Correlations Between Three Indicators of Breadth and Variation in Sources of Stimulation

ABSTRACT - The term "stimulus variation' refers to variety in the types and sources of mental stimulation a person receives. This paper explores the relationships among three indicators of stimulus variation -- role accumulation, mass media utilization, and cosmopolitanism. Empirical evidence of correlation is provided and several conceptual issues are cited.

Citation:

Elizabeth C. Hirschman and Melanie Wallendorf (1979) ,"Correlations Between Three Indicators of Breadth and Variation in Sources of Stimulation", in NA - Advances in Consumer Research Volume 06, eds. William L. Wilkie, Ann Abor, MI : Association for Consumer Research, Pages: 111-117.

Advances in Consumer Research Volume 6, 1979      Pages 111-117

CORRELATIONS BETWEEN THREE INDICATORS OF BREADTH AND VARIATION IN SOURCES OF STIMULATION

Elizabeth C. Hirschman, New York University

Melanie Wallendorf (student), University of Pittsburgh

ABSTRACT -

The term "stimulus variation' refers to variety in the types and sources of mental stimulation a person receives. This paper explores the relationships among three indicators of stimulus variation -- role accumulation, mass media utilization, and cosmopolitanism. Empirical evidence of correlation is provided and several conceptual issues are cited.

INTRODUCTION

The term 'stimulus variation' refers to variety in the types and sources of mental stimulation a person receives. Stimulus variation has been investigated in a non-systematic way in both psychological and sociological contexts, where it has been alternatively termed activity seeking (Fowler, 1976), novelty-seeking (Finger and Mook, 1971), information seeking (Thorelli, 1975), sensation-seeking (Zuckerman, Kolin, Pierce and Zoub, 1964), exploratory drive (Nissen, 1951), spontaneous alternation (Hosada, 1964), optimal stimulation level (Hebb and Thompson, 1954), incongruity seeking (Hunt, 1963), innovation proneness (Rogers and Shoemaker, 1971), exploratory urge (Cattell, 1957), venturesomeness (Robertson, 1970), and role accumulation (Sieber, 1974; Wallendorf, 1976; Zaltman, Wallendorf and Hirschman, 1977; Hirschman and Wallendorf, 1977). Depending on whether a psychological or a sociological approach is taken, engaging in a lifestyle which involves a high level of stimulus variation can be viewed as arising from an internal need or drive, or as being a condition which is socially induced.

Whatever its cause for a particular individual, the concept of stimulus variation is an important one for the study of consumer behavior for several reasons. First, the concept of stimulus variation is important for theories concerning consumer information processing and information overload (e.g., Scammon, 1977; Bauer, 1960; Hunt, 1963; Jacoby, Speller and Kohn, 1974a; Jacoby, Speller and Kohn 1974b; Katona and Mueller, 1955; Osgood and Tannenbaum, 1955; Ray, 1973; Russo, 1974), which posit that individuals may differ in their ability to process and utilize diverse bits of information. This may result in a situation in which, due to the diversity of the bits of information and individuals' differing abilities to handle stimulus variation, one person's information 'overload' constitutes another individual's 'underload'.

The concept of stimulus variation is also important for understanding how individuals choose among various mass media sources of information and stimulation. This may affect the type of media selection strategy which will be most effective in reaching a given target audience. The strategy for reaching target audiences containing large numbers of consumers whose chosen sources reflect stimulus variation will differ from the strategy for reaching audiences characterized by low stimulus variation.

It also appears logical that an individual could obtain stimulus variation in several different ways. For example, it is possible that individuals who are desirous of a high level of stimulus variation (for either internal need-based reasons or for external socially-induced reasons) would have a lifestyle characterized by several types of stimulus variation. That is, the person who belongs to a varied set of groups or organizations and participates in a varied set of social activities might also be one who would expose himself to a varied set of mass media programs and publications and to geographically-dispersed stimuli. In other words, a person may obtain stimulus variation simultaneously in several different ways (e.g., interpersonally, impersonally, and geographically).

