Nine Consumption Lifestyles

Susan Fournier, University of Florida
David Antes, University of Massachusetts
Glenn Beaumier, Westinghouse Savannah River Company
ABSTRACT - This paper presents an empirical application of behavioral segmentation based not upon isolated brand or category purchase habits but on the collective pattern of a household's purchases across different product/service categories over time. Purchase diary records from 1717 households in the 1986 Consumer Expenditure Survey, capturing a full 88% of total discretionary purchases in 122 product/service categories, were organized in terms of the needs and values reflected in each purchase. After controlling for income, the needs-based expenditure categories were cluster analyzed, yielding nine different consumption lifestyle segments. The distinctive purchasing patterns and demographic profiles of the groups are described. The general utility of this lifestyle approach to segmentation is also considered.
[ to cite ]:
Susan Fournier, David Antes, and Glenn Beaumier (1992) ,"Nine Consumption Lifestyles", in NA - Advances in Consumer Research Volume 19, eds. John F. Sherry, Jr. and Brian Sternthal, Provo, UT : Association for Consumer Research, Pages: 329-337.

Advances in Consumer Research Volume 19, 1992      Pages 329-337

NINE CONSUMPTION LIFESTYLES

Susan Fournier, University of Florida

David Antes, University of Massachusetts

Glenn Beaumier, Westinghouse Savannah River Company

ABSTRACT -

This paper presents an empirical application of behavioral segmentation based not upon isolated brand or category purchase habits but on the collective pattern of a household's purchases across different product/service categories over time. Purchase diary records from 1717 households in the 1986 Consumer Expenditure Survey, capturing a full 88% of total discretionary purchases in 122 product/service categories, were organized in terms of the needs and values reflected in each purchase. After controlling for income, the needs-based expenditure categories were cluster analyzed, yielding nine different consumption lifestyle segments. The distinctive purchasing patterns and demographic profiles of the groups are described. The general utility of this lifestyle approach to segmentation is also considered.

INTRODUCTION

In any given culture there exist various "consumption communities," informal groups expressing shared needs, values, or lifestyles through distinctive consumption patterns (Boorstin 1973). That shared needs support common purchase priorities is reflected in many popular marketing segmentation schemes. Users of social class (Coleman 1983) and family life cycle (Wells and Gubar 1966) assume that shared socio-demographic characteristics imply shared needs and goals, and that product choices reflect those goals. Similarly, lifestyle (Wells 1975) and value-based segmentation schemes such LOV (Kahle 1983) and VALS (Mitchell 1978) assume that those with shared psychological profiles acquire similar assortments of products for the expression of their values. "Consumption constellations" acquired in the service of social roles (Solomon 1988) further demonstrate the linkage between shared needs and collective consumption patterns.

It can be argued that the study of the collective consumption patterns of groups with shared goals and values is far from complete. In fact, a call for research that identifies a typology of "consumption lifestyles" remains unanswered in the literature (Sheth 1980; Upah and Sudman 1979). Ideally, such a typology would be grounded in actual consumption behaviors rather than the abstracted values, attitudes, interests, or demographics believed to support those behaviors. Moreover, it would concern the total assortment of products purchased over time by a group member, and the overall statement about identity that this collectivity implies.

The present study identifies such a typology from an analysis of the value-based and need-driven purchases made by 1717 panel households responding to the 1986 Consumer Expenditure Survey. Before presenting the results of this cluster analysis, literature regarding the segmentation of consumers into consumption lifestyle groups is first reviewed, and value-based typologies for the organization of product/service expenditures are considered.

BEHAVIORAL LIFESTYLE SEGMENTATION

Consideration of purchasing as an interrelated, patterned phenomenon was first introduced in the lifestyle segmentation literature (Moore 1963). Before association with cognitive constructs such as interests and opinions, the lifestyle term implied a characteristic pattern of behavior that both determines and is determined by consumption (Anderson and Golden 1984). It stressed the fact that each product purchased is part of a meaningful pattern and that through lifestyle analysis, the interrelated and co-dependent behaviors shared by segment members could be isolated and lifestyle groups identified.

