Being Like Or Being Liked: Identity Vs. Approval in a Social Context

Susan Auty, Lancaster University Management School
Richard Elliott, Exeter University
[ to cite ]:
Susan Auty and Richard Elliott (2001) ,"Being Like Or Being Liked: Identity Vs. Approval in a Social Context", in NA - Advances in Consumer Research Volume 28, eds. Mary C. Gilly and Joan Meyers-Levy, Valdosta, GA : Association for Consumer Research, Pages: 235-241.

Advances in Consumer Research Volume 28, 2001     Pages 235-241

BEING LIKE OR BEING LIKED: IDENTITY VS. APPROVAL IN A SOCIAL CONTEXT

Susan Auty, Lancaster University Management School

Richard Elliott, Exeter University

This study adopts a sociological approach to adolescent consumer behaviour to examine normative influence on fashion brand choice. Bearden’s social influence scale (1989) is applied to the choice of sports footwear brands by teenagers. The informational factor is found to be hardly relevant, and the normative factor divides into two componentsCidentity and compliance. Modifying the concept of compliance to the rather softer notion of 'approval’ results in a scale (based on Bearden’s original) that significantly discriminates between fashion brand buyers and others. The need to be liked by one’s peers appears to be a more important driver of choice than the need to express one’s identity with them.

INTRODUCTION

Researchers into the psychology of consumer behaviour generally agree that for products with a high degree of shared product meanings, product choice depends largely on the consumer’s desire to be associated with those meanings (Douglas and Isherwood 1979; Bourdieu, 1984; Mayer and Belk 1985; McCracken and Roth 1989; Dittmar 1992). Differences among individuals are expressed by alignment with particular images. Recent research has emphasised the identity-construction role of fashion: 'People buy oods solely to be different from others’ (Gabriel and Lang 1995). Thompson and Haytko (1997) state that 'through this logic of self-identity construction, the sense of 'who I am’ is constantly defined and redefined through perceived contrasts to others.’ Bauman (1990) introduced the notion of 'neo-tribes’ to characterise the changing nature of social groups which in contemporary society are based on 'individual acts of self-identification. . . . Tribes exist solely by individual decisions to sport the symbolic traits of tribal allegiance.’ The emphasis therefore is on individual 'self-construction efforts’. But do these self-constructions at the individual level depend on self-definition through likeness to some and difference from others or on the need for social approval? Bauman (1991) acknowledges that the 'postmodern celebration of difference and contingency has not displaced the modern lust for uniformity and certainty.’ As Elliott (1998) notes, 'the development of individual self-identity is inseparable from the parallel development of collective social identity’ and 'self-identity must be validated through social interaction’. The distinction between identity construction and the validation process appears to be an important one, especially in the case of adolescent fashion products, where the need to define an identity is not necessarily the same as the need for social approval.

Translated into the scheme of consumption values devised by Sheth, Newman and Gross (1991), the distinction is between social values, where association with a specific social group drives the choice, and emotional values, where one is satisfying the feeling of security or happiness in being positively perceived by others. Although much research has focused on identity-seeking as an explanation of choice behaviour, approval-seeking has recently been subsumed into discussions on identity: 'Identity then does not mean the creation and projection of any image, but of one that commands respect and self-respect’ (Gabriel and Lang 1995). The distinction made by Kelman (1961) between identification and compliance has effectively been lost, yet their goal orientations are quite different (self-maintenance or enrichment vs. external reward): 'Compliance can be said to occur when an individual accepts influence from another person or from a group because he hopes to achieve a favorable reaction from the other.’ Warde (1997) notes that 'preoccupation with identity value acquired through the purchase of commodities’ has 'arguably . . . been emphasized unduly, thereby encouraging the view that consumer behaviour has primarily been driven by processes of individualization and stylization.’

Identification (according to Burnkrant and Cousineau 1975) is unrelated to the 'visibility’ of a conforming behaviour in contrast to compliance, where 'the individual in a product evaluation situation would be expected to comply with the prior evaluations of others only where his evaluation is visible to others who are perceived by him as mediators of significant rewards or punishments.’ Hence conformity to prevailing fashion is most likely motivated by compliance rather than identity-seeking. Among adolescents especially compliance takes precedence over identity. According to Erikson (1965) 'the growing and developing youths, faced with this physiological revolution within them, . . . are now primarily concerned with what they appear to be in the eyes of others as compared with what they feel they are’. Psychologists agree that 'the conformist stage normally emerges during early adolescence and may be retained through late adolescent and adult life’ (Kroger 1996).

