinions in the light of Hypothesis III.
From
Hypothesis III we would expect comparison to be made mainly with those closest to
oneself. This is indeed true. The support one gets for one’s opinion is derived from those close to one’s
own. However, it will be recalled that, in the case of opinions, comparison with others who are divergent
represents a threat to one’s own opinion. It is for this reason that communication is directed mainly
toward those most divergent but still within the limits where comparison is made. This communication
represents attempts to influence them. Reduction in communication to these extreme opinions indicates
that the existence of these extreme opinions is less of a threat to one’s own opinion. In other words, one
is comparing oneself less with them. In the case of abilities we would not expect to find any such
orientation toward very divergent persons. Comparison behavior in the case of abilities would follow
very closely the simple relation stated in Hypothesis III.
Hypothesis IX: When there is a range of opinion or ability in a group, the relative strength of the three manifestations
of pressures toward uniformity will be different for those who are close to the mode of the group than for those who
are distant from the mode. Specifically, those close to the mode of the group will have stronger tendencies to change
the positions of others, relatively weaker tendencies to narrow the range of comparison and much weaker tendencies
to change their own position compared to those who are distant from the mode of the group.
Some data are available to support this hypothesis, with reference to opinions, from experiments
by Festinger, Gerard,
et al. (10) and by Hochbaum (18). In both of these experiments some persons in
each group were given the impression that the rest of the group disagreed with them while others were
given the impression that most of the group agreed with them. In both experiments there was
considerably more change of opinion among the “deviates” than among the conformers. In both
experiments there were considerably more attempts to influence others made by the conformers than by
the deviates. While there exist no adequate data relevant to the tendency to narrow the range of
comparison, corroboration is suggested in the experiment by Festinger, Gerard, et al. (10). In this
experiment it was found that the deviates actually communicated less to those holding most divergent
opinions than to those somewhat closer to their own position. The conformers showed the more familiar
pattern of communicating most to those with extremely divergent opinions in the group.
The question may also be raised as to the determinants of the extent to which the group actually
does move closer toward uniformity when pressures in this direction exist. In part, the degree of such
movement toward uniformity will be dependent upon the strength of the pressures. In part they will be