Norman Miller

Summary Statement of Research Interests

Norman Miller is a Social Psychologist whose primary research interest is in intergroup relations. He has published five books and over 175 articles on subjects including intergroup relations, attitudes, aggression, and cooperation. He is among the 50 most-cited social psychologists in social psychology textbooks. From 1990-1996, he was the editor of the Journal of Personality and Social Psychology: Interpersonal Relations & Group Processes, which is internationally recognized as the best journal in the field of Social Psychology.

At USC Dr. Miller is known for his graduate seminar in social psychology. This course, although exceptionally demanding, regularly attracts students from graduate departments throughout the University including the Annenberg School of Communications, the Graduate School of Education, Marketing, Sociology, and Exercise Science, and each semester there is a waiting list for registration. Dr. Miller, who is known for his eccentric teaching style and wry humor, emphasizes translating mastery of content into the development of interesting issues for further research.

Detailed Statement of Research Interests

Research of Norman Miller, Silberberg Professor
Dept. of Psychology, 2011

My major current research activities center on four distinct issues: Triggered Displaced Aggression; the Black Sheep effect; Complementary Projection; and Comprehensive Theory-testing in Social Psychology. For each topic, I describe in turn our background research, our ongoing 2011 work, and some indication of the practical value of our efforts.

Triggered Displaced Aggression (TDA)

Background. Displaced aggression is harm that is directed towards an innocent person, rather than the provocateur who elicited the retaliatory aggression. Our content analyses of a large sample of social psychology textbooks published since 1900, shows of brief flurry of interest in displaced aggression in the 1940’s. Since then, however, social psychology had discarded this concept. This abandonment is striking because our quantitative review (meta-analysis) of published experimental studies confirms that displaced aggression is a reliable and ubiquitous phenomenon.

We have found that when displaced aggression occurs, however, its target is frequently not totally innocent. Instead, more often than not, people displace aggression onto someone who, at a later time, instigates its displacement by providing a trivially irritating triggering behavior—as when a wife says, without discernable hostility: “Honey, you said you’d take out the garbage last night. Would you take it out now, before it’s too late?”

Triggered Displaced Aggression (TDA) is of important practical, as well as theoretical interest, because an initial moderately strong provocation (e.g., the husband is publicly, abusively, and unjustifiably chewed out by his boss,) will synergistically interact with a subsequent trivial trigger (the wife’s remark) to disjunctively escalate retaliatory aggression toward the wife. That is, when provoked at Time 1 by his boss, the magnitude of the husband’s retaliatory displaced aggression in response to the wife’s Time 2 remark will exceed the sum of the independent aggression-instigating effects of the boss’s abusive attack and the wife’s remark, were each to be experienced in isolation.

Rumination between the Time 1 provocation and the Time 2 trigger further increases TDA. It focuses attention on potential slights or abuse, causing those slights or “triggers” to be experienced as more egregious and intentionally provoking than they are. These effects occur both for provocation-focused and self-focused rumination.

Our cumulative experimental research on TDA shows that it occurs: (1) with as much as an 8 hour interval between the provocation and trigger; (2) with non-interpersonal (pain) as well as interpersonal (insult) initial provocations; (3) with different types of Time 1 interpersonal provocations (social rejection as well as insult); (4) when the triggering event is a task frustration (a ‘magic box’ that the triggering target easily opens but one which the actor cannot) as well as a trivial interpersonal chide; (5) when the same person provides both the provocation and the trigger; (6) in inter-group as well as interpersonal settings; and (7) more strongly among those intoxicated than those not.

Additionally, we have developed a personality measure of the propensity of individuals to displace aggression. Those who score highly on this Displaced Aggression Questionnaire (DAQ) engage in more domestic abuse, alcohol abuse, and road rage. They are more frequently arrested. In laboratory studies such individuals experience more cardiovascular, gastrointestinal, and respiratory symptoms. They show poor problem-solving coping, a wandering eye while married, and life-dissatisfaction. Our studies indicate that in stark contrast with these relationships uncovered by our DAQ measure, the Buss-Perry Aggression scales, which are the most widely used current measure of aggressive personality, fail to predict any of these effects.

Current work. I am currently expanding this research to study the effects of initial provocations that elicit different emotions including: sadness; the self-conscious emotions of shame, embarrassment, and guilt; as well as anger.

