Religious Differences in Cognitions Regarding Novelty Seeking and Information Transfer

Elizabeth C. Hirschman, New York University
[ to cite ]:
Elizabeth C. Hirschman (1982) ,"Religious Differences in Cognitions Regarding Novelty Seeking and Information Transfer", in NA - Advances in Consumer Research Volume 09, eds. Andrew Mitchell, Ann Abor, MI : Association for Consumer Research, Pages: 228-233.

Advances in Consumer Research Volume 9, 1982      Pages 228-233


Elizabeth C. Hirschman, New York University


Religious affiliation is one of the fundamental elements of social behavior (e.g. Gurvitch 1971, Merton 1931, Greeley 1977, Patai 1977). In various behavioral paradigms, religion has been cast as a palliative used by the ruling class to subjugate and pacify the proletariat (Karl Marx, A Contribution to the Critique of Political Economy, 1983); or conversely as a system of social values that spurs economic growth and industrial development (Max Weber, The Protestant Ethic and the Spirit of Capitalism. 1905).

Regardless of one's theoretical perspective, however, the relationship of religion to consumption behavior must be acknowledged. In United States society, as in most cultures, a consumer's religious affiliation is strongly tied to what Duncan et al (1972) term "life chances." Religion links us through a variety of connections to a style of life that determines not only what and how much we consume, but why we consume (Hirschman 1982). Composing this life style are the SES components such as income, education and occupation (all strongly related to religious identity) and also a set of consumption values, expectations and beliefs that affect behavior (Hirschman 1982).

Religious affiliation has been a major topic of investigation in several of the behavioral sciences. Within sociology there have been studies of the correlation between religious affiliation, socioeconomic status and fertility (Anderson 1970, Greeley 1977, Lenski 1963, Roof 1979). Psychologists have studied relationships between religious affiliation and various personality characteristics (Arieti 1976, Patai 1977); and causal linkages between religion and political orientation have been a major focus in political science (Davidowicz 1977). Findings from these investigations provide evidence that religious affiliation may substantially affect a variety of consumption processes. Yet, despite the empirical linkage established between religious affiliation and consumption, virtually no studies have been conducted on this topic within the consumer research tradition.

Lack of attention to religious affiliation within consumer research may spring from at least three sources. First, it is possible that researchers are unaware of the substantial evidence accumulated linking religion to consumption. This may be because many studies of religion relevant to consumer behavior are conducted within the fields of sociology, anthropology, and political science. Since consumer research infrequently draws from these aggregate social sciences, it is possible that prior investigations of religion are unknown to many in the field.

A second reason may be an implicit normative sanction against the investigation of religious affiliation within consumer research. Perhaps religion is viewed as a subject too sensitive to submit to investigation. If this is the case, it is unfortunate. Topics may be perceived as sensitive, because they are inherently salient to behavior. To avoid studying such an area may act to restrict the accumulation of valid knowledge about consumer behavior.

A final reason for the lack of consumer research incorporating religious affiliation may be the ubiquity of religious influence. Religion is pervasive; it permeates life, whether one is a believer or a nonbeliever. Because religion is an economic force, a political force and social force, it directly or indirectly affects every consumer. The ubiquity of religion provides a major rationale for its investigation. If religious affiliation provides an influential belief system to the consumer. then it is important that we examine its effects.


The purpose here is to investigate three religions Catholicism, Protestantism, Judaism - as cognitive systems (Gurvitch 1971). A cognitive system is a set of beliefs, values and expectations that are shared by members of a group (Berger 1961). In this perspective, adherents to a particular religious creed (e.g., Catholicism) are viewed as possessing an identifying cognitive system. The theological and social characteristics giving rise to the Catholic, Jewish, and Protestant worldviews are described in detail in Hirschman (1982). The present research focuses on differences in the cognitive systems possessed by Catholic, Jewish and Protestant consumers regarding two key consumption dimensions: inherent novelty seeking and information transfer.

