Halo Effects in Marketing Research: Review and Prognosis

Neil E. Beckwith, University of Pennsylvania
Harold H. Kassarjian, University of California, Los Angeles
Donald R. Lehmann, Columbia University
ABSTRACT - The tendency in rating an object on a particular attribute to be influenced by a general impression is well known. Some of the early psychological studies of this halo effect are reviewed along with more recent work in marketing and consumer behavior.
[ to cite ]:
Neil E. Beckwith, Harold H. Kassarjian, and Donald R. Lehmann (1978) ,"Halo Effects in Marketing Research: Review and Prognosis", in NA - Advances in Consumer Research Volume 05, eds. Kent Hunt, Ann Abor, MI : Association for Consumer Research, Pages: 465-467.

Advances in Consumer Research Volume 5, 1978      Pages 465-467

HALO EFFECTS IN MARKETING RESEARCH: REVIEW AND PROGNOSIS

Neil E. Beckwith, University of Pennsylvania

Harold H. Kassarjian, University of California, Los Angeles

Donald R. Lehmann, Columbia University

ABSTRACT -

The tendency in rating an object on a particular attribute to be influenced by a general impression is well known. Some of the early psychological studies of this halo effect are reviewed along with more recent work in marketing and consumer behavior.

INTRODUCTION

Consumers' ratings of a brand (or other object) on a set of dimensions (attributes or traits) are commonly used in consumer research. Examples include image (i.e., profile) studies, advertising effects studies, and product attribute positioning studies. However, an individual's ratings on an attribute may be determined by many other variables besides cues directly relating to the particular attribute.

In a recent Federal Trade Commission case, consumers' perceptions of a brand of bread were at issue. A complaint of the FTC had alleged that ITT Continental Baking Company had falsely represented Wonder Bread to be superior in nutrition to competing brands of white bread.

It was agreed on both sides that some percentage of the public believed that Wonder Bread stood out in nutrition. One issue was the extent to which specific advertising about the nutritional quality of the brand had influenced these beliefs. The bakery company claimed that the nutritional beliefs about Wonder Bread were influenced by consumers' generally favorable overall attitudes toward the brand (including familiarity, price, and higher levels of advertising) and that this positive overall feeling about the brand in turn influenced the belief that Wonder Bread was nutritionally superior. (Eventually this particular charge was dismissed on other grounds.)

The tendency in rating an object on a particular attribute to be influenced by a general impression is widely recognized.

CONCEPT

Rating responses of an object on a set of attribute dimensions is generally determined as a function of several related aspects. First, it is generally held that some adjustment process exists whereby overall attitude and beliefs about the object on the attributes influence each ether, and the individuals' beliefs become adjusted as a result of the overall attitude and other beliefs. This process is neither good nor bad; it exists and must be accounted for.

Second, rating response about the object is influenced not only by true beliefs but also by a variety of measurement effects. This means that a respondent, when rating an object on an attribute, may be responding to stimuli rather than his true belief.

DEFINITIONS

Interestingly, within the research community substantial disagreement exists concerning the definition, importance, circumstances, and even the existence of a halo effect.

Tiffin and McCormick (1965) consider the halo effect to be the domination of all other traits by one particular trait. Krech, Crutchfield, and Bellachey (1962), in a similar vein, point out that if a person is well liked he will be rated as being very high or positive on all other specific traits. If a person is disliked, perceptions of even his most positive attributes will be downgraded. The perception of the person is influenced by the overall feeling toward that person.

English (1934) many years ago provided a rather general definition of the halo effect as the "tendency in rating to be influenced by general impression or attitude when trying to judge separate traits." This view is broader than that of Krech et al, or Tiffin and McCormick in that haloing is seen as an individual phenomenon and the source of bias is the individual's general (overall) impression or attitude, and includes the rating of any objects.

From a more technical point of view, it has also been common to view the halo effect as the excessive partial correlation between belief ratings (e.g., Symonds, 1925). However, some "natural" correlation between beliefs should be expected even without any halo effect. It is difficult to partition the correlation between the halo and natural components (Bingham, 1939). The Federal Trade Commission adopted a definition in its final opinion in the Wonder Bread case which included three sources of the halo effect: (1) widespread usage, (2) generalized advertising claims, and (3) claims stressing a different attribute (Docket 8860, p. 13).

Related Theories

Various cognitive balance theories (individuals' attempts to maintain beliefs which are consistent with each other) also suggest the existence of halo effects. Heider (1958) points out that if an individual likes both person A and object X, he would expect that his friend, person A, would also like object X. If not, the individual would experience cognitive imbalance--an uncomfortable state. "How can my friend whom I like dislike object X or politician Y or a third person. Either object X is not as good as I thought it to be, or my friend is crazy and not quite as likeable as I thought him to be" (see, e.g., Abelson et al, 1969, Feather, 1964; and Rosenberg et al, 1960).

Festinger (1964) calls this imbalance dissonance, again an uncomfortable state that must be corrected. In this unconscious "correction" process the fact that feelings toward one attribute or person or product can systematically influence another object or attribute is very similar to halo effect from some points of view. Nice people have nice attributes, and less nice people have less nice attributes (see, e.g., Venkatesan, 1973; and Wickland and Brehm, 1976).