PURPOSE

The purpose of this exploratory research is to broaden research efforts already underway on the sociological aspects of the stimulus variation phenomenon to two additional areas: mass media utilization and cosmopolitanism. The area previously under investigation was that of role accumulation. A sociological construct, role accumulation refers to the number of groups to which an individual belongs.

Previous work has developed five propositions concerning role accumulation and its relationship to consumer behavior (Wallendorf, 1976). This work states the proposition that role accumulators are more likely to know about innovations earlier than others, and to adopt innovations earlier than others. It also makes the proposition that for role accumulators the time from awareness to adoption of an innovation will be shorter than for other individuals. Finally, this work makes the proposition that role accumulators will initiate and will he sought out for interpersonal communications about innovations more often than others. These relationships are shown in Figure 1. These relationships are presented here to indicate how one form of stimulus variation, role accumulation, may affect some consumption-related activities. These relationships are not, however, tested in the research reported here.

Instead, the purpose of this research is to broaden the applicability of the role accumulation concept and connect it with the more general phenomenon of stimulus variation. Three propositions were formulated and are given below.

1. The greater the degree of role accumulation, the greater the number of different types of mass media vehicles an individual will utilize.

2. The greater the degree of role accumulation, the greater the degree of cosmopolitanism.

3. The greater the number of different types of mass media vehicles utilized, the greater the degree of cosmopolitanism.

The relationships proposed in these three propositions are shown in Figure 2. These three propositions are the hypotheses tested in this research.

FIGURE 1

PROPOSED RELATIONSHIPS BETWEEN ROLE ACCUMULATION AND DIFFUSION OF INNOVATION CONCEPTS

FIGURE 2

HYPOTHESIZED RELATIONSHIPS BETWEEN THREE TYPES OF STIMULUS VARIATION

METHODOLOGY

To test these three hypotheses, survey data from two research projects were utilized. The first survey consisted of 587 telephone interviews conducted with residents of the Birmingham, Alabama SMSA in August 1977. The second survey was composed of 504 telephone inter- views conducted with residents of the Decatur, Illinois SMSA in December 1977. Both surveys were based on random-digit-dialed telephone interviews conducted by professional interviewers. Interviewing was conducted during the evening hours to assure an adequate representation of working persons. Sampling from each telephone exchange in the SMSA was weighted according to population, and the obtained sample was consistent with updated census data for that SMSA in all demographic variables studied (i.e., race, age, occupation, education, and income). During each interview, information was gathered on (1) the individual's group memberships and social activities, (2) mass media utilization, and (3) cosmopolitanism.

Group Memberships and Social Activities

Respondents were asked to name the groups to which they belonged in each of five categories: business, community, social, recreational and religious. They were also asked to give the extent to which they participated in several recreational and entertainment activities.

Mass Media Utilization

Respondents were asked to name the mass media vehicles they regularly used as sources of information. Due to questionnaire length constraints, these were limited to radio, newspaper and television vehicles in the Decatur survey, and newspaper and magazine vehicles in the Birmingham survey.

Cosmopolitanism

Cosmopolitanism refers to the extent to which an individual is involved in affairs, experiences, and relationships outside the geographically local social system. In the literature dealing with cosmopolitanism, it has been variously measured by finding out how often the individual travels to other cities, how much exposure the person has to cosmopolite communication channels such as national organization newsletters, how many people the individual knows outside the local area, and how geographically dispersed these people are (Rogers and Shoemaker, 1971).

In this study, respondents were asked to what extent they went travelling or sightseeing. This question actually measures only one component of cosmopolitanism, that dealing with physical ties to other areas. Other components dealing with relational or organizational ties to other areas were not measured. For the purposes of this study, however, the question on travelling and sightseeing was deemed sufficient. This question was included in the larger set of questions concerning participation in various social activities.

ANALYSIS

Operationalizing Role Accumulation

Two measures of role accumulation were utilized. Because a measure of role accumulation should represent the variety of group memberships and social activities engaged in by the individual, one way of operationalizing the concept is to sum the number of different types of groups and social activities in which the individual participates. Therefore one point was given for each type of organization to which the individual belonged. For example, if a respondent belonged to 2 business groups, 3 community groups, and 1 religious group, the group membership score would be 3 (one point for each type of group for which the individual reported membership). This measure is designed to indicate variety rather than merely sheer number of group memberships.