These two areas were selected for investigation because they are central to several important consumption processes: Innovation diffusion and adoption, promotional effectiveness, interpersonal influence, and decision-making. Hence, if differences are found in the cognitive structures underlying novelty seeking and information transfer across these three religious groups, then the utility of religion as a relevant influence on consumer behavior will be lent preliminary support.



The sample for the study consisted of 96 Catholic, 120 Jewish, and 114 Protestant consumers who responded to survey questionnaires concerning their religious and nationality identification, marital status, socioeconomic status, novelty seeking predispositions, and information transfer frequencies. Subjects were recruited on a systematic, stratified basis from the New York City metropolitan area in a balanced cohort design (Lehmann 1979).

In order to control for SES differences known to exist among these religious groups [This is important, since education and occupation have been found strongly related to cognitive structure and complexity.] (Hirschman 1982), questionnaires were administered to consumers drawn from the Upper Middle social class and above. [This was necessary since over 80% of Jews are found in these SES stata.] All respondents earned $25,000 or more per year in household income, were college-educated, and were drawn from households headed by persons having a managerial or professional occupation; thus they constituted an SES cohort (Glenn 1977). As desired, there were no significant differences among consumers within the three religious groups along any of the demographic variables (e.g. 9 income, education, occupation) measured. Thus, although the sample is representative of a relatively upscale group of Catholic, Jewish, and Protestant consumers, it is not hampered by the SES inequalities that would be present if a random sample of these three denominations were drawn from the population at large. [This was necessary since over 80% of Jews are found in these SES stata.]

The second common factor across all three groups is represented by F1 among Protestants, F2 among Jews, and F3 among Catholics. The coefficients of congruence were .76 C to J, .81 J to P, and .80 C to P. This factor loaded primarily on movies, books, and magazines and appeared to represent Mass Media novelty seeking. In contrast with the first common factor, this second factor exhibited a great deal more religious idiosyncracy, as is reflected in the lower levels of interdenominational congruence. For example, among Catholics, the factor correlated with shopping, among Jews with apparel, but with no other types of inherent novelty seeking among Protestants. As before, there were interdenominational discrepancies in the amount of variance accounted for by this factor. The highest was 261 for Protestants; the lowest was 11% for Catholics.

The third common factor is represented by F4 among Catholics, F3 among Protestants, and F4 among Jews. The coefficients of congruence were .81 C to J; .73 J to P; and .64 C to P, which represent low to moderate levels of factor compositional similarity across the religious group. The amount of explained variance ranged from 12: for Protestants to 8% for Jews. Because the factor loads on religious and political novelty seeking across all three groups, it appears to represent Ideological novelty seeking Among Catholics, this factor also correlates with sports activities, among Jews with apparel, and among Protestants with home furnishings. Thus we again find a common factor displaying several denominationally-linked unique loadings and different levels of explained variance.

From an examination of these three common factors we may tentatively conclude that inherent novelty seeking is primarily oriented toward Sensory/Experiential consumption among Jews and Catholics, and toward Mass Media consumption among Protestants. Ideological novelty seeking tendencies, while present, do not appear strongly characteristic of any of the three groups.

In addition to these three common factors, Catholic consumers shared a factor with Protestants. This was F2 for Catholics, F4 for Protestants. The calculated congruence for this shared factor was .88. Catholics also exhibited two specific novelty seeking factors; the first had a high correlation with home furnishings and the second with transportation. Jewish consumers exhibited two specific factors. The first had a high correlation with shopping, and lower correlations with apparel and hair styles; the second factor loaded on dances, home furnishings, hairstyles and sports. Finally, Protestant consumers exhibited one specific factor with a high correlation with sports activities and lower correlations with dances and political ideas.

Thus, it appears that while an underlying common structure of novelty seeking may be found across these religious groups on three dimensions, this structure displays ample idiosyncracy both in the pattern of specific loadings and proportion of explained variance. Further, there are present within each religious group specific factors which are unique to that group. Therefore, support is given to the proposition that religious affiliation may influence the structure of inherent novelty seeking exhibited by consumers.

