“Social comparison in customization: The effects of
appearance self-esteem and perceived similarity on
the relationship between customization types and
consumer product evaluations”
Master Thesis
University of Groningen
Faculty of Economics and Business
First Supervisor: dr. Debra Trampe Second supervisor: dr. Jia Liu
Abstract
This research aims at understanding when high appearance self-esteem consumers are most likely to engage in negative social comparisons and examines how comparing themselves with an attractive model affects consumer product evaluations. In a pilot study, it turned out that traditional product customization is preferred to customization which shows how the product fits on the attractive model. In an experiment, additional findings have been found. Specifically, high appearance self-esteem consumers are more susceptible to form negative outcome evaluations when the attractive model is consuming (i.e., wearing) the focal product; whereas, low appearance self-esteem consumers are more susceptible to form positive products evaluations when the attractive model is consuming the focal product. Additional analysis show, however, that for high appearance self-esteem consumers the advantage of product customization without the attractive model on product customization with the attractive model disappears when consumer perceive to be highly similar to the model. This research presents counterintuitive evidence of how social comparison mechanism may affect consumer product evaluations in a specific context, that is customization.
Introduction
Patricia, a first-year Research Master student, is customizing her t-shirt, trousers and shoes using H&M’s virtual dressing room (www.hm.com/us/dressingroom/LADIES). She has the opportunity to select different types of clothes which vary in terms of colors, patterns, shapes and sizes. During the customization process, Patricia visualizes how the clothes fit on a very attractive model. Patricia is also very beautiful but she is not satisfied with what she is visualizing on her laptop and decides to interrupt her customization activity.
In the current research we question why Patricia decides to interrupt her customization activity and whether her evaluations of the clothes are affected by the presence of an attractive model during the customization process. Specifically, we are interested in investigating the effects of social comparison on product evaluations, that is, how consumers react when attractive models are consuming (i.e., wearing) the products they are interested in purchasing. Specifically, the current research investigates, adopting the perspective of social comparisons (Festinger 1954), the impact of social information provided by an attractive model in the customization setting on customer product evaluations.
customization approaches (e.g., customerization and individualization) have significantly increased in recent years.
Several researchers have investigated consumer evaluations of customized offerings (Bendapudi and Leone 2003; Dellaert and Stremersch 2005; Wind and Rangaswamy 2001), and psychological consequences of involving customers in the product design process (Franke, Keinz and Steger 2009; Franke, Schreier, and Kaiser 2010). These studies demonstrate how individuals prefer customized products to standard ones because of a greater preference fit – which is the degree to which the customized product is similar to the ideal product - and feelings of accomplishment at designing the product (i.e., “I designed it myself” effect - Franke, Schreier, and Kaiser 2010). Moreover, these studies investigated the role of effort, enjoyment and perceived uniqueness as the moderators of the relationship between customization and customer attitude and dispositions (Franke and Schreier 2008; Franke and Schreier 2010).
In the current research we select the customization setting in order to test our predictions. We contribute to the literature by understanding when those high in appearance self-esteem will be most susceptible to negative social comparison than those low in appearance self-esteem.
Consider the following pilot study. Participants (N=61, 50.8% females,
Mage = 23.21, SDage = 2.16) were randomly assigned to the product customization scenario without an attractive model or the product customization scenario with an attractive model. Respondents then answered an item which measures attitude toward customization (“I do
like the opportunity to customize firms’ offerings”, 1=completely disagree,
practices?”, 1=very low probability, 7=very high probability). On average
participants in the product customization condition without an attractive model showed an higher attitude toward the object of customization than participants in the product customization condition with an attractive model (Mwithout= 6.14, SDwithout=.68; Mwith=4.28, SDwith=1.74). This effect was significant (t(59)=3.52, p<.001). The same results have been found for intention to engage in product customization. On average, participants in the product customization condition without an attractive model showed a higher intention to engage in product customization than participants in the product customization condition with an attractive model (Mwithout= 4.89, SDwithout=.95; Mwith=3.56, SDwith=1.56). This effect was significant (t(59)=3.81, p<.001).
conducted demonstrates that the use of attractive models during customization lowers customer attitude and intention to engage in customization practices. Therefore, our preliminary findings show that the combined use of customization practices and attractive models that consume the focal product may be detrimental in terms of customer reactions.
The current research aims at understanding which factors may lead to this counterintuitive findings. In particular, we examine from the standpoint of social comparisons, the impact of social information provided by an attractive model on consumer product evaluations. Our contribution to the literature is focused on understanding when different levels of appearance self-esteem trigger negative (or positive) social comparison. Differently from Dahl, Argo and Morales (2012) who demonstrated that low self-esteem consumers have the most negative reactions to social comparison, we focus on consumers high in appearance self-esteem and demonstrate that they have the most negative reactions to social comparisons.
Literature background
Social comparison theory and body attractiveness
Social comparison theory (Festinger 1954) asserts that people often search out standards to which they can compare themselves. This comparison can be distinguished in upward and downward social comparison. Upward social comparison occurs when individuals compare themselves to someone whom they believe to be better off than themselves. Conversely, downward social comparison occurs when individuals compare themselves to someone whom they believe to be worse off than themselves.
Festinger (1954) proposed that upward comparison was likely to produce negative consequences, such as decreased self-esteem, whereas downward comparison was likely to produce positive consequences, such as increased self-esteem. Likewise, Argo, White and Dahl (2006) assert that upward comparisons may threaten customer self-esteem, causing them to misrepresent themselves to others. Also Richins (1991) states that upward comparisons can result in decreased self-satisfaction, and downward comparisons can result in increased self-satisfaction. Moreover, further research in laboratory, school, work and health settings has found that upward comparisons can produce jealousy (Salovey and Rodin 1984), hostility (Testa and Major 1990), frustration (Martin, 1986), and lowered self-evaluations (Marsh and Parker 1984; Morse and Gergen 1970), whereas downward comparisons typically enhance subjective well-being (see Gibbons et al. 1991 for a review).
downward comparisons (to those doing worse than the self) produce positive affect, an emerging body of literature suggests that the affective consequences of a given social comparison may not be intrinsic to its direction (Buunk, Collins, Taylor, VanYperen and Dakof 1990). For example, upward comparison may be inspiring when information about successful coping or achievement is provided (Taylor and Lobel 1989) and when success in the comparison domain is perceived to be controllable (Buunk et al. 1990). Likewise, downward comparison have the potential to frighten the comparer when examples of failure or deterioration are provided (Brickman and Bulman 1977; Tayolor and Lobel 1989).
