JOURNAL OF CONSUMER PSYCHOLOGY, 16(1), 79–91 Copyright © 2006, Lawrence Erlbaum Associates, Inc. KEEPING ROSA, GARBARINO, THE BODYMALTER IN MIND Keeping the Body in Mind: The Influence of Body Esteem and Body Boundary Aberration on Consumer Beliefs and Purchase Intentions José Antonio Rosa University of Illinois at Chicago Ellen C. Garbarino Case Western Reserve University Alan J. Malter University of Arizona This research explores the influence of consumers’ body-related information on beliefs and purchase intentions toward products for which the consumption experience is significantly and directly determined by body-related information (e.g., feel, fit, sense of safety) when the products are bought in body-absent purchase environments such as the Internet. We examine the effects of consumers’ body esteem (i.e., like or dislike of one’s body) and body boundary aberration (variation in the perceived location of the edges of one’s body) in the context of apparel purchases that are made on the Internet. Body esteem had a positive influence on involvement with apparel, and body boundary aberration had a negative influence on consumers’ overall concern with the fit of apparel. Involvement with apparel and overall concern with fit, in turn, significantly influenced consumers’ intentions to purchase apparel online. Consequently, consumers with high body esteem were less likely to buy on the Internet and those with high body boundary aberration were more likely to buy. Technology-driven changes in the marketing of products that are traditionally purchased based on feel and fit (e.g., apparel, sports equipment) are bringing to the foreground body-related (often body-generated) information that influences consumers. One important technology-driven change is Internet retailing. With over 40% of all U.S. consumers making at least one online purchase last year (Greenspan, 2004), and a predicted annual growth rate of 17% (Rush, 2004a), Internet shopping has become an important retail channel. Early online retailing focused on product categories with only tacit or limited body involvement such as books, music, and consumer electronics. However, some of the fastest growing online sales are for products that heavily involve the body (Rush, 2004b), most notably apparel. Although many products involve the body at some level, high Correspondence should be addressed to Ellen C. Garbarino, Case Western Reserve University, Weatherhead School of Management, 10900 Euclid Avenue, Cleveland, OH 44106. E-mail: ellen.garbarino@case.edu body-involving products are those for which the consumption experience is significantly and directly determined by body-related information, such as feel (softness, solidity, weight, gripability, etc.), fit (looseness, snugness, flow or constriction of movement), and even sense of safety (i.e., shielding from the elements). Examples include outerwear, fashion and work apparel, bicycles, helmets, sports equipment (tennis rackets, baseball gloves, etc.), furniture, and automobiles. For such high body-involving products, the quality of the consumption experience is directly determined by how the product interacts with the consumer’s body. The lack of direct experiential information increases the apparent risk of purchasing such products in body absent environments such as telephone shopping (Cox & Rich, 1964), mail-order catalogs (Spence, Engel, & Blackwell, 1970), and Internet retailing (Alba et al., 1997). Thus, consumers with a higher need for tactile input for product evaluation appear to make fewer Internet purchases of products such as clothing and flowers (Citrin, Stem, Spangenberg, & Clark, 2003). 80 ROSA, GARBARINO, MALTER Moreover, many consumers prefer shopping for such products in environments that allow for tactile or sensory evaluation prior to choice (McCabe & Nowlis, 2003). The introduction of virtual object-interactive features, for example, increased online purchase intentions for low body involvement products such as digital cameras (Schlosser, 2003). In contrast, the use of virtual three-dimensional product visualization tools (versus two-dimensional product representations) did little to enhance consumer learning or decision quality for products where tactile manipulation is important (Li, Daugherty, & Biocca, 2003). Hence, there appears to be something about high body-involvement products that poses challenges to body-absent purchasing. We propose that a critical missing element in previous studies is the role of body-related information in evaluating such products. The economic and social importance of high body-involving products and their growing online sales volumes draw attention to a necessary but still insufficient understanding of how consumers’ body-related information influences consumers’ beliefs and purchase decisions. When the marketing of high body-involving products occurs solely in a brick-and-mortar world, it is possible to take the influence of body-related information for granted, because consumers tacitly integrate body-related information and more abstract considerations (e.g., price, quality, reliability) into holistic assessments. In traditional retail environments, consumers can try on and assess the overall fit of garments, test drive automobiles, swing tennis rackets, and ride bicycles in store parking lots. In doing so, they seemlessly combine body-related information with abstract knowledge. With the growing use of body-absent environments such as the Internet, however, it becomes more important to understand the unique influence of body-related information on product evaluations, purchase decisions, and satisfaction. Our study focuses on two factors—body esteem (Franzoi & Shields, 1984; Secord & Jourard, 1953) and body boundary aberration (e.g., Fisher, 1986)—which have been linked empirically and conceptually to consumer purchase decisions (Ferraro, Shiv, & Bettman, 2005; Rosa & Malter, 2003). Body esteem is a deeply held and generalized like or dislike of one’s body, which is related to but conceptually distinct from the self-esteem (e.g., Heatherton & Polivy, 1991; Schouten, 1991), and which is manifest in an individual’s positive or negative feelings toward different aspects of his or her body (Franzoi & Herzog, 1986; Franzoi & Shields, 1984; Secord & Jourard, 1953). Body boundary aberration refers to the extent to which an individual’s perceptions of the edges of his or her body do not have a fixed location (Fisher, 1970, 1986). Every person has a sense of where his or her body ends and the rest of the world begins, but not everyone perceives their body boundary equally in terms of location or stability. Using a large-scale national sample in an online survey, we examined the influence of body esteem and body boundary aberration on consumers’ intentions to purchase apparel on the Internet, mediated or complemented by factors such as involvement with the product category, overall concern with fit, and satisfaction with prior online purchases. THEORETICAL BACKGROUND Body Esteem and Body Boundaries As Influencers of Consumer Intentions Different types of body-related information can shape beliefs and intentions toward products. Proprioceptive knowledge, for example, encompasses mental representations of actual and projected posture and movement (e.g., Graziano & Gross, 1998; Witt, Proffitt, & Epstein, 2004). Hence, it influences performance assessments of products for which consumption involves body motion, such as bicycles and tennis rackets. The information generated by the body during the use of such products is integral to an assessment of their adequacy and value. Thus, for example, many consumers tend to swing tennis rackets in store aisles and to have high regard for retailers who allow them to try out products before purchasing (e.g., Barbaro, 2004). In