Attitude toward the in-store experience of a sport apparel store
in non-overweight and overweight customers
Why the experience feels different for
Lukas Janßen 11143304
MSc. in Business Administration – Marketing track University of Amsterdam
Faculty of Economics and Business 27.01.2017 / Final version Supervisor: dhr. dr. Alfred Zerres
Statement of originality
This document is written by Lukas Janßen who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
The aim of the current study was to develop and test two different models that help to a) examine if and b) explain why the in-store experience of a sport apparel store might feel different for overweight customers in comparison to non-overweight customers. To test the two models and the corresponding ten hypotheses, a quantitative survey was carried out. Data were collected from 740 respondents in four different flagship stores of a well-known sport apparel brand. The findings of this study indicate that body weight has multiple significant indirect effects that negatively influence the evaluation of the store atmosphere in a sport apparel store. The reason for this is that overweight customers score lower than their non-overweight counterparts on intrinsic motivation and physical self-worth, which leads them to feel less actual self-congruity in a sport apparel store. In turn, this leads them to evaluate the store atmosphere more negatively. Moreover, body weight was also found to negatively affect the importance of the physical self-concept. As a result, overweight customers perceive less ideal self-congruity, which also adds to a more negative evaluation of the store atmosphere in comparison to non-overweight customers. The main contribution of this study is that it validates one model that accounts for the differences in the attitude towards the atmosphere in a sport apparel store between non-overweight and overweight customers. Moreover, the findings of the current study contribute to closing the gap of self-congruity research in the retail environment by proving the effect of both actual and ideal-self-congruity in a retail setting.
At this point, I would like to thank my thesis supervisor dr. Alfred Zerres from the University of Amsterdam. During the whole project, he always had a friendly ear for any questions and provided me with some helpful feedback in the process. Without him this study would not have been possible.
Furthermore, I would like to thank Asics, especially in the person of Martin Block, for making this whole project possible. Moreover, I would also like to express my gratitude to all employees from the different Asics stores where the data were collected for welcoming us with open arms.
Lastly, I would like to thank my research partners Nastie Schoenmakers, Lara Galka and Emma Kraanen for their support and friendship during this whole process.
Table of Content
1. Introduction ... 1
2. Literature Review, Hypotheses and Conceptual Model ... 5
2.1 How Body Weight and Exercise Motivation affect the Self-Concept ... 5
2.1.1 Construct Definition and Structure of Self-Concept ... 5
2.1.2 Physical Self-Concept ... 7
2.1.3 Direct Effect of Body Weight on the physical self-concept ... 8
2.1.4 Indirect Effect of Body Weight on Physical Self-Concept ... 11
2.2 How the Self-Concept affects Customer Behavior ... 18
2.2.1 Self-Concept in Customer Behavior ... 18
2.2.2 Self-Congruity ... 19
2.3 How Self-Congruity affects Store Evaluation ... 20
2.3.1 Effect of the Physical Self on Self-Congruity in a sport apparel store ... 20
2.3.2 Effect of Self-Congruity on Store Evaluation ... 23
2.4 How Body Weight affects Store Evaluation ... 26
3. Pre-Study: Development of shortened Exercise Motivation scale ... 28
3.1 Aim of Pre-Study ... 28
3.2 Data Collection, Sample and Procedure ... 29
3.3 Data Analysis and Results ... 30
4. Research Design – Main Study ... 33
4.1 Research Design ... 33
4.2 Sampling and Data Collection ... 34
4.3 Measures ... 36
5. Results – Main study ... 38
5.1 Sample Description ... 38
5.2 Statistical procedure ... 39
5.3 Validation of the shortened BREQ-2 scale ... 42
5.4 Analysis of Correlation Matrix ... 44
5.5 Hypotheses testing ... 47
5.5.1 Analytical strategy and Tests of assumptions ... 47
5.5.2 Tests of hypotheses H1 to H4 ... 49
5.5.3 Tests of hypotheses H5a and H6a ... 52
5.5.4 Tests of hypotheses H7 and 8 ... 56
5.5.5 Tests of hypotheses H5b and H6b ... 59
6. Discussion ... 64
6.1 Academic contributions ... 64
6.2 Managerial implications ... 69
6.3 Limitations and future research ... 70
“Want it more”, “Just Do It”, “Forever Faster” – When it comes to brand positioning, sports apparel brands have a clear opinion about how to reach their target group of performance-driven athletes. Brands like Asics, Nike and Puma use symbolic cues in their branding strategy to evoke certain images of the brand or the typical brand user. Advertisements as shown above are a typical example of such cues and portray people that can compete in different kind of sports and engage in heavy and enduring physical-sport exercise. Sirgy, Grewal and Mangleburg (2000) argue that these cues are also used in the retail environment to manifest the identity of the brand and shape the image of the typical store patron, for example in form of in-store communications. On entering a sport apparel store, one cannot avoid to perceive these cues. Videos of athletes running up a hill through the mud, or posters of athletically trained people, are an inherent part of the in-store experience and strengthen the image of the performance-driven, athletic brand user.