Although the direction of the link between global self-esteem and appearance self-esteem has not been established, several researchers posited that satisfaction with one’s physical appearance is the personal variable most strongly linked to one’s global self-esteem (Harter 1999; Brown 1993; Mendelson 1982). Research in advertising has widely demonstrated that self-esteem and body satisfaction decrease when females are exposed to thin media images (e.g., Grabe, Ward and Hyde 2008). Likewise, Richins (1991) demonstrated that women were less satisfied with their own physical appearance after they viewed advertisements featuring attractive and thin models. Moreover, Myers and Biocca (1992) argue that exposure to thin media images can negatively affect body-image perception. However, Smeesters and Mandel (2006) found that women compare themselves spontaneously and automatically with the models in advertisements. They demonstrate that exposure to thin models does not necessarily have a negative impact on one’s self-esteem. On the contrary, exposure to moderately thin (but not extremely thin) models has a positive impact on one’s self-esteem. Other researchers also demonstrate that exposure to thin models in magazines can lead to self-enhancement (Henderson-King and Henderson-King 1997) and thinner self-ratings (Mills, Polivy, Herman and Tiggemann 2002).
levels) and the model’s size that determines the ad’s effect on self-esteem. In their studies, the researchers have analyzed how being exposed to a thin (attractive) model may affect customer self-evaluation (i.e., self-esteem) and demonstrate how being exposed to moderately thin models may have positive consequences for customers. In the current research, customer self-evaluation (i.e., appearance self-esteem) after being exposed to an attractive model is used as a moderator of the relationship between the customization type and customer attitude and dispositions toward the customized outcome. Past researchers have demonstrated that people with low self-esteem may be more sensitive to social comparison information because they are less certain of their own attributes (Campbell 1990; Swallow and Kuiper 1988). However, other investigators have found that people high in self-esteem may be more likely to be self-serving in their use of social comparisons (Crocker and Schwartz 1985; Crocker Thompson, McGraw and Ingerman 1987) and other self-relevant information (Campbell 1986). Our research builds on the latter findings by identifying the specific conditions within customization practices which lead consumers with high appearance self-esteem of being negatively affected by appearance-related social comparisons.
Product customization
Customization consists of customer-tailored definition of the firm’s offering, aimed to approximate as perfectly as possible the ideal product for each customer (Wind and Rangaswamy 2001). Technological developments and marketing practices have progressively allowed for sophisticated forms of customization. Moreover, the diffusion of flexible manufacturing systems integrated into total quality programs (Hayes and Pisano 1994) has enabled firms to modularize their production processes. Specifically, products can be divided into modules (i.e. product modularity), which can be produced and offered to customers in several variants (e.g., colours, dimensions, patterns), and still achieve significant scale economies (Womack, Jones, and Roos 1990). Modules combination, in make-to-order processes, results in mass customization, which represents one of the first customization model proposed by marketing researchers (Pine, 1993). Mass customization can be distinguished in attribute-based and
alternative-based customization (Huffman and Kahn 1998). Through the
Personalization may also rest on interactivity between customers and firms and the use of previously collected customer data. One-to-one
personalization (Peppers, Rogers, and Dorf 1999) is an approach aiming
at enhancing the interaction between customers and the firm, as well as presenting personalized value in terms of services, information, and support. Wind and Rangaswamy (2001) consider one-to-one personalization as mainly focused on interaction flexibility and close communication in the customer-firm relationship, but not featuring large levels of product modules combination. Amazon.com represents an instance of one-to-one personalization based on the proposition of differentiated forms of service and interaction with and among customers. Indeed, Amazon.com collects information about customers during their surfing experience and elaborates on them to offer customized services during the next access to the website. Furthermore, Amazon.com significantly supports and triggers interaction among customers by means of communities, chats and blogs.
to the 3D technology and the excellent graphic features of its website, allows customers to experience positive emotions during the design process.
Companies are increasingly investing in interactive customization toolkits in order to involve customers into the product design process and to reduce customer uncertainty (Lurie and Mason 2007). Customization toolkits have been widely defined in the customization research stream. Franke, Schreier and Kaiser (2010), for instance, define a customization toolkit as the interface between manufacturers and customers and they also propose different ways of naming customization toolkits, that is, configurator, choice menu, design kit, or toolkit for user innovation and design. Franke and Schreier (2010) analyze which factors prompt customers to attribute value to products they design themselves using mass-customization toolkits. These toolkits allow trial-and-error experimentation and deliver immediate feedbacks on the potential outcome of design ideas (Von Hippel 2001; Von Hippel and Katz 2002). Once a satisfactory solution is found, the design can be transferred into a firm’s production system delivered to the customer (Dallaert and Stremersch 2005; Randall, Terwiesch and Ulrich 2007). Randall et al. (2007) distinguish parameter-based toolkits (i.e., where users directly specify values for design parameters of the product, like the size of a personal computer’s hard drive) from need-based toolkits (i.e., where users specify their needs, such as the wish to store a large quantity of data on the PC). They find that, whereas the former type of toolkit seems to suit expert users, the latter type of toolkit offers a better fit for novice users.
Huffman and Kahn (1998) suggest that some consumers may find learning their preferences about a product to be fun, and Dallaert and Stremersch (2005) presume that consumers might enjoy mass customizing a product. Furthermore, some authors (e.g., Franke, Schreier and Kaiser 2010) demonstrate that customers show higher attitude and dispositions toward customized products than standard ones because of the achievement of greater preference fit and the feeling of pride for being the designer of the product (i.e., “I designed it myself” effect).
As already mentioned, past research shows that customers prefer customized products to standard ones because of a greater preference fit and the feeling of pride for being the designer of the customized product (Franke, Schreier and Kaiser 2010). These results are based on studies conducted through customization toolkits where participants have the opportunity to really engage in customization practices. Thus, participants through trial-and-error processes can choose among product modules in order to define the design which best fits their needs and expectations. Companies are increasingly presenting customization by including attractive models in customization toolkits. Although in advertising research it has been widely demonstrated that being exposed to an attractive model may have different effects in terms of customer self-evaluation, little research has been conducted on: 1) the effects of being exposed to an attractive model in customization, and 2) the effects of being exposed to an attractive model on behavioral outcomes.
creations. Their studies demonstrate that preference fit alone does not fully explain consumers’ dominant preferences for self-designed products. More subtle factors, such as social comparisons to other designers, also influence consumers’ evaluations. To our knowledge, their research represents the first attempt to investigate how threats generated by upward comparisons with professional designers can influence evaluations of self-designed products.
attractive model during the customization process may be detrimental in terms of customer reactions toward the customized outcome.