Information Transfer: Factor Analysis Results

The factor pattern (Tables 5, 6, 7) emerging from analysis of the fifteen information transfer items bears some similarities and dissimilarities to that obtained for inherent novelty seeking. As before, both common and specific factors were obtained for the three religious groups. However, these were generally not analogous to those obtained for inherent novelty seeking.

There appear to be two to son factors of information transfer that are exhibited by all three religious groups, although with different proportions of explained variance and with somewhat idiosyncratic loading patterns. The first is represented by F1 for Catholics, Jews, and Protestants, and explains between 24 to 30 percent of the variance in each data set. The computed coefficients of congruence were .86 C to J; .63 J to P and .71 C to P (Table 8). Thus, the level of interdenominational congruence is not high; especially for Catholic and Jewish consumers. This factor loaded commonly on Foods/Restaurants across all three groups. Among Catholics it also correlated with home furnishings, movies, magazines, and vacation spots. Among Jews, the factor was strongly associated with vacation spots and religious ideas, and to a lesser extent with movies. For Protestants, it was additionally associated with books. Thus, although there are some similarities between religious groups in the structure of the dominant factor of information transfer, there are several loadings unique to each group.

A second factor of information transfer that displayed some inter-denominational commonality was F2 for Jews and Protestants, and F3 for Catholics. Computed coefficients of congruence were .71 C to J; .67 J to P; .74 C to E, not a strong pattern of congruence. This factor loaded on Shopping Places/Apparel for all three groups. It additionally loaded on restaurants for Catholics, home furnishings and restaurants for Jews, and home furnishings and hair styles for Protestants. This factor accounted for 12% of variance among Protestants, 14% among Jews and 11% among Catholics.

In addition to these two common factors, each religious group displayed three specific factors that exhibited loading patterns unique to that group. The specific factors each accounted for from 15% to 7% of variance in a given data set. For Catholics, the first specific factor (F2) loaded on movies, books, magazines and sports; the second specific factor (F4) on political ideas, religious ideas, and transportation; and the third specific factor (F5) on dances and hair styles.

For Jews the first specific factor (F3) was associated with dances, home furnishings and hair styles; the second specific factor (F4) with magazines, political ideas and sports; and the third specific factor (F5) with books and transportation.

For Protestants, the first specific factor (F3) was positively associated with information transfer concerning political ideas and transportation; the second specific factor (F4) had high positive loadings on vacation spots and sports; and the third specific factor (F5) was most highly related to dances, movies and hairstyles.

Hence, it is apparent that the factor structure underlying information transfer displays substantial dissimilarity between these three religious subcultures. As was found for novelty seeking, the dimensions typifying each denomination are to some extent unique. This was expected based on the cognitive differences found in prior research to characterize these three groups (Hirschman 1982). While prior research (e.g., Greeley 1977) has sometimes attributed these differences largely to the SES discrepancies between Catholics, Jews and Protestants, the findings reported here indicate that religious differences are present in factor structure even when social class is controlled for, at least among the Upper and Upper Middle class consumers examined here.

Inherent Novelty Seeking: Mean Comparisons

A second perspective regarding cognitive differences in inherent novelty seeking among these three religious groups may be obtained by a direct comparison of group means. To generate these comparisons, an analysis of variance (ONEWAY--SPSS) was performed on each of the 15 items composing the inherent novelty seeking measure.

Therefore, denominational differences in the cognitive structure of novelty seeking and information transfer among the subjects should be primarily attributable to religion-based variation and not to SES factors.

Measuring Religious Affiliation

Respondents were asked the following questions to measure their religious affiliation (Hirschman 1981):

"Do you have a religious affiliation, for example: Catholic, Jewish, Protestant? Please indicate which one and the strength of your affiliation."