Self-Congruity theory proposes that a greater match between the image of the typical store patron and the self perceptions of the customer should lead to a more favorable store attitude (Sirgy et al., 2000). Thus, for customers who feel congruent with these brand images, the store experience should be evaluated positively, as shopping at these stores should reinforce their self-perceptions (Sirgy, 1982). But what about those customers that do not have such a view of themselves? Britt (1960) argues that customers might avoid shopping at a certain store if there is a mismatch with the customer’s self-perceptions. The reason for this can be derived from cognitive-consistency theories (Lecky, 1945; Heider, 1946; Festinger, 1957), which argue that behavior inconsistent with the own self lead to dissonance, a state of discomfort and tension (Sirgy, 1985).
2 Although it seems insensitive and slightly controversial in first instance, I argue that for overweight customers this state of incongruity is more likely to occur in a sport apparel store than for their non-overweight counterparts. Even though the self-congruity matching process is an internal, cognitive process (Sirgy, 1985) that mainly depends on the self-perceptions of the customer, body weight might have a crucial influence on this process through its impact on physical self-perceptions. The underlying rationale of this statement is briefly introduced in the following.
As previously indicated with the impact of self-congruity, the self-concept plays a key role in the behavior of customers. Self-concept refers to the “totality of the individual's thoughts and feelings having reference to himself as an object” (Rosenberg, 1979, p. 7). One part of the self-concept is made up of thoughts and feelings about the own body in terms of appearance and abilities. Fox and Corbin (1989) argue that everyone has perceptions about the own strength, endurance, competence and attractiveness in the physical domain.
Various studies (e.g. Friedman & Brownell, 1995; Miller & Downey, 1999) indicate that perceptions about the self are negatively impacted by overweight. For example, a study by Aşçı (2005) present findings that body fat percentage directly affects perceptions of physical condition and body attractiveness. Moreover, body weight may have an indirect effect on physical self-perceptions through the concept of exercise motivation. Hwang and Kim (2013) argue that overweight individuals are more likely to lack motivation to exercise or feel pressured by external forces (e.g. friends, family, society). Subsequently, this lack of motivation or absence of perceived control was found by Thøgersen-Ntoumani and Ntoumanis (2006) to negatively impact physical self-perceptions.
Considering that a sports apparel store should evoke strong stereotypic images of a person that is confident in the own physical self-perception (i.e. image of a person that believes to be highly competent in different kind of sports etc.), the negative impact of overweight on physical
3 self-perception’s should have an influence on the self-congruity matching process. Based on cognitive-consistency theory, overweight customers should be more likely to perceive discrepancies between the own physical self-concept and the image of the typical sport apparel store patron and in turn, evaluate the store atmosphere more negatively.
However, despite perceived congruity, the store experience does not necessarily need to be evaluated negatively. Self-Congruity Theory argues that customers may also be driven by the need for self-esteem (Sirgy, 1985; Sirgy et al., 2000). People have an ideal image of how they would like to be and behavior consistent with this ideal image helps individuals to enhance their self-concept (Sedikides & Gregg, 2008). It might be that overweight customers aspire to be as athletic and competent in sports as the typical store patron of a sports store. Thus, patronizing such a store could lead them to get closer to their ideal self-image.