Specifically, we aim at demonstrating that the type of customization can influence customer attitude and dispositions and that this is moderated by appearance self-esteem (ASE) – that is, the self-worth a person derives from his or her body-image and weight. We propose that individuals high in appearance self-esteem evaluate customization which includes attractive models less positively than individuals low in appearance self-esteem. Individuals high in appearance self-esteem, who are faced with customization which involves attractive models, are more self-serving in their use of social comparison. This is in line with
compensatory or defensive self-enhancement theory (Allport 1937;
b) willingness to pay for, c) intention to buy, d) intention to customize the product.
Specifically, we hypothesize:
H1: The presence of the attractive model generates lower attitude towards, willingness to pay for, intention to buy and intention to customize the product in the future for customers with high appearance self-esteem than for customers with low appearance self-esteem. Conversely, the absence of the attractive model generates higher attitude towards, willingness to pay for, intention to buy and intention to customize the product in the future for customers with high appearance self-esteem than for customers with low appearance self-esteem.
someone who strictly resembles themselves. Taken together, customers high in appearance self-esteem and their positive feeling of fitting to the model during the customization process should lead to positive evaluations for the customized outcome.On the other hand, we expect that customers who evaluate themselves less positively (i.e., low appearance self-esteem), despite the feeling of being similar to the attractive model (i.e., high perceived similarity), will evaluate the customized outcome less positively. In this case, indeed, fitting to the model leads to negative feelings due to customers’ negative self-evaluations (low appearance self-esteem). Therefore, the customization process should lead to negative evaluations for the customized outcome. Thus, we maintain that the level of appearance self-esteem interacts with the level of perceived similarity to the attractive model in determining customer a) attitude towards, b) willingness to pay for, c) intention to buy, c) intention to customize the product.
Formally, we hypothesize that:
H2: High appearance self-esteem generates higher attitude towards, willingness to pay for, intention to buy and intention to customize the product in the future for customers with high perceived similarity than for customers with low perceived similarity. On the other hand, low appearance self-esteem generates lower attitude towards, willingness to pay for, intention to buy and intention to customize the product in the future for customers with high perceived similarity than for customers with low perceived similarity.
with the attractive model) and customer’s appearance self-esteem. Our second research hypothesis identifies - through the moderating role of perceived similarity - one condition in which respondents with high appearance self-esteem also positively evaluate customization showing the attractive model.
Methodology
Research design
To test our hypotheses, an experiment has been conducted where the independent variables have been: 1) the type of customization (product customization without an attractive model versus product customization with an attractive model), and 2) appearance self-esteem that has been measured using three items (“I feel satisfied with
the way my body looks right now”, “I am dissatisfied with my weight” –
reversed item, and “I feel unattractive” – reversed item, 1 = strongly disagree, 7 = strongly agree - Cronbach’s alpha = .68) from the Appearance self-esteem scale of Heatherton and Polivy (1991). The dependent variables that have been measured are attitude toward, willingness to buy the customized outcome, purchase intention and willingness to engage in customization practices in the future.
Procedure
Having chosen among the different customization options (e.g., color selection, text and pictures uploading), participants answered four 7-point items measuring attitude toward the t-shirt (“I like the design of my customized t-shirt”, “The design of my customized t-shirt is good”, “The
Product customization without the attractive model Product customization with the attractive model
Next participants answered three items measuring self-awareness (Fenigstein, Scheier and Buss 1975 - “I am alert to changes in my mood”, “I reflect about myself a lot”, “I am always trying to figure myself out”, 1=strongly disagree, 7=strongly agree – Cronbach’s alpha = .71). Participants also answered three items measuring self-esteem (Rosenberg 1989 - “On the whole, I am satisfied with myself”, “I feel that I have a good number of qualities”, “I take a positive attitude toward myself”, 1=strongly disagree, 7=strongly agree – Cronbach’s alpha = .85) and three items of the Heatherton and Polivy’s (1991) Appearance self-esteem scale (“I feel satisfied with the way my body looks right now”, “I
am dissatisfied with my weight” – reversed item, and “I feel unattractive” – reversed item, 1 = strongly disagree, 7 = strongly agree –
Cronbach’s alpha = .68). We have also included three items (“I like what I look like in pictures”, “My weight makes me happy”, “My looks help me get dates”, 1=strongly disagree, 7=strongly agree – Cronbach’s alpha = .72 ) to measure the Body self-esteem scale for adults and adolescents (BESAA - Mendelson, Mendelson and White, 2001). Finally, two items have been used to measure perceived similarity to the attractive model (“This person looks like me”, “This person shows my own physical features”, 1=strongly disagree, 7=strongly agree – Cronbach’s alpha = .84).
Results
Model assumptions and data inspection
Normality assumption
To check whether the distribution of our dependent variables is normal, we inspected P-P plots, in which the actual z-score is plotted against the expected score. If the data are normally distributed then the actual z-score will be the same as the expected z-z-score and we will get a straight diagonal line. Figure 2 shows that our dependent variables are normally distributed.
Normality assumption check has also been conducted for the two groups separately. Figure 3 shows P-P plots for the group of respondents who visualized the attractive model during the customization process. The graphs show that the dependent variables follow a normal distribution in the group of customization with the attractive model.
Figure 3 P-P plots on dependent variables:Attitude, WtP, ItB and ITC (sample of respondents who visualized the attractive model)
Figure 4 P-P plots on dependent variables:Attitude, WtP, ItB and ITC (sample of respondents without visualizing the attractive model)
Homogeneity of variance assumption
Table 1 Test of Homogeneity of variance
Levene
Statistic df1 df2 Sig.
Attitude Based on Mean ,016 1 81 ,899
Based on Median ,007 1 81 ,931
Based on Median and with adjusted df ,007 1 80,841 ,931
Based on trimmed mean ,018 1 81 ,895
Willingness to Pay Based on Mean 1,275 1 81 ,262
Based on Median 1,126 1 81 ,292
Based on Median and with adjusted df 1,126 1 80,979 ,292
Based on trimmed mean 1,033 1 81 ,313
Intention to buy Based on Mean 2,043 1 81 ,157
Based on Median 1,001 1 81 ,320
Based on Median and with adjusted df 1,001 1 76,848 ,320
Based on trimmed mean 1,930 1 81 ,169
Intention to customize Based on Mean ,448 1 81 ,505
Based on Median ,502 1 81 ,481
Based on Median and with adjusted df ,502 1 80,023 ,481
Based on trimmed mean ,519 1 81 ,474
For the attitude, the variances are equal for the group with and without the attractive model (F(1,81) = .016, ns). Also for willingness to pay (F(1,81) = 1.275, ns), intention to buy (F(1,81) = 2.043, ns) and intention to customize (F(1,81) = .448, ns) the two groups show homogeneous variance.
