Decisions Past and Decisions Present: Modelling Broader Patterns of Consumer Choice

Margaret K. Hogg, University of Manchester Institute of Science and Technology
Paul C.N. Michell, University of Manchester
ABSTRACT - The creation of meaning via consumption is achieved in different ways and includes positive and negative choices: these are often used to mark inclusion and exclusion. The focus of traditional models and studies of consumer behavior has largely been on positive aspects of choice at the micro level (product or brand level) of decision-making. Examination of choices as a series of often interdependent decisions by consumers contributes to understanding the meanings created by consumption. Broader models of consumption are proposed and the findings from a study of U.K. mail order catalog shoppers indicate that patterns can be found in consumers’ stream of consumption choices.
[ to cite ]:
Margaret K. Hogg and Paul C.N. Michell (1997) ,"Decisions Past and Decisions Present: Modelling Broader Patterns of Consumer Choice", in NA - Advances in Consumer Research Volume 24, eds. Merrie Brucks and Deborah J. MacInnis, Provo, UT : Association for Consumer Research, Pages: 536-541.

Advances in Consumer Research Volume 24, 1997      Pages 536-541


Margaret K. Hogg, University of Manchester Institute of Science and Technology

Paul C.N. Michell, University of Manchester

[The authors would like to acknowledge the generous assistance of BMRB/TGI, Ealing, London with this study.]


The creation of meaning via consumption is achieved in different ways and includes positive and negative choices: these are often used to mark inclusion and exclusion. The focus of traditional models and studies of consumer behavior has largely been on positive aspects of choice at the micro level (product or brand level) of decision-making. Examination of choices as a series of often interdependent decisions by consumers contributes to understanding the meanings created by consumption. Broader models of consumption are proposed and the findings from a study of U.K. mail order catalog shoppers indicate that patterns can be found in consumers’ stream of consumption choices.


Consumption involves the search for, choice, acquisition, possession and disposal of goods and services. Traditional models of consumer behavior (Nicosia 1966, Howard and Sheth 1969, Bettman 1979, Engel, Blackwell and Miniard 1995) and studies of consumer choce (e.g. Bettman and Zins 1977, Bettman and Park 1980, Brucks 1985) largely concentrate on the earlier stages in the consumption process; and on the positive aspects of consumer choice. We propose the development of broader models of consumer behavior-involving consumption configurations-which include both the positive and negative aspects of consumer choice within the stream of consumer decisions.

We present three Models in which consumption configurations represent the combination of positive (constellations) and negative (anti-constellations) aspects of consumers’ choices. We extend existing conceptualizations by developing the construct of 'anti-constellations’ to represent negative choices. This allows us to incorporate both positive and negative aspects of consumption choices into our model building. We argue that the positive and negative aspects of consumption choices-over time-can be linked to the creation of meaning and identity.

In the first Model we draw on the symbolic interactionist perspective within social psychology to model the relationship between consumption, social roles and identities. The first Model integrates consumption with identity by relating levels of product and brand choices with levels of roles and identity. The second Model proposes the major internal and external forces which influence patterns of consumption (configurations). The third Model expands on one part of the first, integrative, model to explore how consumer decisions at the product category/brand level can be linked to the creation of consumption configurations. A preliminary study of U.K. consumers is used to examine consumption configurations and to explore combinations of consumption choices within the context of the third Model.

The central argument in this paper is that consumption is more than a series of single, simple and independent decisions; consumption involves serial, complex and interdependent decisions. When the positive and negative aspects of consumer choices are considered over a period of time and as a series of interdependent-rather than a series of independent-decisions, then it is possible to discover patterns in consumption which can be linked to systems of meaning. Our model-building contrasts with earlier models of consumer behavior which concentrate largely on aspects of search, choice and acquisition; which embrace only the positive aspects of decision making; and which are concerned with decisions about individual products.

In this paper we review the literature on self and identity; on product and consumption constellations; and on the factors which influence consumption. We then describe the model building; outline the research design and methodology; and briefly discuss the findings and the limitations of the study and indicate possible areas of future research.