A wealth of research has examined the constructs of the physical self, exercise motivation and self-congruity in their respective fields of the social and behavioral sciences. However, to my best knowledge, no research has connected these constructs and examined how they might affect the customer experience in a sport apparel store setting. Therefore, the current paper aims to develop and test two different models that helps to a) examine if and b) explain why the in-store experience of a sport apparel store might feel different for overweight customers. The main research questions of this study have been defined as follows: “(1) How does the body weight of a
customer affect the store experience in a sport apparel store? (2) To what extend is this effect mediated by the physical self-concept of the customer, exercise motivation and the construct of actual self-congruity? and (3) How does ideal self-congruity influence the store atmosphere?”.
Even though, there has not been an explicit call for studies combining the different constructs, the current study in the sport apparel store setting warrants examination from both the academic and practical perspective for several reasons. Given the recent estimations of the World
4 Health Organization (2008) that over 50 percent of the WHO European Region are overweight and around 23 percent of women and 20 percent of men obese, findings on the impact of overweight on the in-store experience in a sport apparel are deemed to be of importance for both research on customer behavior and exercise behavior on a large scale.
For one, examining the possible differential effect of body weight on the perceived store experience provides sport apparel retailers with relevant insights. Sales people use physical attributes of customers (e.g. age, gender, clothing style) as proxies for needs and motivations (Sujan, Weitz, & Sujan, 1988). This categorization process is a vital process that helps salespeople to adapt their service behavior to the specific customer (Solomon, Surprenant, Czepiel, & Gutman, 1985). Thus, the awareness of possible differences in the perceived store experience for overweight and non-overweight customers could enable the sales personnel to better target and meet the needs of the different customers.
In general, the current study also contributes to closing the gap of self-congruity research in the field of retailing by adding findings on the effect of self-congruity on store evaluation. Despite the importance of self-congruity in consumer behavior (Aguirre-Rodriguez, Bosnjak, & Sirgy, 2012), the concept has been largely disregarded in this area (Sirgy et al., 2000; Chebat, El Hedhli, & Sirgy, 2009). Considering that image concepts are highly salient and influential in the apparel industry (Zentes, Morschett, & Schramm-Klein, 2008), the sport apparel setting seems to be well-suited to test the effect of self-congruity in retailing.
Lastly, the current study might also have important implications for research on health and exercise behavior. To this day, participation in physical exercise remains generally low (Haskell et al., 2007) and especially overweight individuals often lack any intention to exercise (Hwang & Kim, 2013) despite the important physical (Lee, Sesso, Oguma, & Paffenbarger, 2003) and psychological (Fox, 1999) health benefits. Findings of perceived tension and dissonance in
5 overweight customers when shopping for sport apparel might uncover another perceived barrier to exercise behavior that overweight individuals face.
In the next chapter of this research, the literature review gives insight into the physical self-concept and how this might be affected by body weight and exercise motivation. The literature review builds up to a conceptual model that is aimed to explain a possible differential effect of body weight on store atmosphere. Next, detailed information about the research design and data collection are given. Afterwards, the results of the collected data are presented together with a discussion of the findings. The study ends with academic and practical implications, limitations and suggestions for future research.
2. Literature Review, Hypotheses and Conceptual Model
2.1 How Body Weight and Exercise Motivation affect the Self-Concept 2.1.1 Construct Definition and Structure of Self-Concept
For decades now, understanding the self has played a significant part in the field of the social sciences. Research of the self-concept has a long history and is generally linked to the question of “Who am I?” (Weiten, Dunn, & Hammer, 2014). In a review of self-concept definitions, Shavelson, Hubner, & Stanton (Shavelson, Hubner, & Stanton, 1976) provide a broad definition of the construct as they define self-concept as a person’s self-perceptions of him or herself that are formed by experiences with the environment and are influenced by reinforcements, significant others and attributions for a person’s own behavior. These self-perceptions are related to a person’s “own abilities, limitations, appearance and characteristics, including one’s own personality” (Graeff, 1996, p.5). In other words, self-concept is viewed as the sum of a person's thoughts and feelings regarding themselves as an object (Rosenberg, 1979). Thus, different from other attitudes, which
6 mostly are a result of perceptions about an external object, self-concept is “an image shaped by the very person holding the image” (Hong & Zinkhan, 1995, p. 54).