The second component of this role accumulation index was based on the number of different social activities in which the individual participated. A score of one was given for each of the seven social activities in which the individual participated very often, moderately, or not too often. (The sightseeing and travelling item was not included in this index). For example, if a respondent indicated that he or she participated in tennis very often, home entertaining not too often, and never in the other activities, the social activities component of the role accumulation index would be a score of two. In summary, the first measure of role accumulation was number of different types of groups to which the individual belonged plus the number of different social activities in which the individual participated. This index was created for each individual, and the results are given in Table i for both the Birmingham and Decatur data. As can be seen from these figures, the incidence of role accumulation using this index generally follows a normal distribution, but is slightly skewed to the right. The data in Table 1 further indicate that the distribution of this index of role accumulation is generally consistent between Decatur and Birmingham.

There are some problems inherent in this summed operationalization of role accumulation, because redundance may be present among some of the variables. For example, by summing together an individual's recreational group memberships and his or her participation in tennis and golf there may be some double-counting. Secondly, it is possible that involvement in certain types of clubs or activities may be negatively associated with participation in others. For example, people who are avid concert-goers may be averse to attending spectator sporting events. By summing together all activities and group memberships the occurrence of such negative correlations may be masked.

To overcome these two potential shortcomings of the summed measure of role accumulation, a second operationalization was also used. This consisted of factor analyzing the group memberships and scores on the seven social activities. Such a procedure should serve to reduce any redundancy and extract distinct, orthogonal factors underlying the data. To accomplish this a Principal Factors Analysis combined with a Varimax rotational procedure was employed on the group membership and social activity variables in both cities. The results are given in Table 2A. In both cities three factors were obtained which had eigenvalues substantially greater than one. A total of 44.3% of variance was explained by the three factors in Decatur, and 43.4% total variance was explained in Birmingham.

Two of the factors obtained in each city appeared to be fairly consistent. Factor 1 in Decatur and Factor 3 in Birmingham both loaded very highly on recreational group memberships, golfing, tennis playing, movie attendance and attendance of spectator sporting events. Factor 2 in Decatur and Factor 2 in Birmingham also appeared to be analogous. These loaded highly on membership in business, social, community, recreational and religious group memberships in both cities. The remaining factors in each city -- Factor 3 in Decatur and Factor i in Birmingham -- possess some similarities, but do not appear to be as analogous as the other two sets of factors. For example, both factors are consistent in their high loadings for concert attendance and home entertaining, yet are inconsistently loaded on spectator sports attendance, movie attendance, bridge club participation and community group membership.

The scores an individual had on each factor were added together to create a composite principal factor index of role accumulation. The distributions obtained for this principal factor index of role accumulation are given in Table 2B.

TABLE 1

ROLE ACCUMULATION: SUMMED OPERATIONALIZATION DISTRIBUTIONS

TABLE 2A

ROLE ACCUMULATION FACTOR INDEXES

TABLE 2B

ROLE ACCUMULATION: FACTOR INDEX OPERATIONALIZATION DISTRIBUTIONS

Operationalizing Media Utilization

In each media category the objective was to obtain a measure of the breadth of an individual's media exposure rather than the depth of exposure. Thus, the number of different types of vehicles a person was exposed to was more relevant than the sheer amount of media vehicle exposure. Therefore vehicles in each medium (i.e., radio, television, magazine, and newspaper) were subdivided into several categories which appeared substantively different. An index of mass media utilization breadth was then constructed by summing the number of different categories a person was exposed to within each medium.

Categories were arbitrarily defined for each medium. Newspapers were grouped into four categories: (1) local newspapers, (2) national newspapers, (3) financial newspapers (e.g., Wall Street Journal), and (4) special interest newspapers (e.g., Village Voice). Magazines were divided into five groups based on information content. The categories were: (1) news magazines (e.g., Time), (2) fashion magazines (e.g., Vogue), (3) home magazines (e.g., Ladies Home Journal), (4) men's magazines (e.g. Playboy), and (5) special interest magazines (e.g., Photography, Yachting, and Ebony).