1. Catholic

2. Jewish

3. Protestant

4. Other (other, please list)

5. None

"Is the strength of your religious affiliation:"

Very Strong                        Very Weak

        5          4          3         2        1

Measuring religious affiliation this way has the advantage of allowing the consumer to label himself/herself and also to express the strength of religious identity she/he feels (Cohen 1977). This insures that consumers who may have been "born into" a particular religious tradition; but who no longer feel any ties with it, are not arbitrarily labeled by the researcher. The "strength of affiliation" measure is also valuable in this research, because it allows us to determine if there are differences in the level of identification among members of the three religious groups. This might affect the results because one group may be more/less influenced by its religious philosophy than another; therefore affecting the comparisons. Fortunately, this was not the case; the mean religious strength for Catholics was 3.1, for Jews 2.9, and for Protestants 2.7. These means were not significantly different at the p = .15 level (Scheffe test for multiple comparisons). Hence, each group displayed approximately the same level of religious identification.

Inherent Novelty Seeking/Information Transfer

The measurement of inherent novelty seeking was based on responses to the following question series (Hirschman 1980, 1981) "How willing are you to try something new in each of these consumption areas?" This question was followed by a list of fifteen different products ranging from hairstyles to political ideas, given in Table 1. Responses were measured on a 5-po int scale anchored by the terms "very great willingness" (5) to "very little willingness" (1) (Hirschman 1981).

Information transfer was measured using the same fifteen item product set with the statement, "How often do you give information to others about each of the products listed below?" A 5-point, "very great frequency" (5), to "very low frequency" (1) anchored scale was used in measuring consumers' perceptions of their information transfer (Hirschman 1981)

It should be noted that both of these measures are oriented toward assessing the consumer's self-perceptions and not actual behavior. To the extent that self-perceptions and actual behavior are not congruent, this measure will be a biased indicator of consumption activity. However, since the focus of the research is on cognitive structure (-i.e., beliefs about oneself) rather than upon behavior patterns, such measures are deemed appropriate.

A further assumption of both measures is that the topics covered were germane to the respondents. This assumption appears reasonable given that the sample was drawn from adult households having Upper Middle class socioeconomic status. Product classes such as magazines, vacations, restaurants, political ideas, and apparel should be active consumption areas for most, if not all, respondents.


Of primary interest in this research was whether Catholic, Jewish and Protestant consumers would vary systematically in their cognitive structure for inherent novelty seeking and information transfer. That is, would certain beliefs about novelty seeking and information transfer characterize the members of one religious group; while a different set of beliefs typified members of another group? To answer this hypothesis, two types of analysis were conducted -(1) factor analysis to measure structural similarity and (2) direct mean comparisons across groups to determine relative rankings. Factor analysis (principal components coupled with varimax rotation to achieve simple structure) was first conducted on the fifteen items in each construct measure on an intra-denominational basis, i.e., the responses of Catholic, Jewish, and Protestant consumers were factor analyzed separately. The factor analysis results for inherent novelty seeking are given in Table 2 (Catholic), Table 3 (Jewish), and Table 4 (Protestant); those for information transfer are in Table 5 (Catholic), Table 6 (Jewish), and Table 7 (Protestant). All factors exhibited a substantial level of simple structure - that is, almost all loadings are above .30 or below .10.(due to space constraints, these tables are available from author.) In each of these tables are given the factor loadings of the fifteen novelty seeking and information transfer items, the proportion of variance accounted for by a given factor, its eigenvalue, and the total variance accounted for by the factors extracted. (An eigenvalue in excess of 1.0 was used as the criterion for factor extraction). Table 8 contains the coefficients of congruence calculated between pairs of religious groups for both constructs. Discussion will first center on the results obtained for inherent novelty seeking.

Inherent Novelty Seeking: Factor Analysis Results

Examination of the inherent novelty seeking factor structures for the three groups (Tables 2, 3, 4) reveals that there are three "common" factors exhibited by each denomination, albeit with somewhat varied loadings. There was also one additional specific factor present for Jews and Protestants, one common to Protestants and Catholics, and two additional specific factors for Catholics. The first common factor is represented by the F1 factor for Catholics and Jews, and by the F2 factor for Protestants. The coefficients of congruence for this factor are .90 (C to J), .98 (J to P), and .97 (C to P). Reflecting the high interdenominational congruence, the factor has high loadings on vacation areas, food, and restaurants for all three religious groups. It also correlates positively with transportation for Jews and Protestants; with movies and books for Catholics, and with home furnishings for Jews. Although factor labels are to some extent arbitrary and subjective, this factor might be appropriately termed Sensory-Experiential novelty seeking 9 since it involves primarily the consumption of sensorially-experienced products, such as food.

