Social Balance

(Editor:   Pamela Marek)


Consistency is a driving force behind human thinking and behavior. A variety of consistency theories relate to the idea that if our attitudes are inconsistent with each other we experience imbalance or disequilibrium (Zajonc, 1960). The theories differentially address the cause of this disequilibrium and the manner in which we attempt to restore equilibrium. Related to Gestalt concepts of good form and symmetry (Ickes & Jarvey, 1978), the earliest theory suggested that certain forms of relationships were more balanced than others. Specifically, Heider's Balance Theory (Heider, 1946) primarily focused on perceptions of relationships in the form of a triad. This triad, typically involving two people (the perceiver and another person) and an object of their attitudes, included three relationships. Heider (1958) labeled these attitudes as "sentiments," personal evaluations of other entities that underlie people's actions. A triad of attitudinal relationships is considered balanced if all three relationships are positive, or if one is positive and two are negative. These concepts of balance and imbalance are clarified by the illustrations below, adapted from

Let's say that our perceiver is Neil, the other person is Jane, and the object is Psychology. Figure 1 shows two of the four possible balanced triads. In both pictured triads, Neil has a positive attitude toward Jane. Jones (1953) reported that people rated these two types of balanced triads as more pleasant than other balanced triads (in which, for example, Neil had a negative attitude toward Jane) and as more pleasant than unbalanced triads.

Sample data image from the social balance Data experiment Sample data image from the social balance Data experiment
Figure 1

Now, let's look at two unbalanced triads. In the first unbalanced triad, Neil likes Jane, Jane likes Psychology, but Neil does not like Psychology. In the second unbalanced triad, Neil does not like Jane, but both Neil and Jane like Psychology. Jones (1953) found little difference in pleasantness ratings among the four unbalanced triads.

Sample data image from the social balance Data experiment Sample data image from the social balance Data experiment
Figure 2

To regain equilibrium in an unbalanced triad, Neil might either change his attitude toward Jane or toward psychology. However, what if Neil's attitude toward Psychology was not particularly strong? Heider's theory alone could not answer this question, because he did not address how attitude strength would influence reactions toward imbalance. In contrast, Newcomb's (1968) symmetry model did encompass attitude strength. According the symmetry model, the extent to which Neil experiences tension from the incongruity between his attitude toward Jane and his attitude toward Psychology depends on the potency of these attitudes. In their Congruency theory, Osgood and Tannenbaum (1955) also extended the ideas of balance in a manner than enabled precise predictions. They assigned numbers to represent the intensity of an attitude. Using numbers to measure attitudes permits a more precise analysis of balance. For advanced students, data from this experiment may be analyzed using this form of precise analysis.

Design & Procedures

In this experiment, you will be asked to imagine yourself in a situation in which you have one of five feelings (love, like, neutral, dislike or hate) toward another person named Bill. You will also be given information about Bill's feelings toward a person named John. Based on this information, you will be asked to rate your feelings toward John (the dependent variable). Thus, this experiment uses a 5 (Your feelings toward Bill) x 5 (Bill's feelings toward John) within-subjects factorial design, with two independent variables and one dependent variable. After four practice trials, you will complete a total of 25 trials, one for each possible combination of your feelings toward Bill and Bill's feelings toward John. On each trial, you will rate your feelings toward John on a 9-point Likert type scale ranging from -4 (Dislike) to +4 (Like).

Data Download, Format & Analyses

Data is downloadable in three formats (XML, Excel spreadsheet format, and comma delimited for statistical software packages). Figure 3 shows an excerpt from a sample Excel spreadsheet. The first five columns provide classification data (participant ID number, gender, age, the class ID number, and completion date). The remaining 25 columns provide the ratings of John, for each possible combination of the two independent variables. Your feelings toward Bill are indicated by the term before the dash; and Bill's feelings toward John are indicated by the word after the dash.

Sample data image from the social balance Data experiment
Figure 3

In a mathematical model of balance, to fully support the predictions of balance theory, your feelings toward John should be the product of your feelings toward Bill and Bill's feelings toward John. To illustrate this model, Birmbaum (2001) explained how your feelings toward Bill and Bill's feelings toward John are quantified on a scale ranging from -2 (hate) to + 2 (love). As in this experiment, your feelings toward John (the dependent variable) are scaled so that -4 indicates the strongest dislike, and +4 indicates the strongest like. Thus, as shown in Figure 4, for maximum balance, if you love Bill (+2) and Bill hates John (-2), your feelings toward John would be rated a -4 (+2 x -2). Birnbaum (2001) labels this configuration of interactions as a "Spider of Multiplication" (p. 154), providing advanced, detailed instructions on how to determine whether actual data is consistent with this multiplicative model.