If we widen the context of consumer behaviour from the psychological level to encompass more social and cultural perspectives then it becomes evident that recent developments in social theory have movedfrom conceptualising consumption as a prime force for individualisation towards its playing a more social role. Poststructuralist cultural theory has conceptualised consumption as culturally dependent, that is, consumer choice depends on contextualised characteristics rather than on cognitive traits (Holt 1997), and in opposition to the use of fashion for self-identity construction, it is also used for purposes of the construction of social affiliation and 'to foster an affirming sense of social belonging’ (Thompson and Haytko 1997). Anthropological approaches to consumer behaviour are also beginning to highlight its pro-social functions, where the purchasing of goods is almost always linked to other social relationships (Miller 1998). Maffesoli (1996) focuses on consumption as playing a prime role in building social cohesion, suggesting that we live in 'the time of the tribes’, new social collectivities which attempt to counter the rise of 'identity politics’ and 'lifestyle cultures’, while Cova (1997) identifies the 'linking value’ which supports the social link between persons. This is largely a rediscovery of earlier theoretical approaches to consumption; the role of fashion as not only 'differentiating’ but also 'socialising’ was pointed out nearly a hundred years ago by Simmel (1904).

Firat (1992) observes that 'consumers of postmodernity [seem to be] no longer seeking centered, unified characters, but increasingly seeking to 'feel good’ in separate, different moments by acquiring self images that make them marketable, likable and/or desirable in each moment’, implying that identity-seeking has become secondary to approval-seeking in consumption behaviour. Indeed, for Firat and Venkatesh (1993) consumers do not have an identity, but rather 'assume different images and personalities in different situations to make themselves acceptable in each case’. In this view, identity-seeking has become subsumed in approval-seeking rather than the other way round.

CONCEPTUAL AND MEASUREMENT ISSUES

Social psychologists have long distinguished between informative and normative influences on purchase behaviour, and in the 1970s Burnkrant and Cousineau (1975) looked specifically at reference group influence on brand choices and concluded that 'after observing others evaluating a product favorably, people perceive the product more favorably themselves than they would have in the absence of this observation. . . .People may frequently buy products that others in their groups buy, not to establish some self fulfilling role relationship to the others nor to obtain some reward or avoid some punishment mediated by the others, but rather to acquire what they perceive as a good product.’ They suggest, indeed, that the main influence is informational rather than normative.

A separate stream of research began to try to identify personality differences that make certain people more susceptible to normative influence than others. Snyder’s Self Monitoring Scale (1974) grew out of this research; this attempted to distinguish people on the basis of adaptive behaviour in social situations. Self-monitoring was found to have a modest correlation with purchase behaviour (Auty and Elliott 1998, 1999). Bearden and Etzel (1982) pursued the original motivations of Deutsch and Gerard (1955) as explanatory variables of behaviour. These in turn had been modified by Stafford (1966), who identified knowledge, affectivity (identity) and sanctions (rewards). In devising a scale to measre Consumer Susceptibility to Interpersonal Influence (SII), Bearden, Netemeyer and Teel (1989) started out with the intention of developing a three-factor instrument in keeping with the distinction between 'value expressive’ (identity-seeking) and 'utilitarian’ (compliance) behaviour, in addition to information-seeking behaviour. Their analysis of the data indicated a very high intercorrelation of the value expressive and utilitarian factors, leading them to combine the two into a single 'normative’ factor. However, in the postmodern/poststructuralist cultural context, a 3-factor scale may be more appropriate, with the factors preserving both an identity element and a social approval element. Defining the normative factor more accurately allows marketers to understand what is actually driving the choice in products with shared symbolic meanings. Indeed, practitioners concerned with modelling the emotional attributes of brands (Morgan, 1998) have identified three components of brand affinity: the first (less relevant here) is Authority, which refers to the heritage and trust associated with the brand; the second and third are Identification and Approval, which clearly discriminate between the private meaning and the public prestige of the brand. According to Morgan (1998), the approval element is particularly important 'in market sectors where the visible consumption of a brand is "saying something" about the chooser.’ As Netemeyer, Bearden and Teel (1992) note, susceptibility to interpersonal influence 'is enhanced when individuals are highly concerned with the inferences others make or may make regarding their behavior’ or, in other words, the compliance factor may be more important in consumption behaviour than identity-seeking is.