Shavelson et al. (1976) emphasize that self-concept is a hypothetical construct that can be used to explain and predict a person’s actions. This notion is supported by Prince (1993, p. 161) who states that “the self […] is a driving force for much of human behavior”. In turn, a person’s actions influence his or her self-perceptions (Shavelson et al., 1976). Therefore, Marsh and Craven (2006) noted that self-concept is not only important as an outcome variable, but also as a mediator variable that helps in understanding other outcomes.
Research concerning the structure of the self-concept has a long and controversial history. Structural models of the self-concept have been plotted as either unidimensional or multidimensional (Marsh & Hattie, 1996). In the unidimensional model, self-concept is conceptualized as a one-factor construct or by one general factor that dominates other factors (Marsh & Craven, 2006). The global factor is typically referred to as the general self-concept or sometimes also as self-esteem (Marsh, Parada, & Ayotte, 2004). In contrast, the multidimensional perspective distinguishes between multiple facets of self-concept (Marsh & Shavelson, 1985). The multidimensional model as proposed by Shavelson and colleagues (1976) is set up hierarchically. Like the unidimensional perspective, the multidimensional model places the general self-concept at the apex of the hierarchy. Below the general self-concept, the self is divided into the academic and non-academic self-concept.
The academic self-concept comprises a person’s self-perceptions about the own academic skills and abilities (e.g. language skills, mathematics knowledge) (Byrne, 1984; Shavelson et al., 1976). The non-academic self-concept is distinguished again into the social, emotional and physical self-concept (Shavelson et al., 1976). Social self-concept refers to a person’s self-perception about the own competences in social interactions with significant others or peers (Byrne & Shavelson,
7 1996). The emotional self-concept comprises a person’s emotional states including anxiety, love, anger, depression, and happiness (Shavelson et al., 1976; Kõiv, 2016). Lastly, the physical self-concept consists of a person’s self-perception of the own bodily appearance and physical abilities (e.g. sport competences, physical endurance etc.) (Fox & Corbin, 1989; Shavelson et al., 1976).
After an initial focus on the unidimensional self, contemporary self-concept research in the social sciences advocates the multidimensional and hierarchical model of self-concept as introduced by Shavelson et al. (1976) (Marsh & Craven, 2006).
2.1.2 Physical Self-Concept
With the acceptance of the multidimensional, hierarchical self-concept, research started focusing on studying the different self-concept domains. Especially the physical self-concept has received vast amount of attention in the self-concept literature, reflecting the centrality of the body to the self. In one of the earliest studies conducted by Prelinger (1959), the body was shown to be most linked part to the self with a rating of 2.98 on a 3-point not-self to self-continuum scale, where a higher rating means a closer affiliation to the self. Even though the research design of Prelinger’s study (1959) led to more extreme results, Belk (1988) found support for Prelinger’s results concerning the centrality of the body. In a review of all the elements that form the self, Mittal (2006) concluded that the physical self is an integral part of the self for most individuals. Especially many individuals in western societies are often obsessed with their bodies and see it as the “end-all and the be “end-all of their being” (Mittal, 2006, p. 556).
As stated above, the physical self-concept is a domain of the non-academic self-concept. The physical self-concept is multidimensional and incorporates several more specific sub-domains that lead to a general evaluation of physical self-worth (Maïano et al., 2008). On a sub-domain level, Shavelson et al. (1976) distinguish in their conceptual paper broadly between general physical abilities and physical appearance.
8 Following the multidimensional, hierarchical model of Shavelson et al. (1976), Fox and Corbin (1989) proposed a more nuanced structure of the physical self-concept sub-domain dimensions with the use of factor analysis. Their findings suggest that the physical self-concept is superordinate to four independent, but correlated sub-domains: sport competence, perceived bodily attractiveness, physical strength and physical condition. Sport competence refers to a person’s perception about the own athletic skills and the ability to learn and exert sports. Bodily attractiveness reflects the perception of a person’s attractiveness about the own figure or body type, as well as the level of confidence into the own ability to maintain an attractive body over time. The perception a person has about stamina and fitness is captured by physical endurance. Lastly, physical strength refers to the perceived muscle strength and development (Thornton & Kato, 2012).