Practical implications. The practical implications of this work are obvious and extensive. It isolates a process whereby ordinary people behave in aggressively dysfunctional ways, remaining unaware of the source of their anti-social behavior. It shows that interpersonally close others can readily elicit displaced aggression with a casual remark (or an imposition of a slightly frustrating task) that both they themselves, as well as external observers, think of as inconsequential. An array of moderately negative interpersonal and non-interpersonal experiences can elicit the triggered displaced aggression process, and it is readily directed toward out-groups, as well as individuals. Our work highlights the important role of distraction from the initial provocation (as opposed to ruminating about it) in disrupting the displaced aggression process. Finally, it shows that an array of adverse consequences and life outcomes occur for those habitually predisposed to displace aggression.

Black Sheep and Devil Protection Effects:

Background. In-group favoritism and out-group rejection go hand in hand and this ethnocentrism very pervasively characterizes inter-group relations. Yet, in seeming contradiction to it, a spate of recent research shows that an in-group member who deviates from group or societal norms (hereafter called a ‘deviant’) is more harshly evaluated by comparison with a similarly deviant out-group member – meaning that people are more angry and punitive toward an in-group than an out-group offender. This finding has been termed the Black Sheep Effect.

Though largely ignored by Black Sheep researchers, everyday observations suggested an opposite effect. In-group members often appear to be more protective of a deviant fellow member, yet simultaneously call for especially harsh treatment of an out-group deviant – as when university administrators seek to hide the moral infractions of its own coach, but endorse harsh penalties for such acts by another university’s coach. Such commonly observed occurrences of Devil Protection effects led me to search for the circumstances under which one or the other of these opposing group processes occur.

In our quantitative review (meta-analysis) of the Black Sheep literature, aspects of (1) the degree to which a deviant’s offensive act is public and (2) the specific audience to which it is publicly known, determine which of these opposing effects will be obtained. Likewise, the nature of the audience present during in-group members’ evaluations of the deviant actor also predicts which process — Black Sheep (ostracism/ punishment of the deviant group member) or Devil Protection (shielding the deviant actor from harsh consequences) – will occur.

Current work. We are now conducting experimental laboratory studies that directly examine these audience effects on group members’ reactions to in-group and an out-group members who commits an anti-normative offensive act.

Practical implications. It is crucial to know the circumstances that cause subgroups to hide the anti-social acts of its members from public scrutiny. Likewise, it is important to understand when group members who express an independent, unique, or unpopular opinion will be rejected and ostracized (and thereby remain un-influential) – as was seen in President Kennedy’s security advisory team with respect to the Bay of Pigs fiasco. Understanding the role of audience factors in promoting Black Sheep and Devil Protection effects can help us avoid the negative consequences of each of these processes.

Complementary Projection

Background. People routinely justify their negative traits and actions. In some instances, they do so by distorting the character of others. Complementary Projection refers to the attribution of negative traits or actions to others, which, were they actually characteristic of those others, would justify ones own negative trait or behaviour. Thus, if I am characteristically a hostile, aggressive person, I might distort my view of others and (mistakenly) see them as overly bossy. And were they indeed as bossy as my misperception suggests, in accord with the tit-for-tat rule that governs most social interaction, I would thereby be justified in behaving toward them in the hostile, aggressive manner that I do.

Strikingly, despite a few sporadic and seemingly supportive studies, the most comprehensive and methodologically sophisticated study of Complementary Projection failed to evidence any empirical support for it. We have now shown, however, that the arousal of an acute and relevant motivational state is a necessary precursor for Complementary Projection to be exhibited. For example, when people are momentarily frightened (an acute motivational state), those who are temperamentally fearful will attribute more aggressiveness to others than will those who are not temperamentally fearful. When fright has not been aroused, however, fearful and unfearful people do not differ in their levels of Complementary Projection (viz., the attribution of aggressiveness to others is not affected by the degree to which one is temperamentally a frightened person).