The present research is designed to explore the distinction in today’s society and to understand current purchase motivations of adolescent consumers of symbolic products. In an attempt to discriminate between the motivations of top brand vs. lesser or store brand buyers, the SII Scale was administered together with questions on recent purchases of sports footwear and the associated meanings of the top brands.

HYPOTHESES

The following hypotheses represent expected findings in keeping with previous research into social influence.

H1    The stronger the desire to identify with one’s peer group the more likely one is to purchase one of the top three brands in a fashion category.

H2    The stronger the tendency to comply with group norms the more likely one is to purchase one of the top three fashion brands.

H3    The stronger the tendency to seek information from peer groups, the more likely one is to purchase one of the top three fashion brands.

The next two hypotheses are designed to explore demographic differences with regard to social influence. Previous studies have indicated that females are orientated towards 'an affiliation with disparate parties and attachment of self and other’ (Meyers-Levy and Sternthal 1991) suggesting that girls will be more susceptible to influence than boys. This is also the conclusion of a meta-analysis which showed that "women are somewhat more easily influenced than men, particularly in situations where other group members exert pressure on women to change their minds" (Matlin 1987), although the actual size of the sex difference was small. In keeping with Erikson’s (1965) research into the developing adolescent, it is suggested that even in the narrow age band of middle to late adolescence, a difference in susceptibility will be noticeable.

H4    Girls are more likely than boys to be influenced by peer groups in choosing fashion brands.

H5    High school age adolescents are more likely than college age adolescents to be influenced by peer groups in choosing fashion brands.

A proposition has been put forward to reflect the expectation that in today’s society, the approval of peer groups (which have the sanction of ridicule at their disposal) is a more salient motivation than shaping an identity is.

P1    A unidimensional scale measuring tendency to seek approval of peer groups is more closely associated with brand purchase than the existing two dimensional scale measuring normative and informational influences.

METHODOLOGY

After a series of focus groups at local schools and with university students to determine usage patterns of trainers and brand attitudes, a survey was designed to allow quantitative analysis. Adolescents aged between 15 and 19 were interviewed in dispersed shopping precincts in the United Kingdom, resulting in a convenience sample of 555 young men and women who responded to the full questionnaire. The survey asked for information on sports played and sports trainers owned. Information was elicited on the usage of particular brands for particular sports and for general street wear. Criteria for choosing brands were established with a eries of itemised-scale ratings, and attitudes to the three largest-selling brands (Nike, Adidas and Reebok) were obtained with Likert scale statements for each. Open-ended questions asked for 'one word that comes to mind’ for each of the three brands. Finally, respondents were handed Bearden’s Susceptibility to Interpersonal Influence (SII) Scale (using 5 scale points) for self-completion. Three additional questions were included to ascertain the influence of films, sports stars and top bands on brand choice, but these have been excluded from the current analysis.

TABLE 1

ROTATED COMPONENT MATRIX (LOADINGS OVER .2)

THE SAMPLE

55% of the sample of British adolescents were between 15-17 years old (secondary school age) and the remainder were 18-19 (tertiary education age). 50% were male. The modal figure for number of pairs of trainers worn regularly was two, while the range of the pairs owned was from 1-15. The median cost of the most recently purchased pair was ,43 in a range from ,5-,120. 33% had purchased Nike, 22% Adidas and 16% Reebok. The latest quoted market shares for the UK and Europe are 35% Nike, 24% Adidas and 14% Reebok (Marketing 22 April 1999). Market shares for the specific age range were not available, but a split sample analysis of the ownership variables showed no significant differences, giving confidence in the results.