Despite the fact that this factor displays generally high commonality among the three religious groups, it also possesses some loadings unique to each group, indicating the presence of interdenominational differences in its manifestation. Further, while this Sensory-Experiential factor is the major factor in novelty seeking among Catholics and Jews (explaining over 25 percent of the variance in each data set), it is a secondary factor among Protestants, accounting for only 14 percent of the variance within this group.

Post hoc contrasts using the Scheffe' test were used to compare group means on each item. The results are given in Table 9.3.

The data indicate that in seven out of fifteen instances, Jewish consumers exceeded Catholic ant/or Protestant consumers in self-perceived inherent novelty seeking. The areas in which this condition held were: apparel, books, magazines, vacation spots, foods, restaurants, and political ideas. Catholic consumers were found to exceed their Jewish and/or Protestant counterparts in inherent novelty seeking for three consumption areas: dances, transportation and sports activities. Protestants were found to exhibit higher self-perceptions of inherent novelty seeking in two consumption areas: religious ideas and hairstyles. Thus, Jewish consumers appear to exhibit a substantially higher level of inherent novelty seeking, as compared to Catholic and Protestants.

A discussion of several social and cognitive reasons for the above average levels of inherent novelty seeking among Jewish consumers is put forward in Hirschman (1981, 1982). Basically this rationale centers around normative pressures in Jewish families and religious institutions favoring the acquisition of knowledge by the individual. It is interesting to note that the acquisition of two types of knowledge strongly supported by Jewish norms is present here: Intellectual (e.g., books) and sensory (e.g., food)

Information Transfer: Mean Comparisons

The cross-group mean comparisons for information transfer are given in Table 10. Here again a rather striking pattern of subcultural differences emerges. In eight out of fifteen comparisons, Jewish consumers exceeded Catholics and/or Protestants in self-perceived information transfer. These included: movies, books, magazines, vacation spots, foods, restaurants, transportation, and hairstyles. Catholic consumers exceeded their Jewish and/or Protestant counterparts in five out of fifteen areas: dances, places to shop, political ideas, religious ideas, and sports activities. In no instances did Protestants exceed the other two religious denominations in self-perceived information transfer.

The above average level of Jewish information transfer is believed attributable to ethnic norms advocating such activity (Hirschman 1981). No rationale is immediately apparent, however, for the discrepancy between Catholic and Protestant consumers in reported levels of information transfer. It is possible that the consumption areas covered were, for unknown reasons, especially germane to the Jewish and Catholic subcultures.

Thus, we may conclude this section by noting that distinct religious affiliation differences are present in self-perceptions regarding inherent novelty seeking and information transfer. These differences indicate a higher level of inherent novelty seeking among Jewish consumers vis-a-vis Protestants and Catholics, and a higher level of information transfer among Jewish and Catholic consumers vis-avis Protestants.


The results reported have limitations which should be noted. First, the ethnic nationality of the respondents was not controlled for and may have affected the results obtained. The Catholic group was composed of approximately equal proportions of persons having Irish and Italian nationality; there were few Hispanics in the sample. The Protestant group was made-up primarily of consumers having English nationality, although some German and Scandinavian Protestants were present. The Jewish group was likely (given the sampling site) to be primarily composed of first or second generation Ashkenazic Jews, although some Sephardic Jews and Israeli nationals may also have been included. Should the nationality proportions be radically altered from those present in the data, it is possible that the structural patterns and group means will be affected.

A second limitation is that of the sampling site. Although. the obtained sample is desirably homogeneous with respect: to urbanicity and SES, it is well to remember that the New York City sampling site possesses some rather unique qualities vis-a-vis the United States, generally. First, it is a very ethnically-conscious city, which may serve to intensify subcultural perhaps serving to enhance the strength of the religious affiliation differences found here. In environs where ethnicity possesses less salience, the observed differences may be more subdued. Second, New York City has large concentrations of Catholic and Jewish consumers [The two million Jews residing in New York City comprise the largest concentration of this subculture living outside of Israel.], which is atypical of most of the United States--65 percent of which is Protestant. This atypical proportionality may serve to alter somewhat the self-perceptions of Catholics and Jews,who in New York constitute "majority" groups, but who elsewhere constitute "minority" groups.