As can be seen we have already confused the halo effect defined as "the overall impression influencing specific attributes" with popularity, familiarity, etc. Hence it may be worthwhile to examine what we wish to call halo-like effects.

HALO-LIKE EFFECTS

A number of other sources of rating bias are similar in spirit to the halo effect, but are actually distinct manifestations.

Other Attribute Effects

Belief about an attribute could possibly influence an individual's rating on another attribute beyond its effect via the overall attitude component. For example, if gasoline mileage was totally irrelevant to some individual's overall attitude, it might still influence his rating of miles per fill-up. Many studies have confirmed that the physical attractiveness of a person influences other persons' ratings of that individual on many attributes, as well as influencing the overall attitude (e.g., Miller, 1970).

Peer Attitudes

An individual's belief response may also be biased by his understanding of his own local peers' beliefs about the object on the attribute or overall attitude. Such peer group conformity influences have been considered by Bourne (1956), Sherif and Sherif (1964), and Siegel and Siegel (1957) among others.

Popularity

Individuals may also be biased by their understanding of the general public's beliefs or overall attitudes about the object, as well as their own local peer groups. Individuals may favorably bias their ratings for well known brands and unfavorably bias ratings for less popular products, for example.

Familiarity

Similarly, individuals may systematically influence their ratings differently for more familiar objects than for less familiar objects. The direction (sign) of such influence is difficult to prejudge, since it would seem to depend upon the general satisfaction level realized during familiarization with the object as compared to the individual's expectations of unfamiliar objects. Koltuv (1962) concluded that the magnitude of halo effects decreases with increasing familiarity; although, James and Carter (1977) did not find a similar result in their recent analysis of 14 students' location preferences.

Cross Sectional Estimation

One of the difficulties in assessing the halo effect is estimating the magnitude of the effect. The most critical problem is estimating the individual's true beliefs.

Cross Sectional Bias

"Halo effect" generally denotes the bias of each individual's belief response. However, measurement problems also exist. Huber and James (1976) pointed out that simply averaging the belief responses across individuals (as in Beckwith and Lehmann, 1975) does not provide unbiased estimates of the underlying locations or values. This is because the objects or alternatives will usually have different fractions of the respondents favoring them. The average of belief responses for the popular objects should be more favorably biased than for less popular objects since more individuals will favorably bias the rating for popular objects than they will for unpopular objects. Researchers have attempted to attenuate this source of bias by separately comparing users and nonusers across brands, for example. Beckwith and Kubilius (1977) have attempted to estimate the true locations of objects to remove this bias. Also see Bemmaor and Huber (1977) for a discussion of the necessity for simultaneous estimation of belief and attitude components.

CORRELATES OF THE LEVEL OF HALO

Thorndike (1920) originally concluded that the magnitude of halo effects seems to be surprisingly large. Subsequent investigators have attempted to determine which circumstances seem to involve only small or negligible halo effects, and which seem to encourage larger halo effects. There is substantial disagreement among published studies concerning even which circumstances encourage haloing. Thus, the following should be considered to be candidate, rather than proven, circumstances which tend to evidence more haloing:

1. Low familiarity with the objects (Koltuv, 1962).

2. Ambiguous or subjective attributes (James and Carter, 1977; Beckwith and Lehmann, 1975).

3. High perceived popularity or usage of the objects.

In addition, many other circumstances exist which might be expected to influence the degree of haloing. These include:

1. The importance of an attribute. An important attribute might be expected to show evidence of a larger halo effect since the attribute would more strongly influence overall attitude. However, Beckwith and Lehmann (1976) found a slight (but not significant) tendency for halo effects to decrease with increasing attribute importance.

2. Questionnaire wording can affect the degree of haloing (Wilkie, McCann, and Reibstein, 1973).

3. Relative importance of the object or product class.

4. Personal characteristics such as relative intelligence, education, gestalt-proneness, or interest of the respondent in the product class.

5. Promotional strategies employed. Whether the advertising environment primarily uses very general or very attribute-specific communications and claims may influence halo.

Seemingly divergent results have been reported, even for the same product class. For example, the Roper Organization reported that "The lower your gas mileage, the more uncomfortable the seats turned out to be" (Marketing News, 1975). However, Moore and James (1977) found only a very small amount of haloing in 34 students' ratings of automobiles on several attributes, including gas mileage and comfort. We conclude that unraveling the halo-inducing circumstances will not be an easy task, particularly as the focus of inquiry moves to attitude and belief changes.

SUMMARY

Substantial disagreement exists concerning just when and how much haloing occurs. However it seems clear that a serious degree of haloing occurs in at least some circumstances. Until the nature of these circumstances is better understood it is reasonable for researchers to accommodate the possibility of halo effects within any study using belief ratings on attributes. Since the degree of haloing is expected to vary between individuals, individual level analysis would seem to be appropriate, at least until the halo effect phenomena is better understood.