like manner, body esteem and body boundary aberration, as beliefs and perceptions that form a backdrop to self-concept (e.g., Fisher, 1986; Marsh, 1990), help to shape assessments of how products enhance or preserve personal well-being. Elements of self-concept have been shown to influence consumers’ beliefs and behaviors (e.g., Landon, 1974; Sirgy, 1982; Thompson & Hirschman, 1995). The specific influence of body esteem and body boundary aberration, however, has not been explored extensively. Body esteem. To reiterate, body esteem is a deeply held and generalized like or dislike of the body. It is composed of three correlated factors—physical condition, general (primarily facial) attractiveness, and physique appearance. The components of these factors tends to differ slightly for men and women (Franzoi & Herzog, 1986; Franzoi & Shields, 1984; Mahoney & Finch, 1976). This is not surprising, however, given that different body parts (e.g., eyes, hips, breasts, lips, biceps, hair) convey self-meaning for men and women. Body esteem should be distinguished from body image, which is the conceptual representation or frame that people hold of the physical self (e.g., Cash, 1990). Body image is a stored knowledge structure that includes perceptions of who and what we have been as well as how we see ourselves currently (Schilder, 1935/1950). Body esteem is also distinct from self esteem (e.g., Henriques & Calhoun, 1999; Lerner, Karabenick, & Stuart, 1973). Self esteem can be concerned with performance, social relations, and appearance (Heatherton & Polivy, 1991). Conceptually, body esteem is most strongly related to appearance esteem. It can be viewed as a component of self-esteem, but not identical to it. KEEPING THE BODY IN MIND Evidence that self-esteem affects consumption has been reported (Banister & Hogg, 2004), but the specific influence of body esteem on consumer beliefs and intentions has not been explored. Here we explore two lines of reasoning that lead to competing hypotheses. One possibility is that body esteem leads to higher consumer involvement with body-involving product categories such as apparel. That is to say, high body esteem consumers will be more interested in, and give greater importance to, body-involving product categories. Even high self-esteem individuals seek to continually maintain or enhance their self-concept (Epstein, 1980), and this can sometimes be done through their purchases (Sirgy, 1982). Moreover, research on the consumption of products that impinge directly on the body, such as cosmetic surgery, shows that consumers who like their appearance (i.e., likely have high body esteem) look for ways to enhance different body areas to affirm their self concept (Schouten, 1991; Thompson & Hirschman, 1995). It can thus be argued that consumers for whom body appearance is a significant component of self-esteem (e.g., high body esteem) see high body-involving products as potential affirmers of their self-concept, and consequently display more interest in and give greater importance to such products. Consumers with low body esteem, in contrast, are likely to exhibit low involvement with body-involving products because they do not see body-involving products as affirming their self-concept. These considerations imply that Internet shoppers with high body esteem will exhibit higher involvement with apparel than will shoppers with low body esteem. An alternative argument assumes that self-esteem is positively linked to self-confidence (Hermann, Leonardelli, & Arkin, 2002). To this extent, individuals whose high self-esteem is strongly shaped by high body esteem should be confident of their appearance, and as a result, care less about high body-involving products, because their appearance does not need bolstering. In contrast, low body esteem individuals should see high body-involving products as a means of bolstering appearance and building self esteem, and should consequently care more about such products. In the context of apparel purchases on the Internet, these considerations suggest that consumers with high body esteem will exhibit lower involvement with apparel than will those with low body esteem. The empirical testing will examine the appropriateness of these alternative possibilities. Body boundary aberration. As mentioned, body boundary aberration is the extent to which an individual’s perceptions of the edges of his or her body do not have a fixed location (Fisher, 1970, 1986). It was initially proposed as a research construct by neurologists (e.g., Head, 1926), who found that some brain-damaged patients could not recognize body parts as their own. Although the construct was initially used in research on psychologically disturbed patients (Schilder, 1935/1950), healthy consumers can also differ in 81 their perception of body boundary location (e.g., Fisher, 1970, 1986; Fisher & Cleveland, 1958). Perceived body boundaries can recede from the skin, or they can extend into immediately adjacent objects and surfaces. When boundaries recede, the individual may perceive areas of his or her body as being outside the body boundary and not part of the self (Freeman & Melges, 1977), and this may in turn lead to detachment and depersonalization of body parts (e.g., nose, ears, hands, feet, breasts). When boundaries are extended, they flow or interlace with objects that are in close physical proximity, such as when a tool being wielded becomes part of the hand, or an automobile being driven alters the individual’s sense of physical size (e.g., Graziano & Gross, 1995; Maravita & Iriki, 2004). Body boundary aberration can sometimes cause individuals to extend their boundaries into attached or worn objects and to alter their body schemas accordingly (e.g., Graziano & Gross, 1998). Examples of boundaries being extended and affecting body schemas include differences in self-assessments of reach and effectiveness based on whether a tool is held (Maravita & Iriki, 2004). Similarly, individuals who wear hats with feathers and tall crowns may alter their body schema and hence instinctively discern how far to duck when passing through doorways (Blakeslee, 2004). Another example is the sensation of “nakedness” that arises when one fails to wear a habitually-used item, such as a wedding ring or watch. In all of these instances, body boundaries have been extended to incorporate the items that are in contact with the body. Research has shown that body boundary aberration is correlated with the individual’s perceived vulnerability (see Fisher, 1986, Vol. 2). Individual-level variability in feeling protected when wearing a heavy coat, for example, where some consumers feel more secure and others feel no different, is most likely related to reduced perceived vulnerability that stems from reinforced boundaries (Fisher, 1973). The same research has also shown that body boundary aberration is something of which consumers are seldom aware, in contrast to the more consciously available sense of like or dislike that comprises body esteem. Moreover, research has been inconclusive as to whether the perceived vulnerability engendered by body boundary aberration leads to actively pursued behavioral strategies or to passive avoidance. On the one hand, body boundary aberration was found to be related to the color of apparel favored by women, with high boundary aberration women wearing bold colors (e.g., black, bright red) more often than low boundary aberration women (Fisher, 1970). To the extent that apparel color choices are made consciously to reduce perceived vulnerability, this would suggest an active use of high body-involving products to bolster body boundaries. On the other hand, research also suggests that perceived vulnerability stemming from body boundary aberration engenders discomfort that is sensed but not