Backward segmentation studies were the first to empirically examine the notion of co-occurring, complementary purchases (Alpert and Gatty 1969; Wells 1967). These early cluster analyses of supermarket purchases demonstrated that purchases do fall into meaningful bundles and that groups of consumers with shared consumption patterns also share demographics, tastes and values such that consumption lifestyle groups can be meaningfully identified.

In the sociology and economics literatures, several studies have attempted to identify groups exhibiting shared consumption patterns. Zablocki and Kanter (1976) used ethnographic data to group consumers according to the sources of identity implied in their patterns of consumption. Four broad consumption communities are proposed: (1) those embracing a hedonistic lifestyle, (2) those rejecting evidences of material success and focusing on consumption of ideas and experiences, (3) those seeking cultural identity through ethnic heritage, and (4) those adopting collective behaviors such as fads or socially-sanctioned identities (e.g., the "Yuppie") to define a sense of self.

A fine example of an empirically-based, behaviorally-driven lifestyle segmentation can be found in the work of Uusitalo (1980). Analyzing share of expenditures for each of seven basic categories (i.e., health, transportation, food, furniture, dwelling, clothing, other), she identifies eight Norwegian "consumption lifestyles." In addition to socio-demographics, these groups vary in dollar allocations for necessities versus discretionary choices, spending on mobility and transportation, and relative "consumption modernity" (i.e., the use of modern, "processed" goods and services).

Douglas and Isherwood (1979) also partitioned consumers into lifestyle groups based on actual expenditure patterns. The three-group solution considers both total spending levels and the targets of those expenditures. These include: (1) a small-scale expenditure group with a high proportion of dollars allocated to food, (2) a medium scale group with disproportionate spending on advanced consumer technology, and (3) a large-scale group with a high proportion of total expenditures on entertainment and information systems.

Each of these behavior-based lifestyle analyses has been limited by a focus on total expenditures without controls for income. As a result, low/medium/high expenditure groups generally dominate the solutions. In addition, little theoretical justification has been offered for the product and service categories chosen. Typically, available functional category codes such as food and housing are used to collapse expenditure information, restricting the types of insights that are offered. It is suggested that a lifestyle scheme that takes advantage of the richness of the values construct and its conceptual link to behaviors, while not losing the advantage of the validity of the behavioral measures themselves, will add such richness and insight. The cluster analysis that follows takes this approach.

A CONSUMPTION LIFESTYLES SEGMENTATION STUDY

Product/Service Categorization Schemes

A call for a needs-based categorization scheme for consumer products and services has been raised in the literature (Sheth 1980). Such a scheme would provide insight into consumer purchase motivations and allow inferences about the values that structure those motivations, thus providing a meaningful framework for analysis of collective consumption patterns. By classifying products/services in terms of the needs and values they imply, and then clustering people on their purchase patterns in relation to those categories, a rich and valid typology of consumption lifestyles emerges.

Several researchers have indirectly attempted product categorization along these lines. Csikszentmihalyi and RochbergHalton (1981) studied people's special possessions in terms of the meanings they held for their owners and found that products could be grouped according to the basic values they satisfy. Prentice (1987) had subjects classify favorite possessions into groups reflecting their similarities as sources of value to their owners; based on these data, she developed a four-dimensional solution for favorite possessions, including self-expressive/instrumental, recreational/practical, cultured/everyday, and prestigious/common dimensions. More directly, Settle and Alreck (1989) suggest classifying all consumer products, not just special possessions, according to the needs and values fulfilled through their purchase and usage. The authors propose their own list of fifteen needs categories for this purpose.