ANALYSIS

The survey was analysed using SPSS, and included factor analysis of the SII scale and analysis of variance of brand attitudes, behavioural and demographic characteristics. The behavioural predictors were the number of pairs of sports trainers owned, and the brand and cost of the most recent pair purchased. Brands were grouped for ease of interpretation into the 'top three’ (Nike, Adidas and Reebok); 'lesser brands’ (eg. Fila, Ellesse) and 'unfashionable’ (eg. store brands). Cost of last pair was divided into two groups by the median expenditure (,45). Recency of last purchase was classified according to purchase within the past month, past year or earlier. Respondents were assigned to three groups based on their scores on the relevant items of the SII scale; thus for each manipulation, there were three groups representing high, medium and low scores on a particular factor. The effect of age was considered by dividing the sample into high school and post high school age (15-17 and 18-19). One-way analysis of variance was performed using these groups and the behavioural predictors.

THE FINDINGS

Factor Analysis

Factor analysis of the SII scale with a varimax rotation indicated a three-factor solution, where eigenvalues over 1 is the criterion. A scree-test pointed to a four factor solution, a not unexpected difference between the two extraction criteria (Hair, et al 1998), but this was rejected upon examination of the results, which split one of the items normally associated with information-seeking into its own factor. The three components may be interpreted as being related to identity-seeking, compliance-seeking and informational/consultative (see Table 1). The first factor (identity-seeking) accounted for the highest percentage of variance (39%), and total variance explained was 59%. Individual items, however, did not always fall into the expected category, and item 4 ('To make sure I buy the right product or brand, I often observe what others are buying and using’), suffers from multiple loadings, although the original scale has it clearly in the informational camp. In the current study, this item loads over .5 only in the 'compliance-seeking’ factor. Even more messily, item 5 ('I rarely purchase the latest fashion styles until I am sure my friends approve of them’) loads clearly into the 'identity-seeking’ category, rather than the expected compliance category, while item 1 ('I often consult other people to help me choose the best brand of a particular product’) loads almost equally on informational and identity-seeking, and below .5 in both cases.

TABLE 2

PATTERN MATRIX (OBLIMIN ROTATION WITH 2 COMPONENTS EXTRACTED; LOADINGS OVER .2)

TABLE 3

RELIABILITY COMPARISONS OF SII FACTOR (CRONBACH'S ALPHA)

Replication using the oblimin rotation (as performed in the two earlier studies) indicates that neither items 1 and 4 load above .5 on any category in the 3-factor solution. Correlation between the two normative components was .47, much lower than the 0.92 initially found by Bearden. When analysis was restricted to extracting two components, as in the earlier studies (see Table 2), item 1 remained highly ambiguous, loading almost equally on both factors. Item 4 is firmly in the normative category, rather than the expected informational. Correlation between the two factors was .34, the same as that found by Shroeder in his 1996 replication of Bearden’s work (.44 in Bearden’s study). Removal of items 1 and 4 result in greater clarity whatever method is used, and should perhaps be considered by future researchers.

Analysis of the younger (15-16 year olds) and oldest (19 year olds) was performed separately to see if older adolescents were perhaps more secure in their identity and therefore more likely to make a distinction between identity and compliance. The only difference, however, was found in Item 1 (an informational item), which is more clearly associated with the Identity component among the older age group. The variance extracted for the two normative components was virtually identical for both age groups.

Scale Reliability

Overall, the scale achieves an alpha of .85. The three components achieve .80, .77 and .61. When the items wereanalysed in their original categories, comparison with Bearden’s components shows that reliability of the informational factor is still quite low at .63, especially relative to previous applications (see Table 3). Our sample was younger in age and British, either of which may have affected results, but it does suggest that the informational factor is not robust across experimental conditions. In contrast, the normative factor seems stable as well as reliable. However, the question remains about how effective it is for explaining brand-buying behaviour.