linked directly to the body, and that boundary aberration is not linked directly 82 ROSA, GARBARINO, MALTER to self-esteem (see Fisher, 1986). It is possible, in fact, for individuals to like their bodies (e.g., have high body esteem) and still exhibit high body boundary aberration. The separation of body esteem from body boundary aberration, along with typical consumer unawareness of body boundary aberration, would lead us to expect that the discomfort arising from boundary aberration might not be actively addressed through purchase behaviors. Rather, high body boundary aberration could cause consumers to avoid thinking about their bodies and things related to them. Different body-involving products touch on body boundaries in different ways, some by extending body boundaries in a localized area of the body (e.g., a hand’s perceived dimensions being altered by holding a tool), and others through a more generalized influence on boundaries (e.g., sense of size being altered when in an automobile). In the case of apparel, we expect that the influence will be along the parts of the body touched by the apparel, and that the influence of body boundary aberration will be manifested in part through consumers’ concern with how garments fit their bodies and align with their personal style. When garments fit well and look good, they better define body boundaries. Consequently, if consumers with high body boundary aberration actively seek to reduce their perceived vulnerability through apparel, they should exhibit higher overall concern with fit than consumers with low body boundary aberration (lower vulnerability). On the other hand, suppose consumers with high body boundary aberration seek to reduce their discomfort by avoiding thinking about their bodies and things related to them. If this were the case, they should exhibit lower overall concern with fit than consumers who have low body boundary aberration. These contrasting possibilities are evaluated in the reported research. Controlling for Other Sources of Influence on Apparel Purchase Intentions We have posited thus far that body esteem and body boundary aberration influence different aspects of consumer decision making, with body esteem affecting involvement with apparel and body boundary aberration affecting overall concern with fit.1 To evaluate these influences in a realistic consumer context, we incorporated the proposed effects into a framework that includes other factors known to influence consumer purchase intentions. For example, we assume that consumers’ intentions to make online purchases of apparel are negatively influenced by their concern with fit and positively influenced by their satisfaction with previous online purchases (Anderson & 1We focused on online apparel retailing because it enables a test of the expected effects in an environment that demands consumer action before concerns with fit can be allayed (Alba et al., 1997). The extension of these effects to other body-involving products, although theoretically plausible, remains to be investigated. FIGURE 1 Body esteem and body boundary aberration influence purchase intentions mediated by involvement with apparel and overall concern with fit. Sullivan, 1993; Jones & Sasser, 1995). We further expect that satisfaction will have an indirect effect on purchase intentions through its mediating influence on consumers’ involvement and their consequent concern with fit. (For evidence that satisfaction in a particular domain stimulates interest in products, see Oliver & Bearden, 1983; Shaffer, Root, & Sherrell, 1997.) In addition, we expect involvement to positively influence overall concern with fit. These effects, along with the effects of body esteem and body boundary aberration, are summarized in Figure 1. Gender Differences in Body-Related Information Due to the prevalence of culturally driven body ideals, it is generally believed that men and women differ substantially in how they relate to their bodies. However, although there is substantial evidence that men and women hold different attitudes toward their bodies (see Fiengold & Mazzella, 1998, for a meta-analysis), there is little evidence that these mean differences in attitude have significant impact on how body-related information is used (Koff, Rierdan, & Stubbs, 1990; Muth & Cash, 1997; Sondhaus, Kurtz, & Strube, 2001). Most of these researchers consistently find that women hold lower mean body attitudes than men but that these body attitudes have similar effects on other attitudes and behaviors across genders. Sondhaus et al. (2001), for example, found significant gender differences in mean body attitude for men and women but found that gender only moderates 2 of 13 other attitudes tested. In general, although there may be mean gender differences in body-related information, the effects of this information on attitudes and beliefs may be similar. Correspondingly, we expect some mean differences in body-related information for men and women but do not anticipate substantial differences in how this KEEPING THE BODY IN MIND body-related information would affect other beliefs and purchase intentions. METHOD An Internet-administered survey was used to assess the influence of body esteem and body boundary aberration on intentions to purchase rack apparel (as opposed to custom fit clothing) on the Internet. Respondents were recruited from a national panel designed to represent the U.S. Internet consumer population. A total of 668 respondents (547 women and 121 men) who had recently purchased rack apparel on the Internet were included. The demographics of the sample closely match the U.S. Internet consumer population (ClickZ Stats Staff, 2003; Pastore, 2001). Average online consumer population versus sample characteristics, respectively, were as follows: income—$49,800 versus $53,600; age—41 years old versus 44; education—college graduate versus college graduate; Web sites visited per week—13 versus 12; hours online per week—6.2 versus 10. All U.S. states and the District of Columbia were represented in the sample. The only variable that was poorly matched to the U.S. Internet population was gender (U.S. average, 52% female vs. sample 82%), possibly because of our focus on apparel shopping. Measures Past internet purchase behavior, involvement with apparel, concern with fit, and consumer purchase intentions data were gathered before the body esteem and body boundary aberration data to reduce potential demand effects. Potential gender differences were explored through a two-group test for invariance. To conduct gender comparisons with the small male subsample, parceling (Bandalos & Finney, 2001) was used on three sets of measurement variables: body esteem, body boundary aberration, and involvement with apparel. Parceling involves averaging two or more survey items related to the same construct, and using these parcel scores instead of the individual items when estimating the structural model. Bagozzi and Heatherton (1994) suggested that because parceling results in the estimation of fewer parameters, it allows for a better variable-to-sample-size ratio, which is appropriate for the smaller male subsample. They also suggested that a minimum of two parcels be used for each construct to account for measurement error. In this study, two parcels are derived for body esteem, body boundary aberration, and involvement with apparel. The composition and reliability indexes for the parcels are reported in their corresponding sections. Two individual measurement items are used for satisfaction with past purchases and purchase intentions, and four items are used for overall concern with fit. Body esteem. Using a well-established scale (Secord & Jourard, 1953), respondents were asked to indicate their 83 feelings about 34 parts or aspects of their bodies (e.g., physical stamina, nose, weight, hips, feet, etc.) on a scale from 1 (strong negative feelings) to 5 (strong positive feelings). Consistent with past research on body esteem (e.g., Franzoi & Herzog, 1986; Franzoi & Shields, 1984; Mahoney & Finch, 1976), we found three factors: physical condition, general (primarily facial) attractiveness, and body appearance, the composition of which differs for men and women. To retain consistency across male and female respondents for the structural equation analysis, parcels composed of items that were common across gender lines were derived for each of the three factors. Composite averages were derived for the parcels, and used as measures of body esteem in the structural equation analyses. The items composing each parcel are listed in the Appendix, with the excluded gender-unique items in parentheses. Reliabilities for the common item parcels are as follows: physical condition (α = .90), general attractiveness (α = .88), and body appearance (α = .93).2 Body boundary aberration. Respondents reported their agreement with 49 statements along a scale from 1 (strongly agree) to 5 (strongly disagree). Of these items, 10 pertained to body boundary aberration (Chapman, Chapman, & Raulin, 1978); these items were reversed scored such that higher scores signify higher body boundary aberration. For model estimation purposes, these ten items were aggregated into two parcels: body boundary aberration-extremities (α = .75) and body boundary aberration-body ownership and existence (α = .88). The extremities parcel includes items such as “I have felt as if I am united with objects near me” and “I have had the momentary feeling that things I touch remain attached to me.” The body ownership parcel includes items such as “At least once, I have wondered if my body was really my own” and “I have felt as if I could not distinguish my body from other objects around me.” All items are listed in the Appendix.3 Involvement with apparel. Using established measures (Mittal, 1995), five 7-point semantic differential scales were parceled into importance of apparel (α = .92) and concern with apparel (α = .88). Prior to parceling, items were rescaled as needed, such that larger numbers signify higher involvement. The scale items are listed in the Appendix. Overall concern with fit. Items specific to overall concern with fit in the context of Internet apparel purchases 2Reliability estimates for the full-item gender specific factors are almost identical to those reported earlier and are available from the authors. 3The Chapman, Chapman, and Raulin (1978) scale was developed for use with schizophrenia patients as an alternative to ink blot methods used in other research (e.g., Fisher 1970). A complete list of all filler items is available from the authors. Items not measuring boundary aberration assess deviant behaviors (e.g., “There have been times when I felt like rebelling against people in authority even though I knew they were right”), and were included to disguise the purpose of the measure. 84 ROSA, GARBARINO, MALTER were developed and pretested by the authors. Respondents were asked “How important are the following concerns in your decisions to buy or not buy off-the-rack apparel online?” along with nine statements (e.g., “can’t try it on,” “concern it will not fit right”).4 For each statement, respondents indicated importance on a scale from 1 (not at all important) to 7 (very important). The nine items were highly correlated (average correlation = .87) and could not be used simultaneously in the bootstrapping estimation phase of the structural equation analysis. As a result, four items that touch on different aspects of overall concern with fit were used as measures for the construct: “does it give the right impression,” “does it feel good,” “does it fit,” and “does it match my style.” Satisfaction with past online apparel purchases. Respondents were asked two questions about their off-the-rack apparel purchases: “How would you rate your satisfaction with your most recent online shopping experience?” and “How would you rate your satisfaction with the most recent product you purchased online?” Both were measured on a scale from 1 (not at all satisfied) to 7 (very satisfied). The items were correlated .80 and were used as measures of satisfaction with past purchases. Purchase intentions. Respondents reported their willingness to buy off-the-rack apparel online along a scale from 1 (very unwilling) to 7 (very willing), and reported their likelihood of making a purchase during the coming year along a scale from 1 (very unlikely) to 7 (very likely). These items were correlated .66 and were averaged to form a measure of purchase intentions. RESULTS The relations shown in Figure 1 were tested as a structural equation model using maximum likelihood estimation and bootstrapping. The means, standard deviations and correlations used in the structural model are shown in Table 1. Measurement weights, estimated structural path coefficients, and measures of fit for the full sample are illustrated in Figure 2. Although the χ2 goodness-of-fit statistic for the best fit model is significant (χ2 = 162.3, df = 84, p = .000), several combinations of absolute and relative fit indexes (Hu and Bentler, 1999) suggest that the modeled relations are supported by the data. This model meets the standards proposed by Hu and Bentler (1999) of standardized root mean squared residual (SRMR) values below .08, in combination with values above .95 for the Tucker-Lewis Index (TLI) and normed fit index (NFI), and a root mean square error of approxima- 4The complete set of statements is available from the authors. tion (RMSEA) below .06. SRMR for the full model is .049, TLI is 0.981, NFI is 0.970, and RMSEA is 0.037, all exceeding the recommended thresholds.5 All of the relations shown in Figure 1 were highly significant. Body esteem (.143) had a positive influence on involvement with apparel. That is, the more individuals like their bodies, the higher their involvement with apparel. This is compatible with the idea that consumers seek to affirm their self-concept through the use of apparel. Body boundary aberration had a negative effect on overall concern with fit (–.204), indicating that consumers with high body boundary aberration display less overall concern with how apparel fits their body and aligns with their personal style than do consumers with low boundary aberration. Paradoxically, this negative influence of body boundary aberration on overall concern with fit suggests that high boundary aberration consumers are more likely to purchase rack apparel online than are low body boundary aberration consumers. As expected, involvement with apparel was positively related to overall concern with fit (.160), and concern with fit was negatively related (–.223) to intentions to purchase rack apparel online. Finally, consumer satisfaction with past online apparel purchases had a positive influence on both future online purchase intentions (.465) and involvement with apparel (.177). Consumers who have made satisfying online purchases of rack apparel are more strongly predisposed to making additional purchases, and are more interested in apparel. Alternative Models Several intuitively plausible alternative models were also tested. These models focused primarily on alternate paths by which body esteem and body boundary aberration could influence purchase intentions and related constructs. The rationale for the positive influence of body esteem on involvement with apparel that was revealed by the analysis is that people who like their bodies are interested in apparel because it affirms their positive self-concept, and that high involvement leads to higher concern with fit. It is possible, however, that high body esteem could actually lead