Identity arises from the interdependence of self and society. Following Mead (1934) activities, such as consumption, are behaviors which are 'constituted’ into action and have meaning in creating, confirming, maintaining or transforming situated identities. Self, identity and consumption are viewed as socially constituted and linked (Dittmar 1992).

Social positions or roles; Self image and self concept

The social role or position provides the context for decisions made by individual consumers and is: "a behavioral repertoire characteristic of a person or a position". (Biddle and Thomas 1966:11). McCall and Simmons (1982) argued that collections of products and activities were taken by society as defining social roles. Studies have confirmed that characteristics of self-image are congruous with characteristics of brand image (e.g. Dolich 1969; Green et al 1969; Landon 1974; Ross 1971; Sirgy 1982; Snyder and DeBono 1985). Self concept and self mage (Sirgy 1982, Lee 1990) are used to model the layers of meaning between self and consumption. Situational self concept "... acknowledges that consumers have many self concepts and that consumption of a brand may be highly congruent with the self image in one situation and not at all congruent in another situation" (Schenk and Holman 1980:612). This reflects the view that consumers have a number of 'me’s’ (Solomon 1983:321) which are enacted as part of self, and that some are more central to self-definition than others (Stryker 1968).

When individuals have a number of different social identities or 'me’s’ which derive from their social roles-some of which are partly constituted by their acts as consumers of goods and services-there will not necessarily be congruency across all their consumption choices. However, consumption decisions which can be linked to the enactment of particular roles would be expected to display functional, symbolic or socio-cultural complementarity. Following Rogers (1951) and Alderson (1957) it would be expected that consumers would seek a 'fit’ among consumption choices within the context of the relevant role/identity.


Douglas and Isherwood(1978) argued that the 'exchange of goods...involved shared systems of meaning’ which echoed earlier views (Veblen 1899; Alderson 1957) that individuals use groups of products to communicate messages about themselves relating to status and social position. A number of studies have explored the association between consumers and groups of products (Boyd and Levy 1963; Alpert and Gatty 1969; Kernan and Sommers 1967; Wells 1968; McCall and Simmons 1982; Kehret-Ward 1988). Kehret-Ward (1988) examined a series of studies on the functional complementarity of products (Holman 1976, 1980; Solomon 1983; Rook 1985; Kehret-Ward, Johnson and Louie 1985; Kehret-Ward 1987). Kehret-Ward (1988) discussed the different labels which had been given to product configurations including: 'consumption system’ (Levy 1978); 'product constellation’ (Solomon 1986); and 'ritual artifacts’ (Rook 1985). Solomon (1986) described 'product constellation’ as 'the set of products which support a role’ (Kehret-Ward 198:191). However Solomon and Assael extended Alderson’s work on the functional complementarity of products to symbolic complementarity of product assortments which they also described as 'product constellations’: "clusters of complementary products, specific brands and/or consumption activities" (Solomon and Assael 1987:191). Consumption constellations were seen as "the identification of discriminable clusters of symbolically complementary products, services and/or activities-spanning a range of functionally dissimilar consumption categories" (Solomon and Buchanan 1991:98). Underlying the constructs of both product and consumption constellations is the idea of complementarity in the product and service choices of consumers (Alderson 1957; Green et al 1972; Kehret-Ward 1987). The two terms 'product constellations’ and 'consumption constellations’ would seem to be used often interchangeably in the literature; although there is a suggestion in the defintions above that product constellations demonstrate 'functional complementarity’ whereas consumption constellations demonstrate 'symbolic complementarity’. We would argue for a clarification of existing conceptualizations of product and consumption constellations. We would propose using product constellations to describe goods or services within a product category which can demonstrate some form of complementarity (e.g. functional, symbolic, aesthetic or cultural). An example would be the 'retail constellation’ which would represent a constellation at the product level and which would involve the series of choices made by a consumer among a series of retail outlets and distribution channels. In contrast, we would reserve consumption constellation for multi-category choices which demonstrate some form of complementarity. Consumption constellation would relate more closely to the original notion of the series of choices made by an individual consumer across multi-category products and services.