Sample data image from the social balance Data experiment
Figure 4

However, a more basic analysis invo lves conducting a 5 (Your feelings toward Bill) x 5 (Bill's feelings toward John) repeated-measures analysis of variance (ANOVA), with Your feelings toward John as the dependent variable. Clearly, you would expect to find a significant interaction. To visualize this interaction, if you plotted the means of the 25 conditions, you might check to see how your functions compare with those in Figure 4. However, do not expect a perfect match; rather, look at the general pattern.


In an ingenious experimental test of balance theory using a "live" setting rather than written descriptions, Aronson and Cope (1968) used a 2 x 2 factorial design, combining a manipulation of an experimenter's attitude (pleasant or harsh) with manipulation of a supervisor's attitude toward the experimenter (pleasant or harsh). After participating in what they thought was the primary part of the experiment, participants were asked the extent to which they would be willing to help the supervisor make phone calls. Supporting predictions of balance theory, participants volunteered to make more calls if a supervisor had been harsh to a harsh experimenter or if a supervisor had been pleasant to a pleasant experiment than in the two incongruent conditions.

More recently, balance theory has been extended to apply to larger interpersonal groups (Situngkir & Khanafiah, 2004)), perceptions of groups such as the US Supreme Court (Pilialoha & Brewer, 2006), matchmaking (Chapdelaine, Kenny, & LaFontana, 1994), the connection between voters and their political parties (Ray, 1999), bargaining (Kette, 1986), and developing a comprehensive theory of self-esteem, self-concept, implicit attitudes and stereotyping (Greenwald, Banaji, Rudman, Farnham, Nosek, & Mellot, 2002). Regarding implicit attitudes, you may be interested in trying the Implicit Association Test (Race) in the States of Consciousness category of the Online Psychology Laboratory.


Aronson, E., & Cope, V. (1968). My enemy's enemy is my friend. Journal of Personality 
	and Social Psychology, 8, 8-12.
Birmbaum, M. H. (2001). An introduction to behavioral research on the Internet. Upper 
	Saddle River, NJ: Prentice-Hall.
Chapdelaine, A., Kenny, D. A., & LaFontana, K. M. (1994). Matchmaker matchmaker, can you 
	make me a match? Predicting liking between two unacquainted persons. Journal of 
	Personality and Social Psychology, 3, 83-91.
Greenwald, A. G., Banaji, M. R., Rudman, L. A., Farnham, S. D., Nosek, B. A., & Mellot, 
	D. S. (2002). A unified theory of implicit attitudes, stereotypes, self-esteem, 
	and self-concept. Psychological Review, 1, 3-25.
Heider, F. (1946). Attitudes and cognitive organization. Journal of Psychology, 21, 107-112.

Heider, F. (1958). The psychology of interpersonal relations. New York: Wiley. Reprint 
	available at

Ickes, W., & Jarvey, J. H. (1978). Fritz Heider: A biographical sketch, The Journal of Psychology, 
	98, 159-170.
Jordan, N. (1953). Behavioral forces that are a function of attitudes and cognitive organization. 
	Human Relations, 6, 273-287.

Kette, G. (1986). Attributions restore consistency in bargaining with liked/disliked partners. 
	European Journal of Psychology, 16, 257-277.

Osgood, C. E., & Tannenbaum, P. H. (1955). The principle of congruity in the prediction of 
	attitude change. Psychological Review, 62, 42-55.

Pilialoha, B. R.U., & Brewer, M. B. (2006). Motivated Entitativivity: Applying Balance theory 
	to group perception. Group Processes & Intergroup Relations, 9, 235-247.

Ray, L. (1999). Conversion, acquiescence, or delusion: The contingent nature of the party-voter 
	connection. Political Behavior, 21, 325-347.

Situngkir, H., & Khanafiah, D. (2004). Social Balance theory: Revisiting Heider's Balance 
	Theory for many agents. (Technical Report No. WPN2004BF1). Bandung Fe Institute, 
	Dept. Computational Sociology. Retrieved May 10, 2006 from Cogprints,

Zajonc, R. (1960). The concepts of balance, congruity, and dissonance. Public Opinion 
	Quarterly, 24, 280-296.