Radio stations were separated into three groups according to programming format: (1) rock/soul, (2) easy listening, and (3) special interest (e.g., classical music, all news, religious). Television programs were initially divided into fifteen categories which included situation comedies, daytime soap operas, night-time soap operas, drama, serials, sporting events, police shows, talk shows, movies, local news programs, national news programs, variety shows, miscellaneous PBS programs, religious shows, and specials. However, realizing that this was a rather large set of categories, a reduced set of television program types was also developed. This reduced set of television programs was composed of six categories: (1) local news programs, (2) national news programs, (3) specials, (4) movies, (5) talk shows, and (6) PBS programming. These were selected because they were judged to be more "stimulation rich" than other types of television programming, and therefore to perhaps better represent the concept of stimulus variation.

Mass Media Utilization Index

Decatur. In Decatur respondent utilization of newspapers, radio stations and television programming was measured. To combine these various media vehicles into an index of media utilization which would reflect the concept of stimulus variation, a summing procedure was employed. That is, the number of different categories of each medium (i.e., radio, newspapers, and television) in which the individual utilized at least one vehicle were summed to create an index for that individual.

To further refine the index a secondary analysis was made of "special interest" categories for newspaper readership and radio station listenership for each individual. Where the specialized vehicles in each medium appeared to represent no overlap, (e.g., a classical music station and an all news station), the individual's index was increased by the number of different special interest categories represented.

Finally, a second media utilization index was created for each individual which included the radio station and newspaper categories used in the first index, but the reduced set of "stimulation rich" television program types.

Birmingham. In Birmingham respondent utilization of the two print media, newspapers and magazines, was measured. Similar to the approach used in Decatur, a summed index was created for each individual by adding together the number of different types of magazines and newspapers utilized. Also an analysis was made of the specific vehicles contained in the "special interest" categories for each person and, where the vehicles were not redundant, the respondent's index was incremented.

The distributions for the media utilization indices created in Decatur and Birmingham are shown in Table 3. In summary, three media utilization indices were created: (1) the use of print media in the Birmingham sample; (2) the use of the broadcast media and newspapers in the Decatur sample with a large set of categories for television programming; and (3) the use of the broadcast media and newspapers in the Decatur sample but with a reduced set of categories for television programming.

Operationalizing Cosmopolitanism

Cosmopolitanism was operationalized in this research as being the frequency with which the individual was exposed to sources of external stimulation through sightseeing and travelling. The distribution of responses to this measure is given in Table 4 for both Decatur and Birmingham.

Correlational Analysis to Test Hypothesis One: Role Accumulation and Media Utilization

To measure the degree of correspondence between role accumulation and media utilization, correlational analyses were performed using the two operationalizations of role accumulation, (i.e., the summed index and the principal factors index).

The results for the summed index of role accumulation are given in Table 5. As can be seen, the correlations between the role accumulation summed index and media utilization were consistently positive and statistically significant at the p = .003 level or beyond. However, there were some variations in the degree of correlation. The weakest relationship was found to exist between the "full set" media utilization measure (i.e., that containing all television program types) and role accumulation in the Decatur data. An increased level of correlation was obtained when television program types included in the media utilization measure were "reduced" to those judged to be "information rich" in terms of stimulus variation. By using this reduced set, the correlation increased from .124 to .134 and the significance level from .003 to .001.

An even higher level of correlation and statistical significance was found in Birmingham when the role accumulation summed index was correlated with a media utilization index created from the two print media. The R in this instance was .389 which was significant at p<.001.

This finding appears to indicate that role accumulation, at least as measured by summing across group memberships and activities, is more highly related to stimulus variation for print media, alone, than for broadcast and print media combined. One explanation for this is that perhaps a greater inherent variety of stimulation is found in magazines and newspapers than in television or radio programming. The validity of this conjecture is inadequately addressed in this research , but, it would seem to be an area worthy of further inquiry.