If consumption constellations can be taken to represent a consumer’s set of positive choices across multi-category products, then consumption anti-constellations represent sets of consumers’ negative choices across multi-category products and services. Anti-constellations can also be understood at two levels: product and consumption. Product anti-constellations would represent the negative aspects of product category choices. Consumption anti-constellations would represent the negative aspects of consumer choices across multiple product categories.


Anti-constellations represent the anti-complementarity of negative choices-within and across-categories of goods and services (Hogg and Michell 1996). Anti-constellations involve two aspects of consumers’ negative choices: non choice and anti-choice. Non-choice includes goods and services which are simply not chosen, often because they are beyond the means of the consumer. Anti-choice includes goods and services that are positively not chosen (Wilk 1994) because they are seen as incompatible and/or inconsistent with the consumer’s other consumption choices and preferences. Anti-constellations relate to Bourdieu’s concept (1984) of 'distastes’ and to Wilk’s studies (1994, 1995) of inclusion and exclusion. Wilk’s research demonstrated that distastes are important indicators of socially-linked preferences, notably for food and music. Food and music are two important product categories which can be linked to the socio-cultural aspects of complementarity.

The reason for the emphasis on negative aspects of choice in this model-building is that consumers use the positive and negative meanings, which can be attached to their consumption choices, to create social and cultural identities (McCracken 1986). Consumers use product, brand and activity choices to mark, and also to understand, inclusion and exclusion in relation to different social groups (Bauman 1990).


Products (Tangible and intangible)

One important influence on consumer choice is the nature of the product or service. The functional-symbolic split in product images is well-established in the literature, although questions have been raised about the tendency to see 'product image’ as unproblematic interpretations by individual consumers (Dittmar 1992:99). Consumer goods are viewed as possessing self-referent meanings which serve as symbols for the individual’s personal qualities, attitudes and values, and, at the same time, as markers for social affiliation and position (Dittmar 1992).

Personal (physiological/psychological)

Individuals use consumption to meet a variety of needs. Maslow’s conceptualization of needs and need deficiencies (1943) has been subjected to a number of criticisms. However, evidence has been found to support the pervasiveness of Maslow’s hierarchy and the potency of higher order needs (such as esteem and self-actualization), once the lower order needs (e.g. physiological, safety and social) have been met (Kast and Rosenzweig 1985:291ff). Maslow’s hierarchy will be used here to represent the personal factors-physiological and psychological-which influence consumption choices.


The third important moderating factor in consumer choice, highlighted in this Model, is the social and situational context in which consumption takes place. Bearden and Etzel’s study (1982) highlighted the distinction between public and private consumption of necessities and luxuries and the implications for the selection of products and brands by consumers.


The first Model (Figure 1) proposes that consumers, consumption and consumer goods can be understood via a system of layers of meaning. The Model begins with the micro level of decision making: the product or brand (see Bearden and Etzel 1982 for a discussion of the relevance of product and brand level choices). The Model proposes that a number of linkages exist: firstly between products and brands-and their associated evoked and inert sets-and consumers’ self images and self concepts. Secondly, linkages exist among consumption choices and non choices (represented by product constellations and anti-constellations) which relate to social roles and positions. Thirdly, there are linkages between cumulative consumption choices (represented here by consumption constellations and anti-constellations) and social role repertoires.

An important final linkage is proposed between consumption constellations and anti-constellations which together form consumption configurations. Using configurations, consumption can be linked to social identities, or 'me’s’, and to the creation of self.

Three forces (Figure 2) have been modelled as influencing the 'fit’ of consumption choices. Firstly, the force embodied by the product or service, which involves both symbolic and functional aspects. Secondly, a personal force which involves both physiological and psychological aspects. Thirdly, the social/situational force (involving the public and private aspects of consumption).



The third Model (Figure 3) represents the relationship between the forces which influence consumption (and the respective positive and negative consumption choices) and the different levels of consumption combinations. The focus of the research reported below is an exploratory study to operationalize the third Model in order to elicit consumption configurations. These consumption configurations are then discussed within the context of the second Model.