The categorization scheme used in the current exercise incorporates the above-mentioned research while building from an existing values framework (White 1951). By categorizing the values evidenced in a wide-variety of written and verbal materials collected over an eight-year period (e.g., autobiographies, TAT records, children's stories, ads, psychoanalytic records, speeches), White identified seven basic needs and 14 sub-categories believed to capture "the all-pervasive values that people in our culture talk about when they say they want something" (p.11). The categories include: (1) egoistic needs, including achievement/recognition and dominance/control, (2) need for play, including desires for experience and excitement, beauty, and creativity, (3) need for practicality, (4) need for cognition and knowledge, (5) morality, as reflected in legal and religious codes, (6) physiological needs, including food and health, safety and security, physical comfort, and cleanliness and (7) social needs, including appearance and belonging.

Product/Service Expenditure Categorization Task

Preliminary tests supported the selection of White's scheme over other typologies offered in the literature. Three judges more readily and reliably classified products into White's value categories (83% of products classified with 73% interjudge agreement) than they did using Settle and Alreck's groups (80% classification; 67% agreement) or the categories of Rokeach (65% classification; 65% agreement) and LOV (70% classification; 68% agreement).

Several revisions to White's original scheme were made in order to increase its relevance to the consumption setting. First, pretest diagnostics suggested a split of the practicality need into two categories: functionality and convenience. Segmentation studies in the economics area supported the addition of a separate category for housing expenses. Lastly, other value schemes highlighted two important but missing social needs, namely conformity and nurturance. This resulted in 18 need/value groups for classification and analysis.

One hundred thirty-two product and service expenditure categories were classified into the 18 need/value groups for purposes of this study. The assignment of products/services to the need/value categories was determined by a panel of ten "expert judges," i.e. students and faculty in the marketing and sociology fields familiar with the values literature. First, each of the need/value categories was discussed with the judges as cards summarizing abbreviated definitions were systematically arranged on a table. The judge was then provided with a stack of randomized cards on which the names of the 132 product/service categories were printed. Judges were instructed to "consider the predominant purchase motivation behind consumer acquisition of that particular product or service" and to classify it into one of the need/value categories accordingly. Beyond the 18 categories mentioned above, a "don't know/product ambiguous" category and "multiple needs/no one need dominant" group were also available to improve the reliability and accuracy of classification decisions.

Retained for study were those products for which a dominant need/value group was identified. It was arbitrarily decided that majority (i.e. greater than 50%) interjudge classification agreement would satisfy the criterion of dominance. Ten products were eliminated because of ambiguity, leaving 122 products classified reliably into 18 needs-values categories for analysis. Percentage agreement between judges for the sorting task was calculated at 72%, a result which compares favorably with other studies in which stimuli have been classified along complex value or need lines (Pollay 1983). A list of the final need/value categories and a sample of consumer products assigned to each is provided in Table 1.

The Data

Data for the empirical portion of this paper are from the 1986 Consumer Expenditure Survey (CES), a public domain data set that contains a rich collection of household expenditure and demographic information. Adhering to strict sampling and data collection guidelines set by the U.S. Census Bureau, the survey is considered representative of U.S. households and is used by the Department of Labor's Bureau of Labor Statistics (BLS) to estimate the consumer price index. BLS estimates that 90-95% of total yearly household expenditures are captured in survey panel member diaries.

The CES expenditure data is in diary format; on a quarterly basis, each household reports the dollar amount of expenditures in each of 528 possible categories. As a data management technique for the present study, several expenditure categories were aggregated into comprehensive, non-redundant variables (e.g., six floor coverings were collapsed to form a "carpet" category). Other variables were excluded in light of their savings or investment nature. This left 132 expenditure categories, capturing 88% of the expenditures reported by CES, for the previously-described product classification task.

To remove the effects of seasonality and to maximize the probability of purchase incidence in low frequency categories, only households with purchase data for a contiguous 12-month period were used in the present analysis. This reduced the sample from 21,466 interviews to 6868 interviews within 1717 households.

In order to remove the effects of income from the analysis, to reduce effects of range variability across categories, and to encourage solutions beyond low/medium/high expenditure groups, two transformations were performed on the data. Each category expense was first expressed in terms of its share of total household expenditures; these percentage allocations were then normalized.