Table 5 also shows the results obtained from correlating the principal factors role accumulation index with media utilization. In Decatur, where the two media utilization indexes were composed of newspaper, radio and television usage, correlations were positive but lower than obtained in Birmingham. The lowest correlation, R=.083 was again found between the full measure of media utilization (which included all types of television programs) and the role accumulation principal factors index. The correlation was increased to .113 when the role accumulation principal factor index was correlated with the reduced set measure of media utilization (which included only six types of television programming).

As was found with the summed measure of role accumulation, the role accumulation principal factor index appears to be more highly associated with stimulus variation in print media, alone, than with broadcast and print media combined. Whether this is due to an inherently greater availability of variety in the stimulation obtainable from print media or to some other, unmeasured variable remains unknown.

Correlational Analysis to Test Hypothesis Two: Role Accumulation and Cosmopolitanism

The results obtained from correlating role accumulation with cosmopolitanism are given in Table 5 for both the summed index of role accumulation and the principal factor index of role accumulation. As is shown in Table 5, a high degree of correlation exists between the summed index of role accumulation and cosmopolitanism. In Decatur that R value was .357, while in Birmingham an R value of .522 was found. Both of these correlations were significant at beyond the .001 level.

Similar results were obtained when the principal factor index representing role accumulation was correlated with cosmopolitanism. In Decatur the correlation was .309 and was significant at the .001 level. The Birmingham data indicated an even higher pattern of correlation between the role accumulation principal factor index and cosmopolitanism. The R value was .465 and the significance level .0001.

Correlational Analysis to Test Hypothesis Three: Media Utilization and Cosmopolitanism

To test the third hypothesis that persons using a wider variety of mass media vehicles would exhibit a greater degree of cosmopolitanism, the correlation between these two variables was computed. As is shown in Table 5 in both Decatur and Birmingham the correlations were found to be positive and statistically significant at the p = .001 level and beyond.

It was found that the breadth of the utilization of print media alone (Birmingham) had a higher correlation with cosmopolitanism than did utilization of print and broadcast media combined (Decatur). This finding was reminiscent of the results obtained when media utilization was correlated with role accumulation. The reasoning advanced to explain that finding may be equally applicable here. That is, the stimulus variation provided by television and radio programming may not be as great as that provided by print media. It is also useful to note that when the "full set" of television programs included in the Decatur media utilization index was "reduced" to those shows judged to be richer in stimulation, a large improvement in correlation, rivaling that in Birmingham, was found.

TABLE 3

MEDIA UTILIZATION INDEX DISTRIBUTIONS

TABLE 4

COSMOPOLITANISM INDEX DISTRIBUTIONS

TABLE 5

CORRELATIONAL ANALYSES

SUMMARY AND IMPLICATIONS

The consistently positive and statistically significant correlations among role accumulation, variety in media utilization, and cosmopolitanism found in both cities provide support that these three types of stimulus variation are related. The finding that stimulus variation may be generalized across several areas of an individual's behavior and particularly those relating to information-seeking activities, could have implications for the analysis and understanding of many consumer behavior phenomena.

To discuss these implications, focus will first be placed upon the individual acting within a social system and than upon social systems, generally. First, if stimulus variation activities are normally distributed within a population, and the results reported here appear to support this contention, then it follows that the majority of people will be characterized by a moderate amount of stimulus variation, while others will exhibit higher or lower amounts.

It is the individuals positioned at either end of the continuum who are perhaps most important to consumer behavior research and about whom several conjectures can be made. First, those persons exhibiting a below average amount of stimulus variation may have a reduced ability (psychologically, socially, and/or physiologically induced) to obtain and process information. This below average amount of information seeking and processing may reduce their ability to adequately evaluate relevant alternatives in many consumption activities and generally may impair their functioning in a complex, consumption-oriented society. One would expect people with low amounts of stimulus variation to be largely habitual in their purchasing behavior and possibly unable or unwilling to alter this behavior, even in the face of changing circumstances.