One-way analysis of variance

One-way analysis of variance of the behavioural variables indicated that social influence is closely associated with ownership of the top brands, although the informational factor was not significant at the 95% confidence level (see Table 4). Thus, H1 and H2 are supported, but H3 is not. In order to more clearly isolate the compliance factor from the identity factor, a new scale of seven items (Cronbach’s alpha = 80) comprising those items in which the wording refers to peer approval was tested. The items were 3,4,5,8,9,11,12. These were mostly drawn from the original normative factor, but excluded the two items that refer specifically to identity, items 2 and 6, and included the informational item (4) about observing others to buy the 'right’ brand. Only this 'approval’ factor was aligned with the cost of the last pair (p<.035), with those most seeking approval being more likely to spend more than the median amount. As might be expected, the number of pairs owned and recency of last purchase were not associated with social influence.

TABLE 4

F-STATISTIC (p<) OF BEHAVIOURAL VARIABLES WITH SII VARIABLES

TABLE 5

F-STATISTIC (p<) OF DEMOGRAPHIC VARIABLES WITH SII VARIABLES

TABLE 6

F-STATISTIC (p<) OF DEMOGRAPHIC VARIABLES WITH BEHAVIOURAL VARIABLES

The same analysis of sex and age groups by different factors of social influence indicate that the sex is only significant for information-seeking (see Table 5): girls tend to seek information more than boys do. With regard to age, however, the story is reversed. Younger (high school age compared with college age) adolescents are more susceptible to normative social influence, regardless of whether one is looking specifically at identity-seeking or approval-seeking. However, they are not significantly more likely to seek information than older adolescents. Thus H4 is only partially supported and H5 is supported, with the proviso that information-seeking does not play a large role in influencing them.

In assessing the effects of social influence, it is important to rule out an association between demographics and brands purchased. The analysis using sex and age and all the behavioural variables was carried out to see if other explanations than social influence were associated with purchases (see Table 6). Looking at the relationship between sex and brands purchased, one finds no significant relationship. However, there is a significant relationship between sex and number of pairs owned, recency of last purchase and price paid for last pair, with boys owning more pairs, buying more frequently and spending significantly more than girls. The nature of the product suggests that this finding cannot be generalised beyond sports trainers: boys are more likely to take sports seriously and to require specialis trainers than girls are. Indeed, inspection of the brands owned for sports rather than street wear shows that a large number of boys own Adidas trainers exclusively for sports (much more so than girls own Reeboks just for sports). With regard to age, younger adolescents are barely significantly more likely to buy fashionable brands than older ones are, and are more likely to buy frequently than olderCperhaps because their feet are still growing! There was no significant interaction between age and sex with regard to number of pairs owned. None of the demographic findings argue against the interpretations of the social influence factors.

DISCUSSION

Bearden’s scale is a useful construct for its intended purpose, but when looking specifically at adolescent conformity in fashion consumption, it would appear that only half the scale is required. The 'approval’ subscale seems to predict actual behaviour better than the full SII scale, as measured by choice of brand in most recent purchase of a visible fashion item (sports footwear). There was a highly significant association (p< .000) between a high score on these items and choice of the top three brands. In contrast, the informational element of the scale was not at all associated with brand choice, in that those scoring high on the need to consult were not more likely to choose the top three brands than others were.

The normative factor of Bearden’s social influence scale divides into two components, identity and compliance, which have become blurred in recent research into consumer behaviour. Modifying the concept of compliance, with its connotations of reward and punishment, to a less disciplinarian concept of 'approval’, seems to define the motivation even more meaningfully and to result in a more significant discrimination between fashion brand buyers and others. Scores on an 'approval-seeking’ scale showed more association with actual purchase behaviour (brand last bought and cost) than did the identity-seeking scale, although this modified scale admittedly includes some items that load strongly on identity-seeking in the three factor solution of the SII scale.

The distinction first made from a sociological perspective by Mead in 1934 has most relevance here: the 'me’ aspect of identity ('the self’s image of the others’ image of itself’ in Bauman’s (1990) words) is predominant among adolescents in choosing fashion brands over the 'I’ aspect (the self’s 'inner core’). McCracken’s (1986) emphasis on the transfer of symbolic meanings, the 'systematic appropriation of the meaningful properties of goods’, from the group to the self, which has been embraced by Dittmar (1992), does not explain adolescent brand behaviour so much as the human need for social approval does: teenagers are not trying to be like other people so much as trying to be liked by them. Dittmar acknowledges this when she says that: 'Many aspects of self-definition become associated with people’s possessions because they reflexively evaluate themselves on the basis of how others perceive them and respond to them in terms of material symbols.’ That identity and compliance motives often coalesce is not surprising, given that the apparently simplest way to gain approval is to be like the people one chooses to be liked by. It is important, however, not to lose the approval motivation in elucidating the identity-seeking motivation.