directly to less overall concern with fit, if high body esteem consumers believe themselves to look good in anything. Such an influence would be manifest in a negative direct path between body esteem and concern with fit and no path from body esteem to involvement. Making such changes signifi- 5The structural equation modeling analysis was also conducted using body esteem measures based on the full-item gender-specific factors discussed earlier. Using the gender-specific packet composition resulted in virtually identical path coefficients and model fit statistics (e.g., χ2 = 162.2 versus χ2 = 162.3). Alternate model parameters are available from the authors. Analysis of the residuals and modification indexes does not point to unspecified paths that would significantly improve the model. TABLE 1 Means, Standard Deviations, and Correlations for Measured Variables 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. Physical condition General attractiveness Body appearance Apparel is important Concern with apparel Does it make the right impression? Does it feel good? Does it fit? Does it match my style? Body boundary—extremities Body boundary—body own? Satisfaction with transaction Satisfaction with product Willing to buy Likely to buy M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 3.23 3.58 2.66 5.37 5.41 3.91 4.88 5.15 4.20 3.59 3.87 6.28 6.36 5.92 5.83 0.88 0.70 1.05 1.40 1.33 1.98 1.75 1.70 1.93 0.85 1.04 1.21 1.20 1.52 1.63 1 .584** .704** .134** .104** .038 –.031 .042 –.015 .015 .055 .052 .024 .031 .012 1 .464** .191** .188** .057 .058 .055 .053 .033 .065 .160** .128** .118** .129** 1 .118** .103** .095** .030 .073* .073 –.052 –.020 .004 .026 .001 –.027 1 .818** .140** .076* .084* .117** .036 .000 .148** .115** .092* .076* 1 .165** .103** .109** .137** .003 –.011 .174** .129** .120** .108** 1 .633** .574** .799** –.188** –.193** –.036 –.045 –.153** –.148** 1 .728** .706** –.073 –.073 –.064 –.066 –.159** –.129** 1 .619** –.072 –.087* –.071 –.080* –.212** –.181** 1 –.176** –.167** –.060 –.077** –.201 –.193** 1 .811** .087* .035 .101** .036 1 .106** .033 .126** .021 1 .799** .388** .351** 1 .337** .318** 1 .658** *Correlation is significant at the 0.05 level (two-tailed). **Correlation is significant at the 0.01 level (two-tailed). 85 86 ROSA, GARBARINO, MALTER FIGURE 2 Standardized path coefficients for the full sample. cantly reduces model fit (e.g., χ2 = 172.7, df = 84; proposed model χ2 = 162.3, df = 84),6 and produces a nonsignificant path between Body Esteem and Overall Concern with Fit (.009, ns). It is also possible that high body esteem consumers will have higher intentions to purchase, simply because they enjoy activities that involve the body. This idea can be tested by adding a direct path between body esteem and purchase intentions, while retaining the other paths. Adding this path to the model results in a nonsignificant improvement in model fit statistics (e.g., χ2 = 162.1, df = 83; proposed model χ2 = 162.3, df = 84), and produces a nonsignificant path between body esteem and purchase intentions (.029, ns). Finally, the speculation that body boundary aberration might have a negative impact on concern with fit was based on the assumption that body boundary aberration generates discomfort that is not linked to the body, and that consumers respond to the discomfort by avoiding the direct consideration of factors that touch directly on their bodies, such as overall concern with the fit of apparel. If discomfort due to body boundary aberration is addressed actively, however, high body boundary aberration could lead consumers to purchase clothes as a way of buttressing weak boundaries, which 6Similar degradation is observed for other model fit statistics and not reported here due to space considerations. All fit statistics for all alternative models are available from the authors. should result in a direct positive path from body boundary aberration to purchase intentions without mediation through overall concern with fit. However, adding a path between boundary aberration and purchase intentions reduces the fit statistics (e.g., χ2 = 185.8, df = 84; proposed model χ2 = 162.3, df = 84), and produces a nonsignificant path between Body Boundary Aberration and Purchase Intentions (.050, ns). Thus, this reinforces that consumers do not actively seek (or avoid) apparel purchases in response to body boundary aberration. Based on these results, it appears that consumers with high body boundary aberration do not seek to use apparel to address the condition, but rather try to avoid thinking about it. We believe that the modeled relation between body esteem and involvement with apparel, and body boundary aberration and overall concern with fit, along with the indirect manner in which these body-related factors influence consumer purchase intentions, are accurate approximations of consumer beliefs and intentions.7 7Alternative models that include paths between other constructs, for example, direct paths between Overall Concern with Fit and Satisfaction (χ2 = 177.7, df = 84; proposed model χ2 = 162.3, df = 84), between Involvement and Purchase Intentions (χ2 = 172.1, df = 84; proposed model χ2 = 162.3, df = 84), and redirecting body esteem to be mediated by body boundary aberration (χ2 = 898.8, df = 84; proposed model χ2 = 162.3, df = 84), were also tested. In general, alternative models resulted in poorer fit statistics and nonsignificant path coefficients. Results for all tested alternative models are available from the authors. KEEPING THE BODY IN MIND 87 TABLE 2 Test for Multigroup Invariance: Fit Statistics Comparison χ2 Goodness of Fit Stage 1—Models Compared Unconstrained Constrained measurement weights Constrained measurement intercepts Constrained structural weights Stage 2—Unconstrained Variable Parameters Constrained Measurement Weights—Alla Unconstrained body esteem Unconstrained involvement with apparel Unconstrained body boundary aberration Unconstrained overall concern with fit Unconstrained satisfaction Unconstrained purchase intentions Constrained Measurement Intercepts—Alla Unconstrained body esteem Unconstrained involvement with apparel Unconstrained body boundary aberration Unconstrained overall concern with fit Unconstrained satisfaction Unconstrained purchase intentions Constrained Structural Weights—Alla Unconstrained body esteem → involvement Unconstrained satisfaction → involvement Unconstrained involvement → concern with fit Unconstrained body boundary aberration → concern with fit Unconstrained concern with fit → purchase intentions Unconstrained satisfaction → purchase intentions SRMR TLI NFI RMSEA χ2 = 265.8, df = 168, p = .000 χ2 = 283.2, df = 177, p = .000 χ2 = 360.6, df = 183, p = .000 χ2 = 288.8, df = 174, p = .000 .0928 .0936 .1117 .1198 .977 .976 .961 .974 .952 .948 .934 .947 .030 .030 .038 .031 χ2 = 283.2, df = 177, p = .000 χ2 = 279.9, df = 175, p = .000 χ2 = 283.1, df = 176, p = .000 χ2 = 283.2, df = 176, p = .000 χ2 = 280.3, df = 174, p = .000 χ2 = 272.7, df = 176, p = .000* χ2 = 282.8, df = 176, p = .000 χ2 = 360.6, df = 183, p = .000 χ2 = 318.5, df = 180, p = .000** χ2 = 315.6, df = 181, p = .000** χ2 = 356.1, df = 181, p = .000 χ2 = 355.5, df = 179, p = .000 χ2 = 358.8, df = 181, p = .000 χ2 = 356.2, df = 181, p = .000 χ2 = 288.8, df = 174, p = .000 χ2 = 283.2, df = 173, p = .000 χ2 = 287.9, df = 173, p = .000 χ2 = 287.4, df = 173, p = .000 χ2 = 284.5, df = 173, p = .000 χ2 = 288.5, df = 173, p = .000 χ2 = 278.5, df = 173, p = .000* .0936 .0934 .0936 .0936 .0934 .0935 .0936 .1117 .0901 .0963 .1114 .1105 .1118 .1108 .1198 .1097 .1188 .1155 .1171 .1207 .1144 .976 .976 .976 .976 .976 .976 .976 .961 .969 .970 .962 .961 .961 .962 .974 .975 .974 .974 .974 .973 .976 .948 .949 .948 .949 .949 .948 .948 .934 .942 .942 .935 .935 .935 .935 .947 .948 .948 .948 .948 .947 .949 .030 .030 .030 .030 .030 .030 .030 .038 .034 .034 .038 .038 .038 .038 .031 .031 .032 .032 .031 .032 .030 Note. SRMR = standardized root mean squared residual; TLI = Tucker-Lewis Index; NFI = normed fit index; RMSEA = root mean square error of approximation. aCompare fit results from parameter-unconstrained models to model where all intercepts or weights are constrained. *Significantly different from constrained model p < .01. **Significantly different from constrained model p < .001. Test for Gender Group Invariance There are three ways men and women might differ; they might have different mean levels of the various constructs (identified by the measurement intercepts), different measure compositions (identified by the measurement weights), and different relations among the constructs (identified by the structural weights). The significance of these potential differences between men and women were evaluated by splitting the sample and conducting a two-group test for invariance. The process involves two stages (Byrne, 2004). In Stage 1, a baseline unconstrained model, in which all estimated parameters (measurement and structural) are allowed to differ between the groups, is compared against models where each set of parameters (i.e., all measurement intercepts, all measurement weights, or all structural weights) is individually constrained. Fit statistics for the unconstrained and constrained models from Stage 1 are reported in the top section of Table 2. By comparing the change in fit statistics between the unconstrained and different constrained models, Stage 1 determines the source or sources of group variance, be they measurement intercepts, measurement weights, or structural weights. Higher χ2 goodness-of-fit statistics are expected from constraining different sets of parameters, and large shifts in the χ2 statistic point to significant contributors to group variance.8 Comparing the unconstrained and constrained models from our Stage 1 analysis suggests that measurement intercepts (∆ χ2 = 94.8, ∆ df = 15, p = .000), measurement weights (∆ χ2 = 17.4, ∆ df = 9, p = .043), and structural weights (∆ χ2 = 23.0, ∆ df = 6, p = .000) are significant sources of between-group variance. As would be expected from previous research, the largest source of gender differences is due to mean differences, as represented by the dramatic decline in fit when constraining the measurement intercepts, whereas the constraining of the measurement and structural weights produce only modest changes in fit. Given the significant Stage 1 differences, Stage 2 analysis was consequently performed for each set of parameters. Stage 2 of the group invariance analysis involves comparing 8Other relative and absolute measures of fit (i.e., standardized root mean squared residual, or SRMR; Tucker-Lewis Index, or TLI; normed fit index, or NFI; and root mean square error of approximation, or RMSEA) should change similarly, with SRMR and RMSEA increasing in value relative to the unconstrained model, whereas TLI and NFI decrease in value. 88 ROSA, GARBARINO, MALTER each constrained model from Stage 1 against models constraining each of the specific construct parameters (e.g., measurement weights for body esteem, measurement weights for body boundary aberration, measurement intercepts for body boundary aberration, etc.). Stage 2 pinpoints the constructs responsible for the group variances detected in Stage 1. Lower χ2 goodness-of-fit statistics are expected from removing constraints, and large shifts point to significant contributors to between-group variance. Results from Stage 2 analysis are reported in the lower sections of Table 2.9 Stage 2 results further suggest that the largest significant differences between men and women stem from different measurement intercepts for body esteem (∆ χ2 = 42.1, ∆ df = 3, p = .000) and involvement with apparel (∆ χ2 = 45.0, ∆ df = 2, p = .000). This is confirmed by comparing the mean values for body esteem and involvement with apparel for men and women. We find that men like their physical condition (men = 3.5, women = 3.2, p = .001) and body appearance (men = 3.0, women = 2.6, p = .000) significantly more than women. Conversely, we find that women consider apparel more important (men = 4.6, women = 5.5, p = .000) and are in general more concerned with apparel (men = 4.8, women = 5.6, p = .000) than men. Only 2 of the other 12 comparisons in Table 2 are statistically significant: men give more weight to satisfaction with the product when forming overall satisfaction than women (men = .867, women = .823), and satisfaction with past online apparel purchases has a weaker effect on purchase intentions for men than for women (men = .163, p = .076; women = .535, p = .000). These differences may help explain why male consumers’ satisfaction with past online apparel purchases does not have a significant effect on involvement with apparel (men = .097, p = .321; women = .199, p = .000) or purchase intentions (men = .163, p = .076; women = .535, p = .000), whereas it has a significant influence on both for female consumers (see Figure 3).10 One plausible explanation is that male consumers feel better about their bodies in general and hence do not expect as much from apparel purchases, although like women, they care about affirming their self-concept through apparel. Lower expectations, in turn, result in a lower retrospective influence from satisfaction on involvement with apparel, and in the purchase intentions of male consumers being shaped by factors other than satisfaction. For female consumers, apparel purchases that meet expectations (i.e., produce high satisfaction) elevate their involvement with apparel because of the strong affirmation to their self-concept that comes from such purchases. Their sense of satisfaction also leads to higher future purchase intentions. Our results only pertain to online purchases of apparel, however, and may not be applicable to non-Internet pur9Given the large number of post hoc comparisons, a more conservative alpha of .01 was used. 10The measurement weights for male and female subsamples are available from the authors. FIGURE 3 Standardized path coefficients for the male and female subsamples. chases. The inherent risks involved in purchasing apparel without trying it on may actually result in attenuated expectations, and hence exaggerate the feelings of satisfaction when apparel purchased online looks and fits well. Furthermore, it is important to note that for most of the modeled constructs, the differences between men and women are not significant, suggesting that the general influence of body-related information on consumers is likely to be similar for men and women. Clearly, more research is needed in this area. GENERAL DISCUSSION This study explored the influence of body-related information (body esteem and body boundary aberration) on the purchase of high body-involving products for which consumption is primarily determined by body-related information (apparel), in a context lacking in direct body-centric evaluation opportunities (the Internet). It might be possible to ig- KEEPING THE BODY IN MIND nore the importance of body-related information in traditional shopping environments, even if such information is being applied. The absence of body-centric evaluation opportunities in Internet shopping, however, brings its importance to the fore. Our study reveals that body esteem and body boundary aberration influence consumer involvement with apparel and overall concern with fit, which in turn shape purchase intentions for apparel sold on the Internet. We find that high body esteem has a positive influence on involvement with apparel, which in turn leads consumers to be more concerned with overall fit and subsequently lowers their online purchase intentions. Conversely, high body