Cluster Methodology

One-hundred seven multivariate outliers, arbitrarily defined as those remaining "ungrouped" after 75% of total agglomerations, were first removed from the analysis in light of their potential biasing effects. The remaining sample of 1610 households was randomly split to allow a reliability assessment of the final cluster solution.

A two-stage approach to clustering was selected (Punj and Stewart 1983). Initial solutions, identified using Ward's hierarchical method with squared Euclidean distance as a measure of similarity, provided a preliminary indication of the total number of clusters and identified expenditure category variables that did not discriminate effectively among the groups. As a result of the first phase, the nine cluster solution was chosen and three expenditure categories (i.e., conformity, cleanliness and control) were excluded. The final cluster solution was identified using the Quick Cluster K-Means procedure.

FINDINGS

In this section, profiles of the nine consumption lifestyle clusters are presented. Each group has been assigned a label to reflect the central theme governing its unique value-related spending pattern. Despite the fact that the clustering was determined on the basis of normalized proportionate spending patterns, marked discrimination between the groups in terms of socio-demographic characteristics is also evident. Admitting within-group variation on age, income, social class and family life-cycle, the groups do nonetheless display dominant demographic profiles that effectively distinguish them from other groups. These differences are also highlighted in the profiles that follow. Average spending patterns, in terms of both absolute dollars and shares, are presented in Table II; Table III summarizes key socio-demographic characteristics of the clusters.

Functionalists

True to their name, the Functionalists spend most of their money on essentials necessary for practical and efficient operation of the household, allocating a full 39% of their dollars to items such as household appliances, tools, cookware, window blinds, glassware, dishes and linens (versus 15% of dollars on functional needs for the sample as a whole). This "mainstream" group of average educated, average income, largely blue collar workers possesses life-stage characteristics one would expect to be associated with investment in household amenities; the group is mainly comprised of young (33% less than 35 years of age as compared to 23% on average) and middle-aged (37% are 35-54) dual-earning couples (78% married versus 62% for other groups' average; 65% two earners versus 47% on average) in the full nest stage (45% in full nest versus 29% on average; 50% have kids versus 37% on average).

Nurturers

The second group has been labeled the Nurturers. Also predominantly in the Full Nest stage of the FLC (47% versus 29% on average), expenditure patterns in this group are governed by child-rearing concerns and family values. Almost one-in-five dollars (18.3%) are spent on nurturance needs as compared to only 2.9% for the total sample on average. It appears that household start-up expenses have been completed or perhaps postponed, as expenses in the functional category among this group are the second lowest in the sample (8.7% versus 14.8% on average). Nurturers are younger than the Functionalists (52% under 35 years of age versus 33% for Functionalists), perhaps in initial stages of household formation. Nurturer households are largely middle class (45% versus 30% on average); the heads of these households are more likely than average to be highly educated persons (32% graduated from college versus 26% on average) with white collar jobs (47% versus 39% for the total sample). Interestingly, the group is characterized by lower-than-average income levels ($19,569 vs $25,875 on average), largely reflective of the one-earner status child-rearing duties have commanded for a third of these households.

TABLE 1

CONSUMPTION NEED/VALUE CATEGORIES

TABLE 2

VALUE-BASED EXPENDITURE PATTERNS OF THE CLUSTERS

TABLE 3

SOCIO-DEMOGRAPHIC CHARACTERISTICS OF THE CLUSTERS

Aspirers

This group has been labelled in light of its overwhelming status-seeking purchase behaviors. Spending a full 34.6% of their total dollars on status-related goods and services (versus only 3.9% on average), this group claims an average yearly status expense of $12,823 (versus $1,451 for total sample). Providing further evidence of "enjoying the high life" through excess consumption, members of this group report comparatively high spending on (status) housing accommodations ($3,689 versus $3,064 on average), (privileged) entertainment needs ($1,876 versus $1,319 on average), and (expensive gourmet) foods ($5,292 on survival needs versus $4,916 on average). While Aspirers have the income to support their habits ($35,946 before-tax income versus $25,875 for the total sample), their rank within the sample as only fourth-highest in earnings stands in sharp contrast to their #1 rank in total expenditures.