[See Hogg (1995) for an extended discussion of the application of correspondence analysis to this data set, and the detailed contingency tables and transition formulae which support the subsequent analysis offered in the next sections.]

The primary aim was to identify combinations of consumption choices represented by constellations and anti-constellations. The second aim was to interpret the positive and negative choices within the framework of broader models of consumer behavior. The third aim was to confirm the selection of certain product and service categories for exploring consumption configurations.

Quantitative and qualitative data on shoppers was used from the U.K. in-home catalog shopping industry. For the quantitative data, a two-phase procedure was used to construct and interpret the consumption combinations. The second, qualitative, stage of the research design used interview data to confirm the broader patterns of consumption found in the quantitative analysis.

Earlier studies of cumulative consumption (i.e. consumption of a range of product and service categories over time)(Wells 1968; Alpert and Gatty 1969; Solomon and Buchanan 1991; Fournier, Antes and Beaumier 1992) analyzed large consumer databases. The initial phase of this research design involved the extraction of behavioral data from a large commercial consumer database, across a range of products and services, on the users of eight mail order catalogs.


Users of eight in-home shopping catalog titles (G.U.S., Kays, Littlewoods, Freemans, Grattans, Empire Stores, Next Directory and J.D.Williams) were selected for the study. The first six are mainstream in-home shopping catalogs, generally seen as appealing to the same mass market. J.D.Williams and Next Directory represent niche players: J.D.Williams positioned towards older, larger women; and Next towards a prosperous, younger, fashion-conscious market.


Seventeen product groups were used in the study. Solomon and Assael (1987) had identified three product groups which potentially carried image laden messages, and which had been linked to occupation and social roles: clothing, electronics and automobiles. Other product groups selected were: liquor and tobacco; media; food; personal care products; sports and home equipment; luxury appliances; credit cards; discretionary use of time; and retail outlets. Retail outlets were included as an example of a service. Retail outlets embody both symbolic and functional aspects of consumption decisions and are potentially of significance because of the greater awareness of retail brands among U.K. shoppers.




The matrix of behavioral data consisted of eight columns (the users of the mail order catalogs) by over two thousand rows of product category information. Correspondence analysis was used to identify potential groups and sub groups among users of U.K. catalog shopping; and to identify combinations of product and service choices and non choices which could be linked to mail order shoppers.

A two-phase procedure was adopted for constructing and interpreting consumption combinations in the data. Transition formulae tables were used to identify consumption constellations and anti-constellations among the row profiles. The row variables clearly associated with a predominant axis were included as part of the 'set of consumables’ which could be linked to catalog users also associated with that dimension or axis.

Anti-constellations had not been operationalized in previous studies; however we wished to identify the potential anti-complementarity of goods and services. Absence was treated both explicitly and implicitly: interpretation of explicit non-consumption was based on 'non-usage’ of categories by groups of shoppers, and this information was included in the correspondence analysis of cross-tabulated data. Implicit non-consumption was derived from the absence of specific categories from the constellations.

The second aim was the interpretation of the consumption combinations, found in the quantitative data, within the framework of the model-building. It was expected that consumption constellations and anti-constellations would reflect the differing responses to the forces which influence consumption choices. Where influences were predominantly symbolic-expressive, then constellations would be expected to show a concern with symbolic consumption of goods and services, notably by the use of particular brands (e.g. Harrods store; Levi jeans). Where influences were predominantly functional-instrumental, we expected different consumption configurations would be found with the emphasis on factors such as value for money or shopping convenience. We also predicted that configurations would be influenced by needs (physiological/esteem/self-actualization) reflected in consumer choices. Where the consumption choices were central to self and identity, then the congruence between the series of consumption choices and self image would be important, shown by the selection of particular product and service (including retail) brands.

The second part of the research design involved seven semi-structured interviews with marketing managers from the in-home shopping industry; and one interview with marketing consultancy firm with expertise in the in-home shopping catalog industry. The objective was to use qualitative data to confirm the broader patterns of consumption found in the quantitative analysis.


[In view of space constraints only brief illustrative findings will be given here.]

