Analyzing Social Interaction

Advances in Affect Control Theory

Lynn Smith-Lovin, David R. Heise

• Publisher: Routledge
• Year: 2016
• Print ISBN: 9780677217802
• eBook ISBN: 9781136767685
• Topic: Social Sciences
• Subtopic: Sociology
• Index: Social Sciences

First Published in 1988. This special issue of The Journal of Mathematical Sociology reports continuing work on affect control theory — a theory of social behavior that deals with role actions such as those of doctors toward patients, with deviant behaviors such as those of muggers toward victims, and with creative responses to events such as sanctioning a misbehaved child or labeling a deviant.

IMPRESSIONS FROM EVENTS^*

LYNN SMITH-LOVIN

Department of Sociology, University of South Carolina, Columbia, SC 29208

A large study of event stimuli developed new equations for describing how people react to events. Exploratory work found several new interaction terms affecting the impression formation process. To demonstrate the generality of the impression formation process across subject populations and study procedures, the results from the current study were compared to four others: two earlier studies on U.S. college undergraduates, a study of Belfast, Northern Ireland, high school students, and an Arabic study of well-educated Egyptians and Lebanese. Striking similarities in evaluation dynamics appeared in all studies. All English-speakers had similar potency and activity dynamics, while the Arabic study showed subtantial differences in the processing of these dimensions.

KEYWORDS: Impressions, Affect, Semantic Differential, Cross-Cultural

IMPRESSIONS FROM EVENTS

Events change people’s feelings about things, and this is a matter of utmost importance in everyday life. Consider the kinds of things we talk about:

“John was a friend until he insulted Mother. Then he seemed so obnoxious. That made me notice how vulnerable Mother is, too, you know?”

“I took my baby to Dr. Jones for her cold. He gave her some medication.”

“Jones! Chris, he’s a quack — malpractice suits and everything.”

“Oh no! … Hmm. It was strange the way he gave it to her, and the medicine’s not working.”

Much of our conversation is devoted to discussing events. We’re concerned with defining what occurred, who did what to whom, and how events affect our feelings. Events create new impressions of the people participating in them — both actors and objects, as in the first vignette. The context of an event also casts an aura on behaviors, like the act of medicating in the second vignette.

Feelings produced by events are the central concern in affect control theory; the theory requires an accurate, quantitative statement of how feelings form. While conversation can be exquisitely sensitive in discussing feelings, we have to go beyond verbal discourse to develop formulas describing how events change feelings from one state to another. Such formulas are used directly in affect control theory to predict how events affect people’s impressions and emotions. The formulas also are used indirectly, with mathematical manipulation, to predict what events people will create (Heise, 1977, 1978, 1979, 1985, 1986b, 1987).

This article presents the set of equations currently used in the theory. After a brief discussion of the research tradition, a large, new impression formation study is described. The equations are discussed with reference to the psychological meaning of specific terms. The final section of this article examines the generalizability of the equations by comparing them with results from other U.S. studies and two studies done with respondents from outside the U.S.A.

Impression-Formation Equations

Gollob (1968) was the first to publish an equation defining impression formation from events. He built on earlier attitude-change research which used sentences as cognitive units to describe events (Heider, 1967; Abelson and Rosenberg, 1958) and on a developing literature in impression formation and adjective combinations (Anderson, 1965; Osgood and Ferguson, 1957; Triandis and Fishbein, 1963).

Gollob constructed a sample of event sentences to represent all combinations of positive and negative attitudes toward actor, behavior, and object. For example, some of Gollob’s sentences were “the vicious man likes beggars” and “the kind man praises communists.” Subjects rated the people and behaviors first in isolation; for example, the behavior “to like someone” would be evaluated. Then subjects read each sentence and rated the actor on a good-bad scale. Subjects’ ratings for each stimulus were averaged in order to reduce measurement error; the event is the unit of analysis. The data for all of the sentences were entered into a multiple regression analysis in which the in-context mean ratings of actor (

) were predicted from the out-of-context mean ratings of actor (A_e) and behavior (B_e) along with interaction terms (like B_e*O_e).

A, B, and O refer respectively to the actor, behavior, and object in an event¹. The designator e means we are referring to evaluations — ratings on a good-bad scale assessing “attitudes”. The primed term on the left is an outcome that arises from discerning an event. The unprimed terms on the right represent various attitudes that exist before the event takes place. Equation (1) signifies that an event produces a new evaluative feeling about an actor which is predictable from several pre-event attitudes.

Specifically, the new attitude toward the actor (

) is partly a function of one’s old attitude toward the actor (A_e). In other words, attitudes about people have some stability regardless of events. The new attitude also depends on what the actor did, how good or bad the behavior was (B_e). A good act like Praising raises esteem for the actor, while a bad act like Attacking causes esteem to go down².

The multiplicative term (B_eO_e) indicates that some of the effect of a behavior is conditioned by one’s attitude toward the object person. A nice act toward a nice person produces a good impression. A bad act toward a nice person produces a bad impression. A bad act toward a bad person — negative times negative — turns positive and counteracts some of the negative impact of the behavior. A nice behavior toward a bad person detracts somewhat from the positive impression created by the nice act. As Gollob (1968) noted, the B_eO_e term fits traditional attitude consistency theories (e.g., Heider, 1946; Newcomb, 1953; Osgood and Tannenbaum, 1955; Festinger, 1957; Rosenberg, 1960; Feather, 1964, 1971), producing the same interesting predictions when considering negative attitudes. For example, Obstructing and Thief are both bad, but someone Obstructing a Thief is thought of in a more positive light.

Heise (1969b) replicated Gollob’s finding and showed that the same procedure could be used to predict feelings about behaviors and objects in events. Equations for predicting evaluations of behaviors and objects were found to be similar to those for predicting actor evaluation, though the terms were weighted differently.

Heise (1969b; 1970) also extended the approach to other dimensions of response besides evaluation. Theorists often view social relations in terms of two or more ind...

Table of contents

• Cover
• Title Page
• Copyright Page
• Table of Contents
• Editors’ Preface
• Affect Control Theory: Concepts and Model
• Impressions From Events
• The Affective Control of Events Within Settings
• Modified Social Identities: Amalgamations, Attributions, and Emotions
• Affective Bases of Likelihood Judgments
• Expectations, Intentions, and Behavior: Some Tests of Affect Control Theory
• Affect Control Theory: An Assessment