Outliers
Dependent variables
Figure 5 Outliers inspection of dependent variables
Independent variables
Figure 6 Outliers inspection of independent variables concerning the customization task
Figure 7 Outliers inspection of independent variables concerning respondents’ personality
Descriptive analysis
Dependent variables
Table 2 Descriptive statistics of dependent variables
N Minimum Maximum Mean Std. Deviation
Attitude 83 1.50 7.00 4.59 1.17 Willingness to pay 83 0 30 11.17 6.57 Intention to buy 83 1 7 3.66 1.73 Intention to customize 83 1 7 3.87 1.64 Independent variables
Table 3 shows descriptive statistics of independent variables concerning the customization task. In particular, respondents on average feel the task is not particularly effortful (Meffort = 2.91, SDeffort = 1.02). On the other hand, respondents on average evaluate the task enjoyable (Menjoyment = 4.48, SDenjoyment = 1.33). Respondents also perceive the product neither close nor far from their ideal t-shirt (Mperceived_fit = 3.34, SDperceived_fit = 1.51). Moreover, they show an average value in feeling proud of their customized t-shirt (Maccomplishment = 3.75, SDaccomplishment = 1.65) and in having the design skills to create a good t-shirt (MDS = 3.49, SDDS = 1.42).
There were not significant differences between conditions in terms of perceived effort (F(1,81)=.042, ns), perceived enjoyment (F(1,81)=1.35, ns), perceived fit (F(1,81)=.49, ns) and accomplishment (F(1,81)=1.03, ns).
Table 3 Descriptive statistics of independent variables concerning the customization task
N Minimum Maximum Mean Std. Deviation
Effort 83 1,00 5,33 2,91 1,02
Enjoyment 83 1,33 7,00 4,48 1,33
Perceived fit 83 1 7 3,34 1,51
Accomplishment 83 1 7 3,75 1,65
Table 4 shows descriptive statistics of the independent variables concerning respondents’ personality. On average, they are highly self-aware (MSA = 5.08, SDSA = .97) and show high self-esteem (MSE = 5.49, SDSE = .89). On average, respondents also show high appearance self-esteem (MASE = 4.98, SDASE = 1.14) and BESAA (MBESAA = 5.08, SDBESAA = .95). On the other hand, on average, they do not feel similar or dissimilar to the attractive model shown during the customization task (MPS = 3.58, SDPS = 1.42).
There were not significant differences between conditions in terms of self-awareness (F(1,81)=.16, ns), self-esteem (F(1,81)=.01, ns), appearance self-esteem (F(1,81)=.10, ns), BESAA (F(1,81)=.00, ns) and perceived similarity (F(1,81)=.48, ns).
Table 4 Descriptive statistics of independent variables concerning respondents’ personality
N Minimum Maximum Mean Std. Deviation
Self-awareness 83 2,33 7,00 5,0843 ,96671 Self-esteem 83 1,33 7,00 5,4900 ,89160 Appearance self-esteem 83 1,67 7,00 4,9839 1,14042 BESAA 83 1,33 6,67 4,8434 ,95045 Perceived similarity 83 1,00 7,00 3,5843 1,42884 Bivariate analysis
To gather preliminary knowledge about relations among variables in our data we run correlational analyses.
Dependent variables
correlation between attitude and intention to customize is statistically not significant (r = -.193, ns). Willingness to pay is positively correlated to intention to buy (r = .412, p < .05), while the correlation between willingness to pay and intention to customize is statistically not significant (r = .002, ns). Also the correlation between intention to buy and intention to customize is statistically not significant (r = .069, ns).
Table 5 Correlations among dependent variables
Independent variables
Table 6 shows correlations among independent variables regarding the customization task. Perceived effort shows a significant correlation only with perceived fit (r = .329, p < .05); whereas correlations are not significant when considering enjoyment (r = .024, ns), accomplishment (r = .117, ns) and design skills (r = .172, ns). Perceived enjoyment is positively and significantly correlated to perceived fit (r = .426, p < .05), accomplishment (r = .378, p < .05) and perception of having good design skills (r = .354, p < .05). Perceived fit is also positively correlated to accomplishment (r = .348, p < .05) and perception of having good design skills (r = .302, p < .05). Accomplishment is positively correlated to the perception of having good design skills (r = .428, p < .05).
Variable Attitude Willingness to pay Intention to buy Intention to customize
Attitude 1
Willingness to pay ,736(**) 1
Intention to buy ,455(**) ,412(**) 1
Table 6 Correlations among independent variables regarding the customization task
Effort Enjoyment Perceived fit Accomplishment Design skills
Effort 1
Enjoyment ,024 1
Perceived fit ,329(**) ,426(**) 1
Accomplishment ,117 ,378(**) ,348(**) 1
Design Skills ,172 ,354(**) ,302(**) ,428(**) 1
Table 7 shows the findings of correlations among independent variables regarding respondents’ personality. We expect that BESAA, appearance self-esteem and self-esteem are positively correlated. On the other hand, we expect that BMI is negatively correlated to BESAA and appearance self-esteem. As expected, BESAA is strongly and positively correlated to appearance self-esteem (r = .736, p < .05) and self-esteem (r = .455, p < .05). On the other hand, BMI is strongly and negatively correlated to BESAA (r = -.396, p < .05) and appearance self-esteem (r = -.494, p < .05), which means that the higher the BMI the lower respondents are satisfied with their body and physical appearance. Finally, correlation between self-awareness and the other variables are not statistically significant (p = ns).
Table 7 Correlations among independent variables regarding respondents’ personality
Gender differences: t-tests
In table 8, all gender differences are reported with regard to the dependent and independent variables measured in our experiment. There are not gender differences regarding the dependent variables and the independent variables.