Applied to inter-group contexts, Complementary Projection explains how an individual’s trait can lead to the attribution of a specific negative stereotype onto out-group individuals. In stereotyping research, Whites characteristically see Blacks as threatening. In two studies, we showed that more fearful Whites exaggerated the aggressiveness of Blacks, whereas their perception of the aggressiveness of Whites remained unaffected. In other words, chronically fearful Whites justify their own general fearfulness by exaggerating the degree to which Blacks are aggressive A third study again pointed to the critical role of an acute motivational state. Only when in a competitive situation with business students did stingy LAS students justify their own stinginess by projecting its complement, greediness, onto business students. Non-stingy LAS students showed no such effect.

Current work. We are currently extending this research by examining other complementary trait pairs. For instance, when the insecurity of temperamentally insecure persons is aroused, the greater the degree to which they are typically prying in their interpersonal relationships, the more they see others as being secretive. That is, they use the Complementary Projection of secretiveness to justify their own habitually prying behaviour.

Practical implications. As discussed, an important practical consideration concerns the use of Complementary Projection in intergroup settings. Yet, if group member (e.g. LAS students) can be made aware of the ways in which one of their own characteristics (e.g. stinginess) cause a misperception of out-group members’ (e.g.business students’) greediness and thereby reinforce their negative stereotype of that outgroup, the disruptive effects on mutually effective inter-group bargaining can be constrained. The Complementary Projection process, however, is not constrained to its application to group stereotypes. One can readily see that the relations between a husband and his mother-in-law can hardly be commodious when the mother-in-law is busy justifying her own worst traits by seeing them as merely responsive to aspects of her son-in-laws “bad character.”

Theory Testing in Social Psychology:

Theory testing is central to psychological science in general and social psychology in particular. Almost all theory testing in social psychology is constrained to a critical experiment methodology wherein 2 or 3 key variables that test an important proposition relevant to the theory are experimentally manipulated. Typically the validity of the entire theory rests on the confirmation of that one hypothesis, often in a 2 by 2 factorial experiment. Sometimes, researchers can select an independent variable for which two distinct theoretical accounts make opposing predictions. For example, Festinger and Carlsmith confirmed the Cognitive Dissonance Theory hypothesis that low justification—$1, by comparison with a $20 payment for stating that a boring task was interesting—would more strongly change participants’ attitude about the degree to which the boring task was in fact truly interesting. The Hovland learning–theory–based hypothesis made the opposite prediction. From this critical experiment approach to theory testing, the theory of the Hovland school dies and dissonance theory survives.

For over 20 years I have attempted to overcome some of these limitations on comprehensive theory testing by expanding classic meta-analysis methodologies in a theory testing meta-analytic synthesis approach. This new procedure combines classic meta-analytic tools with a ‘judged variables’ methodology wherein judges (undergrads or graduate students) read the method section of each article included in the meta-analysis to estimate the levels of an array of potential key mediating and moderating variables (cognitive, emotional, and motivational states) that the original experimental participants may have experienced when serving in it. And far more often than not, the judges are thereby providing information about variables that were never systematically explored in prior tests of the relevant theory and almost never measured consistently across the studies subjected to meta-analysis. Application of this procedure adds information about numerous additional variables that can be combined with the variables originally manipulated or measured in the studies comprising the meta-analytic database to sort out when and why a particular effect reliably occurs. To my knowledge, other than the application of computational modeling procedures (see the work of Steve Read in our social psychology program), there is no known procedure other than our theory testing meta-analytic synthesis approach that can simultaneously test the multiple propositions of an elaborate social psychological theory of human behavior and thereby evaluate the validity of that theory.

In our first application of theory testing meta-analytic synthesis procedures, we focused on theories that attempted to explain the relationship between experimentally induced negative moods and increased helping behavior. The most striking feature of this work was its disconfirmation of the richest and most well developed theoretical account at the time—Cialdini’s negative state relief model. A key postulate of this model is that sadness is the critical emotional state that underlies the consistently observed experimental finding that negative mood increases helping. Our application of our theory testing meta-analytic synthesis procedure showed instead that guilt and perceived responsibility are the critical underlying subjective states, not sadness. In response to Cialdini’s published criticism of our work, in a second Psychological Bulletin article we reported 17 additional tests of his model, all based on our original data set. None provided any support for the negative state relief model, and several produced outcomes opposite to those predicted by it. Since then we have published over ten articles that have used this theory testing meta-analytic synthesis methodology to examine a broad array of other social psychological phenomena, including such diverse topics as aggression (Marcus-Newhall, Pedersen, Carlson, & Miller, 2000), experimentally induced stress and cortisol/immune responses (Denson, Spanovic, & Miller, 2009), crossed categorization (Urban & Miller, 1998) and alcohol effects (Ito, Miller, & Pollock, 1996).