The Aspirers possess all the stereotypical "Yuppie" characteristics. Beyond the high incomes mentioned above, group members are highly educated (41% versus 26% have graduated from college) and hold the greatest proportion of white collar positions in the sample (63% versus 39%). The majority have attained middle (48%) or upper-middle (33%) social class standing. Many of these consumers are DINKS: married (75%), dual-earner households (65%) with no children (2.8 HH size; 28% from husband/wife pairs with no kids).

Experientials

This group's distinctive spending characteristic is a disproportionate allocation of dollars to entertainment needs (9.2% versus 5.8% on average) and hobbies (.6% versus .3% on average). Experientials also spend disproportionately in the appearance category (3.3% versus 2.5% on average; $844 versus $624 on average). These fun-lovers appear to exhibit active, "on-the-go" lifestyles, overspending on convenience-oriented products (3.1% versus 1.7% in total) that perhaps save time for things they hold to be more important i.e., having fun.

The Experiential group is disproportionately comprised of singles (32% versus 25% on average; 43% single-earner HH versus 31% on average). Possessing only average educational credentials, members of the group nevertheless hold white collar jobs (52% versus 39% in total) and earn incomes that are more than capable of supporting their vast entertainment needs ($32,048 versus $25,875 on average).

Succeeders

A group of established households, labelled Succeeders, was also identified. Middle-aged members of this cluster (59% are 35-54 compared to 37% for the total sample) possess by far the highest educational credentials in the sample (66% graduated college as compared to 26% for the total, and as compared to 41% and 34%, respectively, for the Aspirers and the Experientials). Their tendency toward dual-earner status (74% versus 47% for total sample, 65% for Aspirers, and 45% for Experientials) in white collar jobs (56%) commands for them the highest household income levels of any group in the sample; on average, Succeeders earn a full $42,779.

And Succeeders spend their money, claiming the highest absolute dollar expenditures in 12 of the 15 value categories. They also report purchases in 61 of the product categories measured, as compared to only 43 on average. Besides this difference in pure volume, a distinguishing feature of the Succeeder is disproportionate spending on knowledge, education and self-advancement (7.6% of total expenditures versus 1.4% on average). In addition, Succeeders invest in their work-world appearances, overspending on self-enhancement needs (3.8% of total expenditures versus 2.6% on average; over twice the average dollar amount, $1407 versus $624 on average).

The Moral Minority

The only distinguishing feature of the Moral Minority is their disproportionate donations to educational organizations, political causes and the church (12.8% of total expenditures versus other groups' average of 1.7%). Moral Minority households contain a disproportionate share of husband/wife pairs in the Empty Nest phase (37% versus 23% on average). Members of this group command the second-highest levels of income in the sample ($39,558) while maintaining traditional husband/wife roles (42% single earners versus 31% on average, the highest percentage of single earners in the sample).

The Golden Years

Golden Years' spending patterns suggest the establishment of second homes or the long-awaited remodeling of original homes; despite their ages (average age 54; 36% over 65 years of age), group members claim the third highest spending levels for housing ($3,909 versus $3,064), report heavy investments in aesthetic improvements to the home (a full 20% of expenses are devoted to this need versus only 2.1% on average), and are investing disproportionately in labor-saving conveniences (2.3% versus 1.7% on average). True to their name, the Golden Years also spend the greatest amount on products that provide physical comforts ($595 versus $362) and entertainment value ($1,895 versus $1,319). In line with their demographics, they also invest in a variety of insurance policies that yield for them emotional security in facing the uncertainties of old age (6.5% on the security need versus 4.3% for the sample on average).

Income levels are exceedingly high in this group ($38,129), second only to Succeeders and Moral Minority members. Much of their money appears to be from pension funds and investments, however, as 36% of the households report no-earner status and 40% are retired.