Table 8 Gender differences
Variable Meanfemale (SDfemale) Meanmale (SDmale) t-value DF p-value
Attitude 4.611 (1.109) 4.548 (1.280) .233 81 .816 Willingness to pay 11.56 (6.687) 10.52 (6.413) .697 81 .488 Intention to buy 3.90 (1.729) 3.26 (1.692) 1.659 81 .101 Intention to customize 3.79 (1.588) 4.00 (1.732) -.567 81 .572 Effort 2.891 (.897) 2.946 (1.214) -.237 81 .813 Enjoyment 4.567 (1.310) 4.333 (1.372) .773 81 .442 Perceived fit 3.33 (1.396) 3.35 (1.704) -.081 81 .936 Accomplishment 3.83 (1.779) 3.61 (1.430) .569 81 .571 Design Skills 3.481 (1.329) 3.508 (1.577) -.084 81 .933 Self-awareness 4.968 (.955) 5.279 (.970) -1.480 81 .157 Self-esteem 5.462 (.886) 5.538 (.914) -.374 81 .709 Appearance self-esteem 4.962 (1.120) 5.022 (1.192) -.230 81 .818 BESAA 4.770 (.872) 4.968 (1.073) 1.132 81 .361 Perceived similarity 3.721 (1.473) 3.355 (1.343) .020 81 .261
Moderation analysis: Appearance self-esteem as moderating variable
customization (product customization without an attractive model = 1, product customization with an attractive model = 0), appearance self-esteem, their interaction and additional predictors which have been investigated by past researchers (i.e., perceived effort, perceived enjoyment, perceived uniqueness, preference fit and feeling of accomplishment to be the designer of the product). It is expected that with high levels of appearance self-esteem, the customization model without the attractive model is preferred to the customization showing the attractive model. In addition, to test for the continuous interaction between customization type and appearance self-esteem, a spotlight analysis (Hayes and Matthes 2009) has been conducted for each regression model to verify how the effect of the type of customization changes with low (i.e., Mean – 1SD), average (i.e., Mean) and high (i.e., Mean + 1SD) levels of appearance self-esteem. Table 9 presents a synthesis of the findings.
Table 9 Moderation analysis: Appearance self-esteem as moderating variable
Dependent Variable Customization type effect (s.e.) Appearance self-esteem effect (s.e.) Interaction estimate (s.e.)
Customization type effect for different levels of appearance self-esteem (s.e.)
Results for attitude
In the attitude model, the direct effect of the customization type is negative and significant (b=-1.50, p < .10), which means that customizing with the attractive model has a less negative effect on attitude than customizing without the attractive model. The direct effect of appearance self-esteem is not significant (b=-.11, ns). Next, the expected interaction effect is significant (b=.31, p < .05). When conducting a spotlight analysis, however, no significant effects of the type of customization have been found for low (b=-.33, ns), medium (b=.02, ns) and high (b=.37, ns) levels of appearance self-esteem. This means that, independently from the level of appearance self-esteem, product customization not showing an attractive model and product customization showing an attractive model are evaluated similarly in terms of customer attitude. However, the expected pattern for the type of customization has been found: with low levels of appearance self-esteem, the type of customization has a negative impact on customer attitude. In simple words, with low levels of appearance self-esteem, customization showing an attractive model has better evaluations than product customization without an attractive model; whereas, with high levels of appearance self-esteem, customization not showing the attractive model has better evaluations than product customization showing the attractive model.
Figure 8 Main and interaction effects for the attitude model
Results for Willingness to pay
In the willingness to pay model, the direct effect of the type of customization is not significant (b=-6.07, ns). The direct effect of appearance self-esteem is also not significant (b=1.32, ns).
Next, the expected interaction is significant (b=1.93, p <.10). When conducting a spotlight analysis, a non-significant effect of the type of customization has been found for low levels of appearance self-esteem (b=1.35, ns); whereas, with medium (b=3.56, p < .01) and high (b=5.76, p < .01) levels of appearance self-esteem significant effects of the type of customization have been found. This evidence means that, with medium and high levels of appearance self-esteem, respondents are more willing to pay for customization without an attractive model than for customization with an attractive model.
Figure 9 Main and interaction effects for the WtP model
Results for intention to buy
In the intention to buy model, the direct effect of the type of customization is not significant (b=-1.70, ns). The direct effect of appearance self-esteem is also not significant (b=-.23, ns).
The expected interaction effect is significant (b=.43, p < .01). When conducting a spotlight analysis, a non-significant effect of the type of customization has been found with low (b=-.05, ns) and medium (b=.44, ns) levels of appearance self-esteem. Instead, a significant effect of the type of customization has been found with high level of appearance self-esteem (b=.93, p < .05). Thus, respondents with high appearance self-esteem are more willing to buy the customized output when the attractive model is not shown during the customization process.
Figure 10 Main and interaction effects for the ItB model
Results for intention to customize
In the intention to customize model, the direct effect of the type of customization is not significant (b=.18, ns). The direct effect of appearance self-esteem is also not significant (b=.12, ns).
The interaction effect is not significant (b=.01, ns). When conducting a spotlight analysis, no significant effects of the type of customization have been found with low (b=.21, ns), medium (b=.22, ns) and high (b=.23, ns) levels of appearance self-esteem. This means that, independently from the level of appearance self-esteem, product customization not showing the attractive model and product customization showing the attractive model are evaluated similarly in terms of customer intention to customize products in the future.
Figure 11 Main and interaction effects for the ItC model
In the next paragraphs, we have run the same analysis by including BESAA and BMI as the moderating variable. In the model where BESAA has been included as the moderating variable, we expect findings in line with the model where appearance self-esteem has been treated as the moderating variable. In the model where BMI has been included as the moderating variable, we expect opposite patterns because of the inverse relationship between BMI and appearance self-esteem (or BESAA).
Moderation analysis: BESAA as moderating variable
Table 10 Moderation analysis: BESAA as moderating variable Dependent Variable Customization type effect (s.e.) BESAA (s.e.) Interaction estimate (s.e.)
Customization type effect for different levels of BESAA (s.e.) Expected pattern Low (Mean – 1SD): 3.89 Mean: 4.84 High (Mean + 1SD): 5.79 Attitude -2.39*(1.30) .09(.20) .45*(.26) -.60*(.35) -.16(.24) .27(.35) Yes Willingness to Pay -16.32**(7.34) -1.93*(1.15) 3.90**(1.49) -1.14(1.98) 2.56*(1.38) 6.26***(1.98) Yes Intention to buy -3.87*(2) -.22(.31) .81**(.40) -.73(.54) .04(.37) .81(.54) Yes Intention to customize -1.87(1.90) .11(.30) .38(.39) -.39(.51) -.03(.36) .33(.51) Yes * p < 0.10, ** p < 0.05, *** p < 0.01
Results for attitude
similarly. However, the expected pattern has been found. With high levels of BESAA, respondents prefer customization without the attractive model more than with the attractive model.
Results for willingness to pay
In the willingness to pay model, the direct effect of the type of customization is significant (b=-16.32, p < .05), which means that customizing with the attractive model has a less negative effect on willingness to pay than customizing without the attractive model. Also the direct effect of BESAA is negative and significant (b=-1.93, p <.10), which means that the better people evaluate themselves, the less is their willingness to pay the customized outcome.
Next, the expected interaction effect is significant (b=3.90, p < .05). With low levels of BESAA the effect of the customization type on willingness to pay is not significant (b=-1.14, ns) which means the two customization types are evaluated similarly. However, with high levels of BESAA, the customization type has a positive effect on willingness to pay (b=6.26, p < .01), which means that respondents are more willing to pay for the customized product in the case of customization without the attractive model than in the case of customization with the attractive model.