Leon Festinger’s Theory of Cognitive Dissonance (1957) is arguably the single most important social psychological theory, rivaled perhaps only by attribution theory (for which new research is now diminishing). Festinger’s theory has produced so much attention and accumulated experimental evidence that the term cognitive dissonance is now sprinkled into the lexicon of current TV news commentary analysts (e.g.: Chris Mathews and Bob Schieffer).

In previous applications of theory testing meta-analytic synthesis, we had relied on correlational, partial correlation, and regression analyses. In our application of it to cognitive dissonance theory, however, we applied the more powerful and elaborate procedures of structural equation modeling (SEM) to examine the shared and unique contributions of the judged variables to the prediction of dissonance effect sizes; (an effect size reflects the difference between the outcomes of the experimental and control group of each study in the meta-analysis

— i.e. the degree too which the manipulation of cognitive dissonance was effective). These SEM analyses were performed within and across five key dissonance research paradigms: induced compliance, insufficient justification, disconfirmed expectancies, selective exposure, and free choice. We tested multiple models that corresponded to each of the major versions of cognitive dissonance theory, namely, Festinger’s original theory; a modified Festingerian model; Aronson’s “threat to self” model of dissonance theory; and a model focusing on consequences of the dissonance-inducing act, as initially developed by Collins and Hoyt and subsequently promoted by Cooper and Fazio.

Our results, as seen in multiple SEM analyses, showed that none of these models supported Festinger’s notion that the unique variance of felt discomfort – the specific subjective state postulated to underlie cognitive dissonance — explains dissonance effects. Although discomfort was correlated with effect sizes, when we controlled for the effect of guilt there was no effect of discomfort. Consistent with our conceptualization of guilt as the key motivational component of dissonance theory, guilt predicted dissonance effect sizes, irrespective of which SEM model was tested. A post hoc theory that integrated the role of guilt with pre-existing dissonance theories is stronger than any of the pre-existing theories in isolation. [This major work was just published in European Review of Social Psychology, 2011, 22:1, 36-113)].

Current work. Having toppled the basic premise of what many believe to be the most important social psychological theory of modern social psychology, we are now in the process of replicating the evidence presented in this newly published monograph with additional judged mediating variables that we had not previously explored, as well as with an expanded sample of studies within each of our previously explored dissonance paradigms. In addition, we are doing so with a different procedure for assessing the magnitude of dissonance produced within each experimental study entered into our judged variable meta-analytic synthesis approach to comprehensive social psychological theory-testing.

Practical implications. The most important goal of science is to test its theories. Yet, other than computational modeling , social psychology lacks any strong tool for doing so. The methodology of theory testing meta-analytic synthesis provides such a tool. Its components, however, have not been invented by me. Meta-analysis is now well accepted as a procedure for quantitatively integrating or summarizing what is currently and reliably known about a relationship between two variables. Likewise, the use of judges to bootstrap information is widely accepted in psychology. For instance, the entire field of Decision Analyses, as pioneered by USC’s Ward Edwards (former director of the Social Science Research Institute) rests on human judgment. Finally, Structural Equation Modeling is now a widely used statistical tool for simultaneously examining a testing theoretically postulated relationships among an array of variables. The simultaneous use of these three procedures provides a powerful theory testing capability that though heretofore not used or promoted, will exponentially advance our theory-testing capability within social psychology.

Conclusion

My ongoing work addresses major issues in social psychology. It concerns processes that are distinctive, in that they have not previously been well addressed – if addressed at all – and that will be important to future development of social psychological science.