Sustainers

In sharp contrast to the Golden Years is a group of mature consumers (59 years of age on average; 44% over 65; 41% retired) labeled the Sustainers. Members of this group devote a full 45% of their total dollars (versus 27% on average) to basic survival needs such as food, electricity, heat and water. This group also spends a higher than average amount in absolute dollar terms on expenses in this category, perhaps reflecting use of more expensive inner-city outlets and high consumption of convenience foods (i.e., "the poor pay more"). Higher than average allocations are also reported in the belonging category (4.5% versus 3.7% for total sample), largely due to alcohol-related expenses. The Sustainers' fate appears largely constrained by economic circumstance. The group exhibits the lowest education levels (47% did not graduate from high school), the second lowest income levels ($14,909 on average), and the lowest social class rankings (75% occupy the bottom two classes) in the sample. Forty percent of the households have no earners.

Subsisters

Subsisters report the lowest expenditures of any group across all 18 need categories. The group is best described as being "house poor;" over one-third (36%) of Subsisters' total expenditures is devoted to housing, which compares markedly to a total sample average of 14.8%. Spending patterns in this group are clearly a function of socio-economic factors. Subsisters report the lowest household income levels ($12,573; 38% with incomes under $7,000), the second lowest education levels (38% less than high school), and the highest percentage of unemployed (22% versus 11% on average). A large percentage of these households live on welfare (38% have no salaried earners yet report income from other sources). A sizeable minority are single-earner families (31%) and single parents (13%), and a disproportionate number have minority racial status (29% versus 16% for total sample).

VALIDATION

Replicability of the final solution was tested using a holdout sample of 823 households. The solution was replicated in terms of both need/value and demographic profiles. The solution's predictive validity was assessed via regression analysis using FLC, social class and income as additional predictor variables. The cluster solution alone accounts for 29% of the variance in total expenditures and uniquely contributes 7% to the total variance explained when FLC, social class and income are in the model (significant at p<.001).

DISCUSSION

The results of this analysis suggest that value-related consumption patterns can be identified and that these patterns reflect more than a simple manifestation of age and income effects. While the transformation of the data into normalized percentage allocations among the value-based categories successfully attenuated income effects in the final solution, it is recognized that income nevertheless exerts some influence through variability in discretionary dollars (e.g., the percent allocated to food decreases when dollars increase). However, the regression results suggest that consumption lifestyles reveal patterns of motivation that are not captured when the focus is on the more income-privileged aspects of choice indicated through social class and income classifications, or on the evolutionary nature of needs reflected in life-cycle schemes. The fact that such strong demographic differences emerge despite controls and in solutions defined not on demographics but on value-related consumption patterns is strong testament to the continued significance of life course, income and social class variables in shaping consumer behavior.

These arguments support the utility of other value-based consumption typologies. In fact, numerous parallels between the consumption lifestyles solution presented here and the VALS typology can be drawn (e.g., Succeeders = Achievers, Aspirers = Emulators). However, several comparative advantages of the present solution can be cited. Importantly, the present scheme has been driven by actual, value-based dollar expenditures rather than abstracted, consumption-irrelevant values (e.g., "Marijuana should be legalized). Additionally, the present scheme represents members from a broad spectrum of ages, capturing each lifestyle adequately, while VALS disproportionately weights older consumer groups in its solution. A test of the predictive validity and explanatory power of a consumption lifestyle scheme such as this versus more traditional value-based schemes such as VALS or LOV could demonstrate the power of behaviorally-based value analyses, perhaps revitalizing interest in value segmentation as a marketing tool.

A final point regarding the coding of products and services into value groups deserves mention. The current procedure used expert judges to classify products and services into 18 theoretically-derived value groups and then applied an arbitrary criterion (50% agreement) to assign codes. This strategy resulted in a strong level of interjudge agreement as compared to previously-reported value classification exercises. Regardless, the procedure does ignore that fact that a given product may address multiple needs by displaying strong secondary value themes or by tapping different values for different consumers, values perhaps overlooked in theoretically-derived coding schemes such as the one used here. To control this source of error, future investigations might consider restricting analysis to products with strong, singular value themes. In addition, having consumers define and derive relevant value categories appears warranted.