Keeping these potential caveats in mind, let us now consider what was learned from the research. First, it was found that although some common factors underlie the structure of cognitions regarding novelty seeking and information transfer, these factors display ample religious affiliation idiosyncracy in both loading pattern and explained variance. Further, for each religious group examined, several unique factors were found, which were not generalizable across groups. Hence, cognitive dimensionality appears to be influenced by religious affiliation.

Second, the mean levels of self-perceived inherent novelty seeking items and information transfer items varied significantly across the three religious groups. Jewish consumers were found to exceed Catholics and Protestants on more specific aspects of both constructs. This was attributed to norms characterizing the Jewish subculture.

To the extent that these self-perception differences translate into actual behavior, then religious affiliation may act as a powerful determinant of consumer innovativeness, information acquisition, media exposure, and opinion leadership - all of which are linked to the constructs of inherent novelty seeking and information transfer. The data presented are not conclusive evidence that such a pattern of causation exists, yet the results are strongly suggestive of the importance of pursuing religious affiliation as a source of influence on consumer behavior. It is hoped that the present findings are sufficiently intriguing so as to stimulate further inquiry.











Anderson, Charles H. (1970), White Protestant Americans, (Englewood Cliffs. NJ: Prentice-Hall, Inc.).

Arieti, Slivano (1976), Crestivity: The Magic Synthesis, (New York: Basic Books).

Berger, Peter L. (1961), The Noise of Solemn Assemblies, (Garden City, NY: Doubleday & Co., Inc.).

Cohen, Ronald (1978), "Ethnicity: Problem and Focus in Anthropology," Annual Review of Anthropology, 7, 379-403.

Davidowicz, Lucy (1977), The Jewish Presence (New York: Harcourt, Brace & Javonovich).

Duncan, Otis Dudley, Festherman, D. L., and Duncan, B. (1972), Socioeconomic Background and Achievement (New York: Seminar Books)

Glenn, Norval (1977), Cohort Analysis (Sage University Paper Series: Palo Alto, CA: Sage Press).

Greeley, Andrew M. (1977), The American Catholic (New York: Basic Books).

Gurvitch, George (1971), The Social Frameworks of Knowledge (New York: Harper & Row).

Hirschman, Elizabeth C. (1980), "Innovativeness, Novelty Seeking and Consumer Creativity," Journal of Consume Research.

Hirschman, Elizabeth C. (June 1981, forthcoming), "American Jewish Ethnicity: Its Relationship to Some Selected Aspects of Consumer Behavior," Journal of Marketing.

Hirschman, Elizabeth C. (1982, forthcoming), "Religious Affiliation and Consumption Processes: An Initial Paradigm," Research in Marketing, Vol. 6.

Lehmann, Donald (1979), Market Research and Analysis (Homewood, IL: Richard D. Irwin, Inc.).

Lenski, Gerhard (1963), The Religious Factor (Garden City NY: Doubleday and Company).

Mars, Karl (1843), Rritick Des Hegelschen Staatsrechis (Moscow and Frankfurt Min: M.E.G.A.)

Merton, Robert [. (November 1937), "The Sociology of Knowledge," Isis, 75, 3, 493-503.

Patai, Raphael (1972), The Jewish Mind (New York: Charles Scribner Sons).

Roof, Wade Clark (September 1979), "Socioeconomic Differentials Among White Socioreligious Groups in the United States," Social Forces, Vol. 58: 1, 280-288.

Weber, Max,, Die Protestantische Ethic und der Geist Des Kapitalismus, 1, 11, Archiv Fur Sozialwisserschaft and Socialpolitick, 20 pp. 1-54.

Zaltman, Gerald, and Wallendorf, Melanie (1979), Consume Behavior (John Wiley and Sons, NY).