Conference Presentations

• The Displaced Aggression Questionnaire and health. , Silverwind Conference on Social Interdependence. Mpls. MNTalk/Oral Presentation, Invited, Mpls. MN, 2006-2007
• The Effect of Intergroup Goal Structure on Projection: Threat Produces Complementary Projection. Society for Personality and Social Psychology (APA Division 8) , Annual Convention, Memphis, TN.Poster, Memphis, TN, 2006-2007

Book Chapters

• Miller, N., Collins, B. E. (2010). A new method for theory testing in social psychology: The case of dissonance. In: Most underappreciated: 50 Prominent social psychologists talk about hidden gems. New York: Oxford.
• Miller, N., Spanovic, M., Stenstrom, D. (2010). The effects of crossed categorization in intergroup interaction. The Psychology of Social and Cultural Diversity pp. 193-233. Wiley-Blackwell.
• Denson, T. F., Pedersen, W. C., Ronquillo, J., Miller, N. (2008). Trait displaced aggression, physical health, and life satisfaction: A process model. Hauppauge, NY: Nova Science Publishers.

Encyclopedia Article

• Miller, N., Yang, L. (2010). Aggression. Encyclopedia of Peace Psychology. (D. J.Christie, Ed.).9. Encyclopedia of Peace Psychology. Hoboken, New Jersey: Wiley-Blackwell.
• Miller, N. (2010). Desegregation: Encyclopedia of Group Processes and Intergroup Relations. (J. L. Levine & M. Hogg, Ed.).199-204. Thousand Oaks, CA: Sage.

Journal Article

• Miller, N. (2009). Lashing out after stewing over public insults: The effect of public provoctaion, provocation intensity, and rumination on triggered displaced aggressionn of public provocation , provocation intensity, and. Aggressive Behavior.
• Stratton, G., Miller, N., Lickel, B. (2011). The integrated deviant protection/rejection model: A meta analytical explanation for the varying treatments of ingroup deviants. Group Dynamics.
• Kenworthy, J. B., Miller, N., Collins, B. E., Read, S. J., Earleywine, M. (2011). A trans-paradigm theoretical synthesis of cognitive dissonance theory: Illuminating the nature of discomfort. European Review of Social Psychology. Vol. 22, pp. 36-113.
• Pedersen, W. C., Denson, T. F., Goss, R. J., Vasquez, E. A., Kelley, N. J., Miller, N. (2011). The impact of rumination on aggressive thoughts, feelings, arousal, and behavior. British Journal of Social Psychology. Vol. 50, pp. 281-301.
• Denson, T. F., Spanovic, M., Aviles, F. E., Pollock, V. E., Earleywine, M., Miller, N. (2010). The effects of accute alcohol kintoxication andself-focused rumination on triggered displaced aggression. Journal of Aggression, Maltreatment & Trauma..
• Denson, T. F., Spanovic, M., Aviles, F., Pollock, V. E., Earleywine, M., MIller, N. (2010). The effects of acute alcohol intoxication and self-focused rumination on triggered displaced aggression. Journal of Aggression, Maltreatment & Trauma. 20, 128-147.. Vol. 20, pp. 128-147.
• Pedersen, W. C., Denson, T. F., Goss, R. J., Vasquez, E. A., Kelley, N. J., Miller, N. (2010). The impact of ruminationon aggressive thoughts, feelings, arousal, and behavior. British Journal of Social Psychology.
• Denson, T. F., Spanovic, M., Miller, N. (2009). Stress and specificity: Reply to G. Miller. Psychological Bulletin.
• Denson, T. F., Spanovic, M., Miller, N. (2009). Cognitive appraisals and emotions predict cortisol and: immune Responses: A meta-analysis of acute laboratory social stressors and emotion inductions. Psychological Bulletin.
• Ensari, N., Stenstrom, D. M., Pedersen, W. C., Miller, N. (2009). The role of integral affect and category relevance on crossed categorization. Group Dynamics.
• Miller, N., Pedersen, W., Bushman, B., Bonacci, A., Vasquez, E. (2008). Chewing on it can chew you up: Effects of rumination on triggered displaced agggression. Journal of Personality and Social Psychology/American Psychological Association. Vol. vol. 88, pp. 969-983.
• Miller, N., Ensari, N. (2008). Prejudice and intergroup attributions: The role of personalization and performance feedback. Group Processes and Intergroup Relations. Vol. vol. 8, pp. 391-410.
• Stenstrom, D. M., Lickel, B., Denson, T. F., Miller, N. (2008). The roles of ingroup identification and outgroup entativity in intergroup retribution. Personality and . Vol. 34, pp. 1570-1582.
• Pedersen, W. C., Bushman, B. J., Vasquez, E. A., Miller, N. (2008). Kicking the (barking) dog effect: The moderating role of target attributes on triggered displaced aggression. Personality and Social Psychology Bulletin. Vol. 34 (1382-), pp. 1395.
• Denson, T. F., Aviles, F., Pollock, V., Earleywine, M., Vasquez, E. A., Miller, N. (2008). The effects of alcohol and the salience of aggressive cues on triggered displaced aggression. Aggressive Behavior. Vol. 34, pp. 25-33.
• Urada, D. I., Stenstrom, D. M., Miller, N. (2007). Crossed categorization beyond the two-group model. Journal of Personality and Social Psychology. Vol. 92, pp. 649-664.
• Vasquez, E. A., Bartsch, V. O., Pedersen, W. C., Miller, N. (2007). The impact of aggressive priming, rumination, and frustration on prison sentencing. Aggressive Behavior. Vol. 33, pp. 477-485.
• Miller, N., Vasquez, E. A., Ensari, N., Pedersen, W. C., Tan, Y., R., M. N. (2007). Personalization and differentiation as moderators of triggered displaced aggression towards out-group targets. European Journal of Social Psychology. Vol. 33, pp. 297-319.
• Miller, N., Denson, T., Pedersen, W. (2006). The Displaced Aggression questionnaire. Journal of Personality and Social Psychology/American Psychological Association. Vol. vol. 90, pp. 1032-1051.
• Miller, N., Kenworthy, J., Canales, C., Stenstrom, D. (2006). Explaining the effects of crossed categorization on ethnocentric bias. Multiple social categorization: Processes, models and applications/Taylor and Francis. pp. p.160-188.
• Miller, N., Vasquez, E., Denson, T., Stenstrom, D., Pedersen, W. (2005). The moderating effect of trigger intensity on triggered displaced aggression. Journal of Experimental Social Psychology. Vol. vol. 41, pp. pp. 61-67.
• Miller, N., Othmer, S., Pollock, V. (2005). The subjective response to neurofeedback. Mind altering drugs: The science of subjective experience/Oxford University Press. pp. p.345-366.