IMPLICATIONS FOR FUTURE RESEARCH

This study has attempted to broaden the context of consumer behaviour from the individual towards a more socially-situated person, but this is only a first step towards locating choice behaviour in a truly cultural context. Although the conceptualisation of consumer behaviour as being driven by individual preferences has dominated the field since its inception, recent radical reframings of consumer choice suggest that rather than choice being a psychological variable it should be thought of as a cultural variable; 'the basic choice that a rational individual has to make is the choice about what kind of society to live in. According to that choice, the rest follows' (Douglas 1996). This extends the social affiliation role of brands to one of the symbolic expression of a basic cultural value system and points us away from simple preferences and towards the negative pole of 'hostility'. Rather than focus simply on what people want, it may be more informative to ask what they do not want, as people do not only seek social approval from some social groups, they also want to express hostility to other groups. We therefore need to enrich our conceptualisation of the symbolic meaning of brand choice as not just a badge of allegiance but also as an act of rejection and hostility. 'To understand shopping practices we need to trace standardised hates, which are much more constant and more revealing than desires' (Douglas 1996). The work on anti-choice behaviour (e.g. Hogg, 1998) has made a start in re-contextualising consumer choice, but despite being based on Bourdieu's (1984) sociological theory of 'refusal of taste' it has not embraced the cultural perspective taken by Douglas and remains focused on individuation. Future research should explore the extent to which affiliation, the building and maintenance of social relationships and the expression of cultural values are key factors in understanding brand choice.

REFERENCES

Auty, S and R Elliott (1998) "Fashion Involvement, Self-Monitoring and the Meaning of Brands," Journal of Product and Brand Management, 7, 109-121.

Auty, S and R Elliott (1999) ' "Reading" Advertising and "Writing" Identity' in Advances in Consumer Research, ed. E Arnould and L Scott, 26, pp. 439-444.

Bauman, Z (1990) Thinking Sociologically. Oxford: Basil Blackwell.

Bauman, Z (1991) Modernity and Ambivalence. Cambridge: Polity Press.

Bearden, W and M Etzel (1982) 'Reference Group Influence on Product and Brand Purchase Decisions', Journal of Consumer Research, 9, pp. 183-194.

Bearden, W, R Netemeyer and J Teel (1989) "Measurement of Consumer Susceptibility to Interpersonal Influence," Journal of Consumer Research, 15, pp. 473-481.

Bourdieu, P. (1984) Distinction: A Social Critique of the Judgement of Taste. London: Routledge.

Burnkrant, R and A Cousineau (1975) 'Informational and Normative Social Influence in Buyer Behavior, Journal of Consumer Research, 2, 206-215.

Cova, B (1997) 'Community and Consumption: Towards a Definition of the "Linking Value" of Products or Services', European Journal of Marketing, 31, pp. 297-316.

Deutsch, M and H Gerard (1955) 'A Study of Normative and Informational Social Influences Upon Individual Judgment', Journal of Abnormal and Social Psychology, 51, pp. 624-636.

Dittmar, H (1992) The Social Psychology of Material Possessions. Hemel Hempstead: Harvester Wheatsheaf.

Douglas, M. (1996), Thought styles: Critical Essays on Good Taste. London: Sage Publications

Douglas M and B Isherwood (1979) The World of Goods, London: Allen Lane.

Elliott, R (1998) 'A Model of Emotion-Driven Choice', Journal of Marketing Management, 14, pp. 95-108.

Erikson, E (1965) Childhood and Society, Harmondsworth: Penguin Books.

Firat, A F (1992) 'Fragmentations in the Postmodern' in Advances in Consumer Research, ed. I Sherry and B Sternthal, 19, pp. 203-205.

Firat, A F and Venkatesh (1993) 'Postmodernity: the Age of Marketing' International Journal of Research in Marketing, 10, pp. 227-249.

