Chapter Two from

5                                                                                                                         Festinger (1954) A Theory of Social Comparison Processes                     

enough, then there should be some attraction to groups where others are relatively close with respect to 

opinions and/or abilities. There are data confirming this for both opinions and abilities. 

 Festinger, 

Gerard, 

et al. (mo) report an experiment in which after each person had written down 

his opinion on an issue he was handed back a slip of paper presumably containing a tabulation of the 

opinions in the group. Some in each group were thus given the impression that most of the others in the 

group held opinions close to their own. The rest were given the impression that most others in the group 

held opinions quite different from their own. After the experiment they were each asked how well they 

liked the others in the group. In each of the eight different experimental conditions those who thought 

that the others held divergent opinions were less attracted to the group.

2

 

 

The previously mentioned experiment by Dreyer (5) has as one of its main purposes the testing 

of this derivation in connection with abilities. He used a “level of aspiration” situation and falsified the 

scores he reported to the subjects so that some thought they were scoring very far above the group, some 

thought they were scoring very far below the group, while others thought they were scoring about at the 

same level as the average of others like them. After each trial they were asked whether they wanted. to 

continue for another trial or whether they would prefer to stop. The reasoning was that if those scoring 

well above or well below the group average were not able to evaluate their ability accurately, the 

situation would be less attractive to them and they would stop sooner. On the average, those scoring 

very much above the group stop after the fifth trial, while those scoring below or at the average of the 

group stop after the ninth trial.

3

 There is no difference between those scoring at and those scoring well 

below the average of the group.

 

The derivation in the case of abilities seems confirmed for deviation 

from the group in one direction then but not in the other. This is probably due to the presence of another 

pressure which we shall discuss in detail later, namely, the value placed in our culture on being better 

and better with the result that the subjects scoring below the group wanted to, and felt that they might, 

improve and achieve comparability with the group average. 

 

This result from the experiment by Dreyer (5) is also corroborated in the previously mentioned 

experiment by Hochbaum (18). It will be recalled that half the subjects were made to feel that their 

ability in judging situations of the kind they were to discuss was extremely good and very superior to the 

abilities of the others in the group. The other half of the subjects were made to feel that their ability was 

poor and considerably worse than the ability of the others in the’ group. At the end of the experiment all 

the subjects were asked whether, if they returned for another session they would like to be in the same 

group or a different group. Of those who felt they were very much above the others in the group, only 38 

per cent wanted to return to the same group. Of those who felt that they were considerably inferior to the 

others, 68 per cent wanted to return to the same group. 

 

With the qualification concerning the asymmetry with regard to abilities the derivation may be 

regarded as confirmed. We will discuss the unidirectional drive upwards for abilities, which produces 

the asymmetry, in more detail later. 

Derivation D (from I, II, III): The existence of a discrepancy in a group with respect to opinions or abilities will lead 

to action on the part of members of that group to reduce the discrepancy. 

 

We have stated in Hypotheses I, II and III and in the corollaries to these hypotheses that there is a 

drive to evaluate accurately one’s opinions and abilities, that this evaluation is frequently only possible 

by comparison with others and that the comparison tends to be made with others who are close to 

oneself on the particular ability or opinion in question. This implies that the drive to evaluate one’s 

ability or opinion will lead to behavior which will produce for the person a situation where those with 

whom he compares himself are reasonably close to him, in other words, there will be action to reduce 

discrepancies which exist between himself and others with whom he compares himself: 

 

Before we can discuss the data relevant to this derivation it is necessary to point out two 

important differences between opinions and abilities which affect the behavioral manifestations of the 

action to reduce discrepancies. We will state these differences in the form of hypotheses. 

---

A Theory of Social Comparison Processes 

   

              6

 

Hypothesis IV: There is a unidirectional drive upward in the case of abilities which is largely absent in opinions. 

 

With respect to abilities, different performances have intrinsically different values. In Western 

culture, at any rate, there is a value set on doing better and better which means that the higher the score 

on performance, the more desirable it is. Whether or not this is culturally determined, and hence 

culturally variable, is an important question but one with which we will not occupy ourselves here.

1

 

With respect to most opinions, on the other hand, in the absence of comparison there is no 

inherent, intrinsic basis for preferring one opinion over another. If we thought of opinions on some 

specific issue as ranging along a continuum, then no opinion in and of itself has any greater value than 

any other opinion. The value comes from the subjective feeling that the opinion is correct and valid. 

Hypothesis V: There are non-social restraints which make it difficult or even impossible to change one’s ability. These 

non-social restraints are largely absent for opinions. 

 

If a person changes his mind about something, deserts one belief in favor of another, there is no 

further difficulty in the way of consummating the change. It is true that there are sometimes 

considerable difficulties in getting someone to change his mind concerning an opinion or belief. Such 

resistance may arise because of consistency with other opinions and beliefs, personality characteristics 

that make a person lean in one direction or another and the like. But the point to be stressed here is that 

once these resistances are overcome, there is no further restraint which would make it difficult for the 

change to become effective. 

 

There are generally strong non-social restraints, however, against changing one’s ability, or 

changing one’s performance which reflects this ability. Even if a person is convinced that he should be 

able to run faster or should be more intelligent, and even if he is highly motivated to improve his ability 

in this respect, there are great difficulties in the way of consummating the change. 

 

We may now examine the implications of Derivation D. Considering Hypothesis IV is clear that 

the action to reduce the discrepancy which exists is, in the case of opinions, a relatively uncomplicated 

pressure towards uniformity. When and if uniformity of opinion is achieved there is a state of social 

quiescence. In the case of abilities, however, the action to reduce discrepancies interacts with the 

unidirectional push to do better and better. The resolution of these two pressures, which act 

simultaneously, is a state of affairs where all the members are relatively close together with respect to 

some specific ability, but not completely uniform. The pressures cease acting on a person if he is just 

slightly better than the others. It is obvious that not everyone in a group can be slightly better than 

everyone else. The implication is that, with respect to the evaluation of abilities, a state of social 

quiescence is never reached. 

 

Competitive behavior, action to protect one’s superiority, and even some kinds of behavior that 

might be called cooperative, are manifestations in the social process of these pressures which do not 

reach quiescence. We shall now elaborate this further in considering the specific nature of the social 

action arising from pressures toward uniformity. There are three major manifestations of pressure 

toward uniformity which we shall list below together with the relevant data. 

Derivation D1: When a discrepancy exists with respect to opinions or abilities there will be tendencies to change 

one’s own position so as to move closer to others in the group. 

Derivation D2: When a discrepancy exists with respect to opinions or abilities there will be tendencies to change 

others in. the group to bring them closer to oneself: 

 Considering 

Hypothesis V in relation to the above two subderivations we can see that a 

difference is implied between the resulting process for opinions and for abilities. Since opinions are 

relatively free to change, the process of changing the positions of members of a group relative to one 

another is expressed in action which is predominantly socially oriented. When differences of opinion 

                                                 

1

 There is some evidence, for example, that among the Hopi Indians this preference for better performance is absent (2) 

---