Results for purchase intention
However, the expected pattern has been found. Indeed, with low levels of BESAA, the customization model has a negative effect on purchase intention (b=-.73, ns) which means that, with low levels of BESAA, respondents prefer customization with the attractive model more than customization without the attractive model. Conversely, with high levels of BESAA, the customization type has a positive effect on purchase intention (b=.81, ns), namely customization without the attractive model generates higher purchase intention than customization with the attractive model.
Results for intention to customize
For the intention to customize model, the direct effect of the type of customization is significant (b=-3.87, p < .10), which means that customizing with the attractive model leads to higher intention to customize than customizing without the model. The direct effect of BESAA on intention to customize is not significant (b=-.22, ns).
Moderation analysis: BMI as moderating variable
We run the same analysis as before in order to investigate the mediating role of BMI on the relationship between customization type and customer attitude and dispositions. Since BMI is inversely related to appearance self-esteem and BESAA, we expect that with low levels of BMI (i.e., high levels of ASE and BESAA), respondents prefer customization which does not show the attractive model; whereas, with high levels of BMI (i.e., low levels of ASE and BESAA), respondents prefer customization which shows the attractive model. Table 11 shows a synthesis of the findings.
Table 11 Moderation analysis: BMI as moderating variable
Dependent Variable
Customization type effect (s.e.)
BMI
(s.e.)
Interaction estimate
(s.e.)
Customization type effect for different levels of BMI (s.e.)
Expected pattern Low (Mean
– 1SD): .19
Mean: .22 High (Mean + 1SD): .25 Attitude -1.05(1.92) -9.01(6.68) 4.02(8.71) -.29(.37) -.17(.26) -.05(.37) No Willingness to Pay 13.25(10.68) 10.16(37.13) -49.01(48.40) 4.02*(2.04) 2.54*(1.43) 1.07(2.04) Yes Intention to buy -2.63(2.84) -17.43*(9.88) 12.18(12.88) -.34(.54) .03(.38) .39(.54) No Intention to customize -2.83(2.72) -9.34(9.46) 12.82(12.33) -.42(.52) -.03(.36) .36(.52) No * p < 0.10, ** p < 0.05, *** p < 0.01
Results for willingness to pay
customize model. Specifically, in the willingness to pay model, respondents with a low BMI are more willing to pay for the customized t-shirt when the attractive model is not shown (b=4.02, p < .10). The effect of the customization type on willingness to pay becomes not significant with high levels of BMI meaning that for high BMI respondents, the willingness to pay for the customized t-shirt does not differ for the customization showing or not showing the attractive model.
Discussion of the results
Moderation analysis shows that for the attitude model and the intention to customize model, no significant effects of the type of customization have been found with low, medium and high levels of appearance self-esteem. Thus, the level of appearance self-esteem is not moderating the relationship between the type of customization and customer attitude and intention to customize in the future. However, the expected pattern has been found. Participants with high appearance self-esteem have a more positive attitude and are more willing to customize when the attractive model is not used to simulate how the t-shirt fits on her/him.
self-esteem customers evaluate customization showing the attractive model less positively than customization not showing the attractive model. These findings are in line with compensatory or defensive
self-enhancement theory (Allport 1937; Swann, Pelham and Krull 1989)
which suggests that individuals with positive self-views (i.e., high appearance self-esteem) are less motivated to self-enhance than are people with negative self-views (i.e., low appearance self-esteem).
Further analyses which include BESAA and BMI as moderating variables provide support for our previous findings. In particular, the same evidence as in the appearance self-esteem analysis has been found when considering BESAA as the moderating variable between the customization type and customer attitude and dispositions. All the expected patterns have been found. Specifically, for low BESAA respondents, customization with the attractive model generates higher attitude and dispositions than customization without the attractive model. Conversely, for high BESAA respondents, customization not showing the attractive model generates higher customer attitude and dispositions than customization showing the attractive model. In the analysis where BMI has been included as the moderating variable, only one out of four patterns have been found. Specifically, low BMI respondents are more willing to pay for customized products when the attractive model is not shown during customization.
model. We assume that customers who perceive to be highly similar with the proposed model and evaluate themselves positively (high appearance self-esteem), should evaluate customization showing the attractive model more positively than customers who evaluate themselves less positively (low appearance self-esteem) despite perceiving the model being highly similar in terms of physical features.
The role of perceived similarity
In order to understand which other variables may intervene to explain our previous results, we only consider the group of respondents who visualized the attractive model and categorized respondents into two groups (i.e., high versus low appearance self-esteem) by using median split. In particular, the median is equal to 5, thus participants with a score higher than 5 were categorized as having high appearance self-esteem; whereas, participants with a score lower than 5 were categorized as having low appearance self-esteem. We have found no significant differences in perceived similarity between female (Mfemale = 3.77) and male (Mmale = 3.00; t(39.417) = 1.85, ns).
Table 12 Moderation analysis (only respondents who visualized the attractive model)
° the model is not significant
* p < 0.10, ** p < 0.05, *** p < 0.01
From our analysis, it turns out that none of the four models is statistically significant. However, some interesting patterns emerged from our spotlight analysis. In particular, for the attitude and willingness to pay models, respondents with high levels of appearance self-esteem in the low perceived similarity condition evaluate the customized outcome negatively and are less willing to pay for the product. This effect remains negative for higher levels of perceived similarity, however it becomes smaller in magnitude. Such an evidence partially supports our expectations, that is, respondents high in appearance self-esteem show higher attitude and dispositions toward the customized outcome when they perceive to be highly similar to the attractive model. However, no definitive support can be drawn from this analysis in order to support H2.
Dependent Variable Appearance self-esteem (s.e.) Perceived similarity (s.e.) Interaction estimate (s.e.)
Appearance self-esteem effect for different levels of perceived similarity (s.e.)
Discussion of the results
In our previous section, we consider all the respondents who had the opportunity to visualize the attractive model during the customization process. We tested the interactive role of appearance self-esteem and perceived similarity on customer attitude and dispositions. Results show that all the four models are not statistically significant. However, for the attitude and willingness to pay model the expected pattern have been found. Specifically, for respondents with high appearance self-esteem the effect of perceived similarity enhances customer attitude and dispositions toward the customized outcome.
Not statistically significant results might be explained by the small sample that has been used to run our analysis. Indeed, only 42 respondents visualized the attractive model during the customization process and, thus, selected for our spotlight analysis.