Lastly, it is worth noting that in recent experimental manipulations, subjects were found to halo extensively, even when they had sufficient information to allow for independent assessments on the attributes. Furthermore, the subjects actually believed that the influence ran the other way, i.e., that their assessments on the attributes prompted their overall evaluative rating (Nesbill and Wilson, 1977; Ryan, 1977).

REFERENCES

R. P. Abelson, E. Aronson, W. J. McGuire, T. N. Newcomb, and M. J. Rosenberg, eds., Theories of Cognitive Consistency: A Sourcebook (Skokie: Rand McNally, 1969).

Neil E. Beckwith and U. Victor Kubilius, "Empirical Evidence of Halo Effects in Store Image Research," Advances in Consumer Research (1977 ACR Proceedings).

Neil E. Beckwith and Donald R. Lehmann, "The Importance of Halo Effects in Multi-Attribute Attitude Models," Journal of Marketing Research, 12 (August 1975), 265-275.

Neil E. Beckwith, "Halo Effects in Multiattribute Attitude Models: An Appraisal of Some Unresolved Issues," Journal of Marketing Research, 13 (November 1976), 418-421.

Albert C. Bemmaor and Joel C. Huber, "Econometric Estimation of Halo Effects: Single vs. Simultaneous Equation Models," Advances in Consumer Behavior (1977 ACR Proceedings).

W. V. Bingham, "Halo, Invalid and Valid," Journal of Applied Psychology, 23 (1939), 221-229.

F. S. Bourne, "Group Influence in Marketing and Public Relations," in R. Likert and S. P. Hayes, eds., Some Applications of Behavioral Research, UNESCO, Paris, 1957, 207-257.

Horace B. English, A Student's Dictionary of Psychological Terms (New York: Harper Brothers, Inc., 1934, fourth edition).

N. T. Feather, "A Structural Balance Model of Communication Effects," Psychological Bulletin, 71 (1964), 291-313.

Leon Festinger, Conflict, Decision, and Dissonance (Stanford: Stanford University Press, 1964).

Fritz Heider, The Psychology of Interpersonal Relations (New York: John Wiley & Sons, 1958).

Joel C. Huber and William L. James, "A Theory of Halo," working paper, Krannert School of Industrial Administration, Purdue University, March 1976.

Joel C. Huber, "A Concept of Halo," Proceedings of the Association for Consumer Research, 1977.

William L. James and Forrest G. Carter, "Halo Effects and Location Preferences," Advances in Consumer Research (1977 ACR Proceedings).

Barbara Koltuv, "Some Characteristics of Intrajudge Trait Intercorrelations," Psychological Monographs, 76 (1962).

David Krech, Richard S. Crutchfield, and Egerton L. Ballachey, Individual in Society: A Textbook of Social Psychology (New York: McGraw Hill, 1962), 67.

Marketing News, "Marketing Briefs," June 6, 1975, 5.

Arthur G. Miller, Role of Physical Attractiveness in Impression Formation," Psychonomic Science, 19 (1970), 241-243.

William L. Moore and William L. James, "A Study of the Amount of Halo in the Perceptions of Automobiles," Advances in Consumer Behavior (1977 ACR Proceedings).

Richard E. Nesbitt and Timothy DeCamp Wilson, "The Halo-Effect: Evidence for Unconscious Alteration of Judgments,'' Journal of Personality and Social Psychology, 35 (1977), 250-256.

M. J. Rosenberg, C. I. Hovland, W. J. McGuire, R. P. Abelson, and J. W. Brehn, Attitude Organization and Change, An Analysis of Consistency Among Attitude Components (New Haven: Yale University, 1960).

Michael J. Ryan, "An Experimental Investigation into the Effects of Halo on Attitudinal and Normative Structures,'' Advances in Consumer Behavior (1977 ACR Proceedings).

M. Sherif and C. W. Sherif, Reference Groups, Exploration into Conformity and Deviation of Adolescents (New York: Harper and Row, 1964), 170-172.

A. E. Siegel and S. Siegel, "Reference Groups, Membership Groups, and Attitude Change," The Journal of Abnormal and Social Psychology, 55 (1957), 360-364.

Percival Symonds, "Notes on Ratings," Journal of Applied Psychology, 9 (1925), 188-195.

E. L. Thorndike, "A Consistent Error in Psychological Ratings," Journal of Applied Psychology, 4 (January 1920), 25-29.

Joseph Tiffin and Ernest J. McCormick, Industrial Psychology (Englewood Cliffs: Prentice Hall, 1965), fifth edition.

M. Venkatesan, "Cognitive Consistency and Novelty Seeking,'' in Scott Ward and Thomas S. Robertson, eds., Consumer Behavior: Theoretical Sources (Englewood Cliffs: Prentice Hall, 1973), 354-384.

Robert A. Wickland and Jack W. Brehm, eds., Perspectives in Cognitive Dissonance (New York: L. Erlbaum Assoc., 1976).

William L. Wilkie, John M. McCann, and David L. Reibstein, "Halo Effects in Brand Belief Measurement: Implications for Attitude Model Development," Advances in Consumer Research, I (1973 ACR Proceedings), 280-290.

----------------------------------------