At the other end of the spectrum may be found those persons who experience a high amount of stimulus variation. From a broad-based consumer behavior perspective these individuals would be the consummate consumers --continuously being exposed to information about new products, new knowledge and new experiences. They are of particular interest because they may be the people most likely to be reached by any mass media information flow, whether it be commercially sponsored, public interest or otherwise.

Their value to an increased understanding of consumer behavior, however, arises not only from their heightened rate of product consumption and information acquisition activities, but from their above average rate of information distribution, as well. Individuals who have a high amount of stimulus variation may play a central role in the diffusion process. Due to their diverse and extensive contact with new information, such people may be the first to learn about new products, and among the earliest to adopt them. Additionally, because of their widespread network of social activities and group memberships, they could serve as disseminators of information and ideas about new products. These implications can be illustrated by mentally connecting Figures 1 and 2.

Shifting focus to the influence which stimulus variation may have on social consumption patterns generally, one obvious issue is the longitudinal distribution of stimulus variation within the population. Sociologist Peter Blau (1975) has hypothesized that a process of increasing multiform heterogeneity is endemic to most Western societies, with the United States serving as the archetype. Blau's thesis asserts that societies are not only becoming more complex, but that people's networks of social contracts are becoming wider and more fractionated. That is, people are becoming exposed to an increasing variety of stimuli from diverse sources --the media, friends, business associates, educational institutions and special interest groups.

From a sociological point of view one of the key concerns is the effects these increasing levels of multiform heterogeneity may have on society. The most constructive scenario, perhaps, would be the occurrence of a gradual and nondisruptive movement by the entire population toward adaptation to greater levels of social complexity. That is, the distribution of stimulus variation within the population would maintain its normal shape but be shifted steadily rightward.

Another, more disconcerting possibility is that instead of gradually adapting themselves to higher levels of stimulation, a large portion of the population would begin to increasingly reject change and move instead toward desiring lower levels of stimulus variation. If this situation holds, and some recent research supporting an apparent resurgence in ethnicity suggests that it may (Glaser and Moynihan, 1975), then society may become bipolar with respect to stimulus variation. The distribution mean would remain constant, but one portion of the population would withdraw from stimulus variation, while another portion moved forward to embrace it, giving the distribution a bimodal character.

If the first of these scenarios is the more valid one, what alterations can be expected in consumption behavior? First, a primary result may be an increase in the rapidity with which new products (broadly defined) diffuse. A higher level of mass media monitoring and a broadened pattern of interpersonal contacts should prove highly conducive to new product adoption and diffusion. Closely related to this, however, is the notion that there would be a correspondingly shorter life cycle for a given product, as it would be more readily discarded in favor of something novel.

Conversely, if the second scenario of a bipolar society should become a reality, its implications for consumer behavior are also significant. First, marked differences between large population segments in their willingness and ability to process information may effectively create a two-tier consumption system. Within one tier high-stimulus-variation individuals may manifest the type of behavior described earlier. That is, new products would be learned about, adopted and discarded rapidly. Within the second tier, the diffusion of new products would be retarded as people maintained habitual consumption patterns and were unwilling or unable to seek out new information or to adopt novel consumption alternatives.

The implications of these two possible societal scenarios are large not only for consumer behaviorists but for political scientists and sociologists as well. Although longitudinal analyses are required for studying such a phenomenon, some clues as to the relative validity of each scenario may be gained by studying societies in different stages of development and looking for shifts toward bipolarization or increased overall levels of stimulus variation.

The research reported here is unable to address these larger issues because of the similarities in size and social composition of the two social systems studied. However, the genuine value such analyses would have should serve as adequate impetus to researchers for extending investigation beyond the preliminary boundaries established here.

REFERENCES

Raymond A. Bauer, "Consumer Behavior as Risk Taking, in R. S. Hancock (Ed.), Dynamic Marketing for a Changing World, Proceedings of the 43rd National Conference of the American Marketing Association (1960), 389-440.

Peter M. Blau, "Presidential Address: Parameters of Social Structure," American Sociological Review, Vol. 39 (October 1975), 615-635.

R. B. Cattell, Personality and Motivation Structure and Measurement (New York: Harcourt, Brace & World, 1975).

Frank W. Finger and Douglas G. Mook, "Basic Drives," Annual Review of Psychology, Vol. 22 (1971), 8-9.