BIBLIOGRAPHY

Alpert, Lewis and Ronald Gatty (1969), "Product Positioning by Behavioral Life-styles," Journal of Marketing, 33 (April), 65-69

Anderson, W. Thomas and Linda L. Golden (1984), "Lifestyle and Psychographics: A Critical Review and Recommendation," in Advances in Consumer Research, Volume XI, ed. Thomas Kinnear, Provo, UT: Association for Consumer Research, 405-411

Boorstin, Daniel J. (1973), The Americans: The Democratic Experience, New York: Random House

Coleman, Richard P. (1983), "The Continuing Significance of Social Class to Marketing," Journal of Consumer Research, 10 (December), 265-280

Csikszentmihalyi, Mihaly and Eugene Rochberg-Halton (1981), The Meaning of Things. New York: Cambridge University Press

Douglas, Mary and Baron Isherwood (1979), The World of Goods: Towards an Anthropology of Consumption, New York: W. W. Norton

Kahle, Lynn R. (1983), Social Values and Social Change: Adaptation to Life in America, New York: Praeger

Mitchell, Arnold (1983), The Nine American Lifestyles, New York: Macmillan

Moore, David G. (1963), "Life Style in Mobile Suburbia," in Toward Scientific Marketing, ed. William D. Wells, Chicago, IL: American Marketing Association, 243-266

Pollay, Richard W. (1983), "Measuring the Cultural Values Manifest in Advertising," in Current Issues and Research in Advertising, eds. James Leigh and Claude Martin Jr., Ann Arbor: Division of Research, Graduate School of Business Administration, University of Michigan, 71-92

Prentice, Deborah (1987), "Psychological Correspondence of Possessions, Attitudes and Values," Journal of Personality and Social Psychology, 53(6), 993-1003

Punj, Girish and David W. Stewart (1983), "Cluster Analysis in Marketing Research: Review and Suggestions for Application," Journal of Marketing Research, 20 (May), 134-148

Rokeach, Milton (1973), The Nature of Human Values, New York: Free Press

Settle, Robert B. and Pamela L. Alreck (1989), Why They Buy, New York: John Wiley & Sons

Sheth, Jagdish (1980), "The Surpluses and Shortages in Consumer Behavior Theory and Research," Journal of the Academy of Marketing Science, 7(Fall), 414-427

Solomon, Michael (1988), "Mapping Product Constellations: A Social Categorization Approach to Consumption Symbolism," Psychology and Marketing, 5(Fall), 233-258

Upah, Gregory and Seymour Sudman (1979), "The Consumer Expenditure Survey: Prospects for Consumer Research," in Advances in Consumer Research, Volume 6, ed. William L. Wilkie, Ann Arbor, MI: Association for Consumer Research, 262-266

Uusitalo, Llisa (1980), "Identification of Consumption Style Segments on the Basis of Household Budget Allocation," in Advances in Consumer Research: Volume VII, ed. Jerry C. Olson, Ann Arbor, MI: Association for Consumer Research, 451-459

Wells, William D. (1967), "Patterns of Consumer Behavior," in Marketing for Tomorrow...Today, eds. M.S. Moyer and R.E. Vasburgh, Chicago: American Marketing Association, 134-138

Wells, William D. (1975), "Psychographics: A Critical Review," Journal of Marketing Research, 12 (May), 196-213

Wells, William and George Gubar (1966), "Life Cycle Concept in Marketing Research," Journal of Marketing Research, 3 (November), 355-363

White, Ralph (1951), Value Analysis: The Nature and Use of the Method, New Jersey: Libertarian Press

Zablocki, Benjamin D., and Rosabeth Mass Kanter (1976), "The Differentiation of Lifestyles," Annual Review of Sociology, 2, 269-298

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