Gabriel, Y and T Lang (1995) The Unmanageable Consumer: Contemporary Consumption and its Fragmentation, London: Sage Publications.

Hair, J, R Anderson, R Tatham and W Black (1998) Multivariate Stafford, J (1966) 'Effects of Group Influence on Consumer Data Analysis, 5" ed. London: Prentice Hall International.

Hogg, M (1998), Anti-constellations: Conceptualisation and Critique. In B. Engel and A. Oloffson (eds.) European Advances in Consumer Research, 3, pp. 44-49.

Holt, D (1997), 'Poststructuralist Lifestyle Analysis: Conceptualising the Social Patterning of Consumption in Postmodernity', Journal of Consumer Research, 23, pp.326-350.

Kelman, H (1961) 'Processes of Opinion Change', Public Opinion Quarterly, 25, pp. 57-78.

Kroger, J (1996) Identity in Adolescence: The Balance Between Self and Other, 2" ed. London: Routledge.

Maffesoli, M (1996), The Time of the Tribes: The Decline of Individualism in Mass Societ.v. trans. D. Smith, London: Sage Publications.

Matlin, M (1987), The Psychology of Women. New York: Holt, Rinehart and Winston.

Mayer, R and R Belk (1985) 'Fashion and Impression Formation Among Children' in The Psychology of Fashion, ed. M. Solomon, Lexington MA: Lexington Books, pp. 293-307.

McCracken, G (1986) 'Culture and Consumption: A Theoretical Account of the Structure and Meaning of Consumer Goods,' Journal of Consumer Research, 13, pp. 71-84.

McCracken, G and V Roth (1989) 'Does Clothing Have a Code? Empirical Findings and Theoretical Implications in the Study of Clothing as a Means of Communication,' International Journal of Research in Marketing, 6, pp. 13-33.

Mead, G (1934) Mind, Self and Society. Chicago: Chicago University Press.

Meyers-Levy J and B Sternthal (1991) 'Gender Difference in the Use of Message Cues and Judgments, Journal of Marketing Research, 28, pp. 84-96.

Miller, D. (1998), A Theory of Shopping. Cambridge: Polity Press

Morgan, R (1998) 'Linking Brand Imagery with Preference' in Brand Choice Modelling, ed. I Greig, ESOMAR, pp.63-84.

Netemeyer, R, W Bearden and J Teel (1992) 'Consumer Susceptibility to Interpersonal Influence and Attributional Sensitivity' Psychology and Marketing, 9, pp. 379-394.

Cova, B. (1997) "Community and consumption: Towards a definition of the "linking value" of products or services", European Journal of Marketing, 31, 3/4, 297-316.

Simmel, G. (1904), "Fashion" Reprinted in American Journal of Sociology, 62 (May 1957),541-548.

Douglas, M. (1996), Thought styles: Critical Essays on Good Taste. London: Sage publications

Miller, D. (1998), A Theory of Shopping. Cambridge: Polity Press

Bourdieu, P. (1984), Distinction: A social critique of the judgement of taste. London: Routledge.

Sheth, J, B Newman and B Gross (1991) 'Why We Buy What We Buy: A Theory of Consumption Values' in Journal of Business Research, 22, pp. 159-170.

Schroeder, J (1996) 'An Analysis of the Consumer Susceptibility to Interpersonal Influence Scale', Journal of Social Behavior and Personality, 11, pp. 585-599.

Simmel. G (1904). 'Fashion' Reprinted in American Journal of Sociology, 62 (May 1957), pp. 541-548.

Snyder, M (1974) "Self- Monitoring of Expressive Behavior," Journal of Personality and Social Psychology, 30, pp. 526-537.

Stafford, J (1966) 'Effects of Group Influence on Consumer Brand Choice Preference, Journal of Marketing Research, 3, pp. 68-75.

Thompson, C and D Haytko (1997) 'Speaking of Fashion: Consumers’ Uses of Fashion Discourses and the Appropriation of Countervailing Cultural Meanings’ Journal of Consumer Research, 24 (June), pp 15-42.

Warde, A (1997) Consumption, Food and Taste, London: Sage Publications.

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