boundary aberration reduces overall concern with fit and consequently increases online purchase intentions. Satisfaction with past online apparel purchases increases consumer involvement with apparel and purchase intentions, although more for women than for men in our sample. In general, we find the results are robust across gender. Our study provides evidence of direct influence from two body-related factors that have thus far received limited attention in consumer research. Although some calls for additional research into the role of body-related information have been issued (e.g., Joy & Sherry, 2003; Rosa & Malter, 2003), progress has been slow. Our finding that body esteem and body boundary aberration can influence consumer beliefs about high body-involving products such as apparel enhances the field’s understanding of consumer behavior. Moreover, showing that both factors influence consumer intentions to purchase in the technology-rich realm of Internet retailing also contributes to our understanding of how new media may bring heretofore underappreciated factors into more prominent roles in the consumer’s decision processes. The statistically robust findings invite further exploration of these and other body-related factors and their influences. The negative influence of high body boundary aberration on overall concern with fit may help explain some product returns of Internet apparel purchases. Consumers with high body boundary aberration appear less concerned than low boundary aberration consumers when buying apparel online. However, they may be just as likely as other consumers to find fault if the apparel does not fit well when they actually try it on. Thus, consumers with high body boundary aberration are likely to be insufficiently concerned with fit when ordering clothes online, leading them to less-than-optimal decisions, lower postpurchase satisfaction, and higher product returns. This study is limited to apparel (a high body-involvement product category) and to online retailing (an inherently body-absent environment). It is possible that the noted effects are elicited, at least in part, by the idea of buying something very personal with only verbal descriptions and a digitized visual representation on which to base decisions. It would be informative to explore if these results are replicated with a range of body-involving products (e.g., books and automobiles) and if body-related information has unique or 89 similar, but as of yet undocumented, effects in environments that allow active body participation in product assessments (e.g., traditional retail). In the traditional retail context, the effects of concern for fit may be muted due to the ability to address them directly, or alternatively, they may be heightened by the salience of the fit experience. In any event, the potential influence of body-related information on consumer beliefs and intentions across a wider array of categories and contexts merits further investigation. An additional consideration in evaluating our results is that all effects are based on correlations at a single point in time. Longitudinal and experimental investigations of these phenomena might better identify causal relations and boundary conditions. Furthermore, respondents were generally familiar with the purchase of apparel on the Internet and thus may not be representative of all consumers. Restricting the sample to consumers who have bought apparel online may actually depress the demonstrated impact of body-related information, as these individuals have proven themselves willing to engage in body-absent purchases of apparel. Overall, our research provides insights into the operation of body-related factors that have not been studied extensively in consumer psychology. However, body esteem and body boundary aberration are only two types of body-related information that have been identified across the behavioral disciplines. Although our study provides empirical evidence that body-related information can be important to consumer beliefs and intentions toward body-involving products, future research should explore the effects of these and other types of body-related information on a wider range of body-involving products in a wider array of contexts. ACKNOWLEDGMENTS We thank Rick Bagozzi for comments on a prior version of this article, and the faculty and doctoral student attendees of the Research Seminar Series at Case Western Reserve University for their valuable observations. We thank the Journal of Consumer Psychology reviewers for their invaluable assistance in preparing this manuscript. We also thank the Weatherhead School of Management for financial support through the Summer Research Grants program. REFERENCES Alba, J., Lynch, J., Weitz, B., Janiszewski, C., Lutz, R., Sawyer, A., et al. (1997). Interactive home shopping: Consumer, retailer, and manufacturer incentives to participate in electronic marketplaces. Journal of Marketing, 61, 38–53. Anderson, E. W., & Sullivan, M. W. (1993). The antecedents and consequences of customer satisfaction for firms. Marketing Science, 12, 125–143. 90 ROSA, GARBARINO, MALTER Bagozzi, R. P., & Heatherton, T. F. (1994). A general approach for representing multifaceted personality constructs: Application to state self-esteem. Structural Equation Modeling, 1, 35–67. Bandalos, D. L., & Finney, S. J. (2001). Item parceling issues in structural equation modeling. In G. A. Marcoulides & R. E. Schumacker (Eds.), New developments and techniques in structural equation modeling (pp. 269–296). Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Banister, E. N., & Hogg, M. K. (2004). Negative symbolic consumption and consumers’ drive for self-esteem. European Journal of Marketing, 38, 850–869. Barbaro, M. (2004, July 8). In-store testing, a recipe that sells. The Washington Post, p. E01. Blakeslee, S. (2004, July 13). When the brain says, ‘Don’t get too close.’ The New York Times, p. F2. Byrne, B. M. (2004). Testing for multigroup invariance using AMOS Graphics: A road less traveled. Structural Equation Modeling, 11, 272–300. Cash, T. F. (1990). The psychology of physical appearance: Aesthetics, attributes, and images. In T. F. Cash & T. Pruzinsky (Eds.), Body images: Development, deviance, and change (pp. 51–79). New York: Guilford. Chapman, L. J., Chapman, J. P., & Raulin, M. L. (1978). Body-image aberration in schizophrenia. Journal of Abnormal Psychology, 87, 399–407. Citrin, A. V., Stem, D. E., Spangenberg, E. R., & Clark, M. J. (2003). Consumer need for tactile input: An internet retailing challenge. Journal of Business Research, 56, 915–922. ClickZ Stats Staff. (2003). June 2003 Internet usage stats. Retrieved July 21, 2005, from http://www.clickz.com/stats/big_picture/traffic_patterns/article.php/2237901 Cox, D. F., & Rich, S. J. (1964). Perceived risk and consumer decision-making: The case of telephone shopping. Journal of Marketing Research, 1, 32–39. Epstein, S. (1980). The self-concept: A review and a proposal of an integrated theory of personality. In E. Staub (Ed.), Personality: Basic issues and current research (pp. 121–134). Englewood Cliffs, NJ: Prentice Hall. Feingold, A., & Mazzella, R. (1998). Gender differences in body image are increasing. Psychological Science, 9, 190–195. Ferraro, R., Shiv, B., & Bettman, J. R. (2005). Let us eat and drink, for tomorrow we shall die: Effect of mortality salience and self-esteem on self-regulation in consumer choice. Journal of Consumer Research, 32, 65–75. Fisher, S. (1970). Body experience in fantasy and behavior. New York: Meredith Corp. Fisher, S. (1973). Body consciousness: You are what you feel. Englewood Cliffs, NJ: Prentice Hall. Fisher, S. (1986). Development and structure of the body image (Vols. 1 & 2). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Fisher, S., & Cleveland, S. (1958). Body image and personality. Oxford, England: Van Nostrand. Franzoi, S. L., & Herzog, M. E. (1986). The body esteem scale: A convergent and discriminant validity study. Journal of Personality Assessment, 50, 24–31. Franzoi, S. L., & Shields, S. A. (1984). The body esteem scale: Multidimensional structure and sex differences in a college population. Journal of Personality Assessment, 48, 173–178. Freeman, A. M., & Melges, F. T. (1977). Depersonalization and temporal disintegration in acute mental illness. American Journal of Psychiatry, 134, 679–681. Graziano, M. S. A., & Gross, C. G. (1995). From eye to hand. In J. King (Ed.), Scale in conscious experience: Is the brain too important to be left to specialists to study? (pp. 119–129). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Graziano, M. S. A., & Gross, C. G. (1998). Spatial maps for the control of movement. Current Opinion in Neurobiology, 8, 195–201. Greenspan, R. (2004). Online Asians lead in income, E-commerce. Retrieved March 15, 2005, from www.clickz.com/stats/big_picture/demographics/article.php/3326041 Head, H. (1926). Aphasia and kindred disorders of speech. London: Cambridge University Press. Heatherton, T. F., & Polivy, J. (1991). Development and validation of a scale for measuring state self-esteem. Journal of Personality and Social Psychology, 60, 895–910. Henriques, G. R., & Calhoun, L. G. (1999). Gender and ethnic differences in the relationship between body esteem and self-esteem. The Journal of Psychology, 133, 357–368. Hermann, A. D., Leonardelli, G. J., & Arkin, R. M. (2002). Self-doubt and self-esteem: A threat from within. Personality & Social Psychology Bulletin, 28, 395–408. Hu, L., & Bentler, P. (1999). Cutoff criteria for fit indices in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. Jones, T. O., & Sasser, W. E., Jr. (1995). Why satisfied customers defect. Harvard Business Review, 73, 88–99. Joy, A., & Sherry, J. F., Jr. (2003). Speaking of art as embodied imagination: A multisensory approach to understanding aesthetic experience. Journal of Consumer Research, 30, 259–282. Koff, E., Rierdan, J., & Stubbs, M. L. (1990). Gender, body image, and self-concept in early adolescence. Journal of Early Adolescence, 10, 56–68. Landon, E. L., Jr. (1974). Self concept, ideal self concept, and consumer purchase intentions. Journal of Consumer Research, 1, 44–51. Lerner, R. M., Karabenick, S. A., & Stuart, J. L. (1973). Relations among physical attractiveness, body attitudes, and self-concept in male and female college students. Journal of Psychology, 85, 119–129. Li, H., Daugherty, T., & Biocca, F. (2003). The role of virtual experience in consumer learning. Journal of Consumer Psychology, 13, 395–407. Mahoney, E. R., & Finch, M. D. (1976). The dimensionality of body-cathexis. Journal of Psychology, 92, 277–279. Maravita, A., & Iriki, A. (2004). Tools for the body (schema). Trends in Cognitive Sciences, 8, 79–86. Marsh, H. W. (1990). A multidimensional hierarchical model of self-concept: Theoretical and empirical justification. Educational Psychology Review, 2, 562–589. McCabe, D. B., & Nowlis, S. M. (2003). The effect of examining actual products or product descriptions on consumer preference. Journal of Consumer Psychology, 13, 431–439. Mittal, B. (1995). A comparative analysis of four scales of consumer involvement. Psychology & Marketing, 12, 663–681. Muth, J. L., & Cash, T. F. (1997). Body-image attitudes: What difference does gender make? Journal of Applied Social Psychology, 27, 1438–1452. Oliver, R. L., & Bearden, W. O. (1983). The role of involvement in satisfaction processes. Advances in Consumer Research, 10, 250–255. Pastore, M. (2001). Online consumers now the average consumer. Retrieved July 12, 2004, from http://www.clickz.com/stats/big_picture/demographics/article.php/800201 Rosa, J. A., & Malter, A. J. (2003). E-(embodied) knowledge and e-commerce: How physiological factors affect on-line sales of experiential products. Journal of Consumer Psychology, 13, 63–75. Rush, L. (2004a). E-commerce growth spurred by maturation. Retrieved January 23, 2005, from http://ww.clickz.com/stats/markets/retailing/article.php/3303311 Rush, L. (2004b). Women, comparison shopping help boost E-commerce holiday revenues. Retrieved January 15, 2005, from http:// www.clickz.com/stats/markets/retailing/article.php/6061_3299531 Schilder, P. (1950). The image and appearance of the human body. New York: International Universities Press. (Original work published 1935) Schlosser, A. E. (2003). Experiencing products in the virtual world: The role of goal and imagery in influencing attitudes versus purchase intentions. Journal of Consumer Research, 30, 184–198. Schouten, J. W. (1991). Selves in transition: Symbolic consumption in personal rites of passage and identity reconstruction. Journal of Consumer Research, 17, 412–425. Secord, P. F., & Jourard, S. M. (1953). The appraisal of body cathexis: Body cathexis and the self. Journal of Consulting Psychology, 17, 343–347. KEEPING THE BODY IN MIND Shaffer, T., Root, T., & Sherrell, D. L. (1997). Consumer satisfaction with health-care services: The influence of involvement. Psychology & Marketing, 14, 261–285. Sirgy, M. J. (1982). Self-concept in consumer behavior: A critical review. Journal of Consumer Research, 9, 287–300. Sondhaus, E. L., Kurtz, R. M., & Strube, M. J. (2001). Body attitude, gender, self-concept: A 30-year perspective. The Journal of Psychology, 135, 413–429. Spence, H. E., Engel, J. F., & Blackwell, R. D. (1970). Perceived risk in mail-order and retail store buying. Journal of Marketing Research, 7, 364–369. 91 Thompson, C. J., & Hirschman, E. C. (1995). Understanding the socialized body: A poststructuralist analysis of consumers’ self-conceptions, body images, and self-care practices. Journal of Consumer Research, 22, 139–153. Witt, J. K., Proffitt, D. R., & Epstein, W. (2004). Perceiving distance: A role of effort and intent. Perception, 33, 577–590. Received: April 28, 2005 Accepted: July 20, 2005 TABLE A1 Composition of Dimensions—Body Esteem Men Physical condition Stamina Reflexes Energy Health Physical condition Coordination Agility (Biceps) (Arms) (Muscle strength) Women General attractiveness Nose Lips Ears Chin Eyes Cheeks and cheekbones Face Shoulder Feet Body appearance Waist Body build Physique and figure Stomach Weight Physical condition Stamina Reflexes Energy Health Physical condition Coordination Agility General attractiveness Nose Lips Ears Chin Eyes Cheeks and cheekbones Face Shoulders Feet (Legs) (Buttocks) (Chest) (Hips) (Thighs) Body appearance Waist Body build Physique and figure Stomach Weight (Biceps) (Hips) (Legs) (Thighs) (Arms) (Buttocks) TABLE A2 Composition of Dimensions—Involvement With Apparel Apparel Is Important Concerned With Apparel To me, apparel is important–unimportant. To me, apparel matters–does not matter. To me, apparel is significant–insignificant. To me, apparel is of no concern–concern. To me, apparel means nothing–a lot. TABLE A3 Composition of Dimensions—Body Boundary Aberration Extremities Body Ownership and Existence I have felt as if I am united with objects near me. I think it is possible that one of my arms or legs is disconnected from the rest of me. I can remember when it seemed as though one of my limbs took on an unusual shape. I have has the momentary feeling that things I touch remain attached to me. Now and then when I look in the mirror, my face seems different than usual. At least once, I have wondered if my body was really my own. I have felt as if I could not distinguish my body from other objects around me. I have felt as if my body does not exist. I have had the momentary feeling that my body has changed shape. At least once, I have sometimes felt confused as to whether my body was really my own.