Our findings provided confirmation of the relationship between social roles, social identity and consumption combinations; illustrated the differential impact of consumption forces on consumer choices; and indicated the key product categories in consumption combinations.

Correspondence analysis of the behavioral database showed evidence of constellations and anti-constellations of consumption from which consumption configurations could be derived. The qualitative data, collected during the second stage of the empirical study, confirmed the existence of identifiable and separable groups amongst consumers, in their roles as mail order shoppers: firstly Next Directory shoppers; secondly J.D.Williams shoppers; and thirdly users of the mainstream in-home shopping catalogs: Grattans, GUS, Kays, Littlewoods, Freemans and Empire Stores.

Consumption configurations

The consumption configuration of Next Directory shoppers was influenced primarily by the interaction of symbolic and expressive forces. The consumption constellation was characterised by extensive patronage of restaurants, branded clothing goods such as Levi 501 jeans and L.A. Gear, Reebok and Nike trainers, upmarket foreign automobiles, expensive electronic equipment, particularly cameras, as well as a range of telephonic machinery, and patronage of a range of fashionable retailers. The consumption anti-constellation included non patronage of mass market mail order catalogs as well as non use of package vacation trips, non patronage of food stores such as Gateway, Morrisons and Kwiksave for grocery shopping, non purchase of such women’s magazines as Woman, Woman’s Own and Woman’s Journal and non readership of the tabloid press. The emphasis in many of the consumer choices was on meeting such psychological needs as esteem and self-actualization. This was an important influence on the formation of the consumption configuration for Next Directory shoppers.



Product categories

Evidence was found to support the choice of product categories associated with particular roles (and identified by Solomon et al in earlier studies, 1987, 1988, 1991) to explore cumulative consumption including: automobiles, clothing and electronic goods. In this study automobiles had the potential to reflect the different interaction of the forces which influence consumption among groups. Other important product categories included electronic goods, clothing and discretionary use of time (e.g. use of leisure time) and retail patronage. The qualitative data also confirmed the importance of understanding brand choices within the stream of decisions made by the consumer.


Evidence was found of broader patterns in the stream of consumers’ decisions. The quantitative analysis of the behavioral data demonstrated a number of consumption configurations which could be linked with different groups of in-home catalog shoppers. A set of product categories, which had been successfully linked to social roles in earlier studies, were used to explore the broader patterns of consumption choices and non choices. Discriminators amongst constellations have been found, and these are particularly linked to branded clothing products, electronic products and the discretionary use of time and these can all be related to Solomon’s work on intermediat patterns of cumulative consumption. Retail outlets represented a new component within the consumption constellations and exhibited the potential to differentiate amongst the various groups.The other important product constellation was automobiles and this product category tended to overlap a number of profiles, suggesting that automobiles represented 'fuzzy sets’.


No study at this stage of exploratory research into the complex area of consumers’ cumulative consumption choices and non choices can offer a comprehensive picture of consumer profiles. There was only limited consideration of some product categories (e.g. branded clothing and electronic goods).


The creation of social and cultural meaning by consumption notably in the formation of tastes and distastes to mark inclusion and exclusion; and the creation of identity via material goods-such as branded jeans and trainers-as well as via the consumption of services such as leisure activities (e.g. pubs, discos, rave parties, football) are both areas which would reward further study. An important framework has been proposed within which to explore the broader patterns of consumption and the linkages among consumers’ product and service choices, self images and social identities.


[a full set of references is available on request].

Dittmar, Helga (1992): The Social Psychology of Material Possessions: To Have Is To Be. Harvester Wheatsheaf, Hemel Hempstead.

Solomon, Michael R. (1988): "Mapping Product Constellations: A Social Categorization Approach to Consumption Symbolism", Psychology and Marketing 5 (3, Fall), 233-258.

Wilk, Richard R. (1994): "'I hate Pizza’: Distaste and Dislike in the Consuming Lives of Belizeans," paper read at American Anthropological Association Meeting, November, Atlanta.

Wilk, Richard R. (1995): "Learning Distaste: The Social Importance of Not-Wanting", conference paper read at "Learning to Consume", Lund University, August.