General Discussion
evaluations when exposed to attractive (thin) models. Likewise, Dahl, Argo and Moreau (2012) demonstrated in the retail environment, that low self-esteem consumers are more susceptible to form negative evaluations when confronted with attractive peers. Differently, our research demonstrates that high appearance self-esteem consumers may be more susceptible to form negative evaluations of the focal product when exposed to attractive models.
Contribution
The present research contributes to the literature along different ways. First of all, this research is one of the few demonstrations of how social comparison affects subsequent behavior. Most social comparison research focused primarily on the self-evaluative consequence of comparing with others, while essentially ignoring behavioral effects (Mussweiler 2003; Smeesters and Mandel 2006; Smeesters, Mussweiler and Mandel 2010). In this research, indeed, we demonstrate that social comparison results in negative (or positive) attitudes and evaluations of the focal product, as consumers are likely to project the negative (positive) feelings and perceptions generated by the negative (or positive) comparison to the focal product being tried on by the attractive model during the customization activity. Therefore, we demonstrate that the influence of comparison information is more widespread, extending beyond self-evaluations to evaluations of products that are customized.
appearance self-esteem. Our results demonstrate that consumers high in appearance self-esteem may be more susceptible to form negative attitudes and evaluations when exposed to an attractive model during the customization process. This evidence contrasts to past research which has instead emphasized that the affective consequences of a given social comparison is intrinsic to its direction (e.g., Argo, White and Dahl 2006; Dahl, Argo and Morales 2012). Our results are consistent with
compensatory or defensive self-enhancement theory (Allport 1937;
Swann, Pelham and Krull 1989) which suggests that individuals with positive self-views (i.e., high appearance self-esteem) are less motivated to self-enhance than are people with negative self-views (i.e., low appearance self-esteem).
Managerial Implications
Several companies are investing in customization toolkits which allow customers to visualize how the customized product (e.g., clothes, make-up, tattoos) fits on attractive models. This work has implications for those companies which strive to optimize the design experience for consumers and to increase both self-evaluations and product satisfaction. In particular, firms which decide to show attractive models during the customization process should carefully evaluate how customers might react to social information triggered by those models.
One the one hand, our research demonstrates that especially high appearance self-esteem consumers are more susceptible to form negative attitudes and evaluations when confronted with attractive models during the customization process. Indeed, those consumers prefer to customize only by visualizing the focal product. On the other hand, consumers low in appearance self-esteem are less susceptible to form negative attitudes and evaluations when confronted with attractive models during the customization process. It is thus advisable that companies which make large use of customization toolkits, allow customers to spontaneously decide - instead of forcing them - whether they want to visualize how the product fits on attractive models.
their attitudes and evaluations. As an instance Land’s End my virtual
model is an interactive virtual reality application that enables customers
to build a virtual image of themselves and then “try” on clothing. A recent report shows that Lands’ End online customers who use My Virtual Model have a 34% higher conversation rate and 8% higher average order value.
Limitations and further research
Our research is not free from limitations which can be addressed by future research. First of all, we measured appearance self-esteem (Heatherton and Polivy 1991) instead of manipulating it. In a new study it might be interesting to manipulate appearance self-esteem through bogus feedback procedures (e.g., Hoyle, Insko and Moniz 1992). This would allow us to define causal relationships between the dependent and independent variables.
Second, in our study the sample is represented by undergraduate students. A different sample should be considered in order to generalize our findings across respondents. Moreover, in our study we considered only t-shirts as the focal product. In order to generalize across categories it might be interesting to consider different type of clothes (e.g., trousers, shoes) or product categories (e.g., car interiors, make up). In a future study, it might be interesting to measure individuals motivation to self-enhance. This would allow us to test for mediation and to investigate the psychological mechanism underlying our empirical evidence.
validate our findings, it might be interesting to analyze how customers react to customization when their own physical aspects are simulated through the customization toolkit. Consistently with our results, high appearance self-esteem consumers should evaluate the customized outcome more positively because in that case, their perceived similarity to the model is high. Moreover, further studies might be conducted to analyze which specific features of customization practices might affect social comparison mechanism in customization settings.
With regard to this last point, it might be interesting to check which objectives drive individuals who engage in product customization when confronted or not with the attractive model. For example, one might check how customers high (or low) in appearance self-esteem perceive the customization task. Do low self-appearance customers perceive the customization task as a game because they are aware that the attractive model used to visualize the product is impossible to pursue for them ? Perceiving the customization task as a game, would that generate positive feeling such as fun and enjoyment that in turn generate positive customers evaluations of the customized product ? Next, do high self-appearance customers perceive the customization task more seriously because the attractive model is easier to pursue for them ? Perceiving the customization task seriously, would that generate negative feelings such as effort and frustration which in turn generate negative evaluations of the customized product ?
References
Allport G. W. (1937), Personality: A psychological interpretation. New
York: Henry Holt
Argo, J.J., White, K. and Dahl, D.W. (2006), “Social Comparison Theory and Deception in the Interpersonal Exchange of Consumption Information,” Journal of Consumer Research, 33 (2), 99-108.
Beck, A. T. (1967), Depression: Causes and treatment. Philadelphia: University of Pennsylvania Press.
Bendapudi, N., and Leone, R. P. (2003), Psychological Implications of Customer Participation in Co-Production. Journal of Marketing, 67, 14– 28.
Blascovich, J., and Tomaka, J. (1991), Measures of self-esteem. In J. P. Robinson, P. R. Shaver, & L. S. Wrightsman (Eds.), Measures of personality and social psychological attitudes (pp. 115–160). New York: Academic.
Brickman, P., and Bulman, R. J. (1977), Pleasure and pain in social comparison. In J.Suls & R.Miller (Eds.), Social comparison processes: Theoretical and empirical perspectives (pp. 149–186). Washington , DC : Hemisphere.
Brown, G. W., Andrews, B., and Bifulco, A. T. (1990), Self-esteem and depression: I. Measurement issues and prediction of onset. Social Psychiatry and Psychiatric Epidemiology, 25, 200–209.
Brown, J. D. (1993), Motivational conflict and the self: The double-bind of low self-esteem. In R. F. Baumeister (Ed.), Self-esteem: The puzzle of
Buss, A. H. (1980), Self-Consciousness and Social Anxiety. San Francisco: Freeman
Buss, D.M. and Scheier, M.F. (1976), “Self-Consciousness, Self-Awareness, and Self-Attribution,” Journal of Research in Personality, 10 (4), 463–68.
Buunk, B. P., Collins, R. L., Taylor, S. E., Vanyperen, N. W., & Dakof, G. A. (1990), The affective consequences of social comparison: Either direction has its ups and downs. Journal of Personality and Social Psychology, Vol. 59, pp. 1238–1249.