Harry Fowler, "Satiation and Curiosity: Constructs for a Drive and Incentive-Motivational Theory of Exploration,'' in K. W. Spence and J. T. Spence (Eds.), Psychology of Learning and Motivation, Vol. 1 (New York: Academic Press, 1967), 157-277.

Nathan Glaser and Daniel P. Moynihan (Eds.), Ethnicity: Theory and Experience (Cambridge, MA: Harvard University Press, 1975).

D O. Hebb and W.R. Thompson, "The Social Significance of Animal Studies," in G. Lindsey (Ed.), Handbook of Social Psychology (Reading, MA: Addison-Wesley, 1954), 551-2.

Elizabeth C. Hirschman and Melanie Wallendorf, "Role Accumulation and Early Adoption," Working Paper #260, Graduate School of Business, University of Pittsburgh, Pittsburgh, PA (1977).

K. Hosada, "Curiosity and Spontaneous Alternation," Japanese Journal of Psychology, Vol. 35, 3 (1964), 126-32.

J. M. V. Hunt, "Motivation Inherent in Information Processing and Action," in O. J. Harvey (Ed.), Motivation and Social Interaction - Cognitive Determinants. (New York: Ronald Press, 1963), 35-94.

Jacob Jacoby, Donald E. Speller and Carol A. Kohn, "Brand Choice Behavior as a Function of Information Load," Journal of Marketing Research, Vol. II (1974), 63-9.

Jacob Jacoby, Donald E. Speller and Carol A. Kohn, "Brand Choice Behavior as a Function of Information Load: Replication and Extension,", The Journal of Consumer Research, Vol. 1 (1974), 33-42.

G. Katona and E. Mueller, "A Study of Purchase Decision," in L. H. Clark (Ed.), Consumer Behavior: The Dynamics of Consumer Reaction (New York: New York University Press, 1955), 30-87.

Henry W. Nissen, "Phylogenetic Comparison," in S. S. Stevens (Ed.), Experimental Psychology (New York: John Wiley & Sons, 1951).

C. E. Osgood and P. H. Tannenbaum, "The Principle of Congruity in the Prediction of Attitude Change," Psychological Review, Vol. 62 (1955), 42-45.

M. Ray, "Psychological Theories and Interpretation of Learning," in S. Ward and T. S. Robertson (Ed.), Consumer Behavior: Theoretical Sources, (Englewood Cliffs, NJ: Prentice-Hall, 1973), 45-117.

T. S. Robertson, Consumer Behavior, (Glenview, IL: Scott Foreman Co., 1970).

Everett M. Rogers and E. Floyd Shoemaker, Communication of Innovations, (New York: The Free Press, 1971).

J. E. Russo, "More Information is Better: A Reevaluation of Jacoby, Speller and Kohn," The Journal of Consumer Research, Vol. 1 (1974), 68-72.

Debra L. Scammon, "Information Load and Consumers," Journal of Consumer Research (December, 1977), 148-155.

Sam Sieber, "Toward a Theory of Role Accumulation," American Sociological Review. Vol. 39 (August, 1974), 567-578.

Hans B. Thorelli, Helmut Becker and Jack Engledow, The Information Seekers: An International Study of Consumer Information and Advertising Image (Cambridge, MA: Ballinger, 1975).

Melanie Wallendorf, "Exploring the Impact of Role Accumulation on the Diffusion of Innovations," Working Paper, Graduate School of Business, University of Pittsburgh, Pittsburgh, PA (1977).

Gerald Zaltman, Melanie Wallendorf and Elizabeth C. Hirschman, "Role Accumulators Versus Innovators, Opinion Leaders and Gatekeepers as Important Groups in Consumer Research," Working Paper, Graduate School of Business, University of Pittsburgh, Pittsburgh, PA (1977).

M. Zuckerman, E. A. Kolin, L. Price and I. Zoub, "Development of a Sensation-Seeking Scale," Journal of Consulting Psychology, Vol. 38 (1964).

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