• American Psychological Association: Kurt Lewin Memorial Award, SPSSI Division 9, 2000
• American Educational Research Association: Outstanding Contribution to Cooperative Learning Award, 1999
• American Psychological Association: Fellow, Division 8, 1996
• Fellow, Western Psychological Association, 1993
• American Psychological Society, Charter Fellow, 1987
• USC Endowed Professorship, Mendel B. Silberberg Professorship in Social Psychology, 1974/09/01
• Fellow, Division 9 APA, 1971
• Fellow, NSF Social Psychophysiology Summer Program, U of Iowa, Spring 1986
• Fulbright Award, Research Fellowship, 1984-1985
• Guggenheim Fellowship Recipient, Research Fellowship, 1984-1985
• James McKeen Cattell Fellow, 1978-1979
• John Randolph Haynes & Dora Haynes Foundation Fellow, 1974-1975
• NIMH Special Research Fellowship, 1968-1969

Other Advisement or Time Devoted to Students

• Provided guidance and supervision to 18 undergraduates on research and topics of interest., 2008-2009
• Kim Gossweiller: Chair of Undergraduate Honors Thesis, 2008-2009
• AdamTakeda:Chair, Undergraduate HonorsThesis, 2008-2009
• Laboratory Research Seminar: Weekly: Wed., at 2PM – 4PM, 2006-2007
• Laboratory Research Seminar: Weekly: Wed., at 2PM – 4PM, 2006-2007

Editorships and Editorial Boards

• Reviewer, Aggressive Behavior(No of articles reviewed – 2), 2006-2007
• Reviewer, European Journal of Social Psychology (No of articles reviewed – 1), 2006-2007
• Reviewer, Group Processes & Intergroup Relations (No of articles reviewed – 1), 2006-2007
• Reviewer, Journal of Experimental Social Psychology (No ofarticles reviewed – 3), 2006-2007
• Reviewer, Personality and Social Psychology Bulletin (No of reviewed articles reviewed – 2), 2006-2007
• Editor, Journal of Personality and Social Psychology: Interpersonal Relations & Group Processes, 1990 – 1996

Reviewer for Publications

• J. Exp. Social Psychology: 5 submissions, Reviewed 5 submitted manuscripts, 2009-2010