Campbell, J.D. (1986), Similarity and uniqueness: the effects of attribute type, relevance, and individual differences in self-esteem and depression, Journal of Personality and Social Psychology, Vol. 50, pp. 281–294.
Campbell, J. D. (1990), Self-esteem and clarity of the self-concept.
Journal of Personality and Social Psychology, Vol. 59, pp. 538-549.
Carver, C.S. and Scheier, M.F. (1981), Attention and Self-Regulation: A
Control-Theory Approach to Human Behavior. New York:
Springer-Verlag.
Cheng, H., and Furnham, A. (2003), Personality, self-esteem, and demographic predictions of happiness and depression, Personality and
Individual Differences, Vol. 34, 921–942.
Crocker, J. and Schwartz, I. (1985), Prejudice and Ingroup Favoritism in a Minimal Intergroup Situation: Effects of Self-Esteem,ǁ Personality and
Social Psychology Bulletin, 11 (4), 379–86.
Self-Esteem and Threat,ǁ Journal of Personality and Social Psychology, 52 (5), 907–16.
Dahl, D.W., Argo, J.J., and Morales, A.C. (2012), Social information in the retail environment: The importance of consumption alignment, referent identity, and self-esteem, Journal of Consumer Research, Vol. 38(5), pp. 860-871.
Dallaert. B.G.C. and Stremersch, S. (2005), Marketing Mass-Customized Products: Striking a Balance Between Utility and Complexity, Journal of
Marketing Research, Vol. XLII, 219–227.
Dutton, K. A., & Brown, J. D. (1997), Global esteem and specific self-views as determinants of people’s reactions to success and failure.
Journal of Personality and Social Psychology, Vol. 73, 139–148.
Duval, S. and Wicklund, R.A. (1972), A Theory of Objective
Self-Awareness. New York: Academic Press.
Fejfar, M.C. and Hoyle, R.H. (2000), “Effect of Private Self-Awareness on Negative Affect and Self-Referent Attribution: A Quantitative Review,”
Personality and Social Psychology Review, 4 (2), 132–42.
Fenigstein, A., Scheier, M.F. and Buss, A.H. (1975), “Public and Private Self- Consciousness: Assessment and Theory,” Journal of Consulting and
Clinical Psychology, 43 (4), 522–27.
Fenigstein, A. and Levine, M.P. (1984), “Self-Attention, Concept Activation, and the Causal Self,” Journal of Experimental Social
Psychology, 20 (3), 231–45.
Franke, N. and von Hippel, E. (2003), Satisfying Heterogenous User Needs via Innovation Toolkits: The Case of Apache Security Software, Research
Policy, Vol. 32, 1199–215.
Franke, N. and Piller, F. (2004), Value Creation by Toolkits for User Innovation and Design: The Case of the Watch Market, Journal of
Product Innovation Management, Vol. 21, 401–15.
Franke, N. and Schreier, M. (2002), Entrepreneurial Opportunities with Toolkits for User Innovation and Design, Working Paper, A later version of this paper is published in The International Journal on Media Management, Vol. 4(4), pp. 225-235.
Franke, N. and Schreier, M. (2008), Product Uniqueness as a Driver of Customer Utility in Mass Customization, Marketing Letters, Vol. 19, 93– 107.
Franke, N., Keinz, P. and Steger, C. (2009), “Testing the Value of Customization: When Do Customers Really Prefer Products Tailored to Their Preferences?” Journal of Marketing, 73 (5), 103–121.
Franke, N., Schreier, M., and Kaiser, U. (2010). The ‘‘I designed It Myself’’ Effect in Mass Customization, Management Science, Vol. 65, 125–40.
Franke, N. and Schreier, M. (2010) Why Customers Value Mass-customized Products: The Importance of Process Effort and Enjoyment. Journal of Product Innovation Management, 27(7). pp. 1020-1031.
Gibbons, F. X., Gerrard, M., Lando, H. A., and McGovern, P.G. (1991), Social comparison and smoking cessation: The role of the “typical smoker” . Journal of Experimental Social Psychology, Vol. 27, pp. 239 – 258.
Goffman, E. (1959), The Presentation of Self in Everyday Life. Oxford: Doubleday.
Grabe, S., L.M. Ward, and J.S. Hyde (2008), The role of the media in body image concerns among women: A meta-analysis of experimental and correlational studies, Psychological Bulletin, Vol. 134, pp. 460–476
Hayes, R.H. & Pisano, G.P. (1994), Beyond World-Class: The New Manufacturing Strategy, Harvard Business Review, 72(1), 77-86.
Hayes, A.F. and Matthes, J. (2009), “Computational Procedures for Probing Interactions in OLS and Logistic Regression: SPSS and SAS Implementations,” Behavior Research Methods, 41 (3), 924-36.
Harter, S. (1999), The construction of the self: A developmental
perspective. New York: The Guilford Press.
Henderson-King, E. and Henderson-King, D. (1997), Media effects on women's body esteem: Social and individual difference factors. Journal
of Applied Social Psychology, 27(5): 399–417.
Hirshman, E.C. and Holbrook, M.B. (1982), “Hedonic consumption: emerging concepts, methods and propositions”, Journal of Marketing, Vol. 46 No. 3, pp. 92-101.
Huffman, C. and Kahn, B.E. (1998). Variety for Sale: Mass Customization or Mass Confusion. Journal of Retailing, Vol. 74(4), 491–513.
Johnson, C.S. and D.A. Stapel (2007), “No Pain, No Gain: The Conditions under Which Upward Comparisons Lead to Better Performance,”
Journal of Personality and Social Psychology, 92 (6), 1051–67.
Leary, M.R. and Kowalski, R.M. (1990), Impression Management: A Literature Review Two-Component Model, Journal of Personality and
Social Psychology, Vol. 107, pp. 34-47.
Leahey, T.M., Crowther, J.H. and Mickelson, K.D. (2007), The frequency, nature, and effects of naturally occurring appearance-focused social comparisons, Behavior Therapy, Vol. 38, pp. 132–143.
Levav, J., Heitmann, M., Herrmann, A., and Iyengar, S. (2007), "Order in Product Customization Decisions: Evidence from Field Experiments," Working paper, Columbia Business School.
Lockwood, P. and Pinkus, R.T. (2008), “The Impact of Social Comparisons on Motivation,” in Handbook of Motivation Science, ed. James Y. Shah and Wendi L. Gardner, New York: Guilford, 251–64.
Lurie, N.H. and and Charlotte H. Mason (2007), “Visual Representation: Implications for Decision Making,” Journal of Marketing, 71(January), 160–77.