Contemporary measures of the physical self-concept (Eklund, Whitehead, & Welk, 1997; Maïano et al., 2008; Marsh, Richards, Johnson, & Roche, 1994; Ninot, Bilard, Delignieres, & Sokolowski, 2000) all give support to the complexity of the physical self. Although the model proposed by Fox and Corbin (1989) remains one of the most frequently used physical self-concept measures, conceptual refinements of this model by Marsh et al. (1994) led to the current view of the physical self-concept as a hierarchical construct consisting of the following sub-domains: strength, body fat, activity, endurance and fitness, sport competence, coordination, health, appearance, and flexibility (Dishman et al., 2006).
2.1.3 Direct Effect of Body Weight on the Physical Self-Concept
In their cornerstone article, Fox and Corbin (1989) propose that physical characteristics could serve as antecedents to physical self-perceptions. Thus, physical attributes have received vast attention in the self-concept literature over the last decades (Miller & Downey, 1999). Special focus has been directed on how body weight and deviations from the cultural norms influence perceptions of
9 the self. Not conforming to standards of society about being thin was regularly hypothesized to have a strong and negative impact on global self-esteem (Pesa, Syre, & Jones, 2000). Findings on whether being overweight negatively influences the self-concept have been inconsistent, with some studies rejecting this notion (J. Crocker & Major, 1989; Jarvie, Lahey, Graziano, & Framer, 1983), and others finding support (Friedman & Brownell, 1995; Sarwer, Thompson, & Cash, 2005).
From a theoretical perspective, prejudice against and the stigmatization of overweight individuals in the Western culture indicates that overweight should have a negative impact on esteem (Miller & Downey, 1999). Stigmatization leads to a socially devalued identity as self-esteem is influenced by the appraisal of others’ (J. Crocker, Major, & Steele, 1998). The awareness of overweight individuals that others have negative prejudices against them depresses self-esteem (Miller & Downey, 1999). Another major factor for the negative impact of overweight on self-esteem is derived from them self-efficacy model, which is concerned with one’s belief in the own ability to accomplish a certain task (Bandura, 1986). With each failure to lose weight diminishing the belief in the own ability, self-esteem is subsequently bound to suffer (Miller & Downey, 1999). Society’s belief that overweight individuals could lose weight easily if they only had the willpower and self-discipline to follow through with weight loss, only further reinforces the negativity spiral (Crandall & Biernat, 1990).
One reason that might account for the inconsistent findings is that not for all overweight individuals, self-esteem or physical self-worth suffer to the same degree. Factors including the degree of overweight (Schwartz & Brownell, 2004) and internalization of negative messages (Cafri, Yamamiya, Brannick, & Thompson, 2005) influence the relationship between overweight and esteem. Moreover, there is a significant difference between actual overweight and the self-classification of being overweight. In a study by Cash & Hicks (1990), the authors find that only the belief to be overweight and not the actuality lead to poorer self-esteem. However, in a
meta-10 analysis of the relationship between overweight and self-esteem by Miller & Downey (1999), the authors report that both actual overweight and self-perceived overweight influence self-esteem. Although the negative effect is significantly stronger for self-perceived overweight (d=-.72, r=-.34) than for actual overweight (d=-.24, r=-.12), both factors significantly contribute to lower self-esteem.
While research has studied extensively the negative effect of overweight on self-esteem, the impact of overweight on the physical self-concept has received less attention. Despite the scarcity of research in this field, I predict that overweight also has a negative effect on physical self-worth. I discuss the rationale behind this statement in the following.
To understand why overweight should also affect the physical self-concept, one must acknowledge the hierarchical nature of the self-concept (Shavelson et al., 1976). Low self-esteem should be a result of negative self-perceptions in one or more of the academic or non-academic domains (Shavelson et al., 1976; Tiggemann, 2004). Thus, if overweight negatively affects the general concept (esteem), this effect should mainly be a result of a low physical self-concept, since this domain comprises thoughts and feelings about the own body (Fox & Corbin, 1989). In line with that arguing, O’Dea (2006) presents findings that indicate that overweight status negatively impacts the physical self-concept and in turn decreases general self-worth.
The main reasons why overweight should negatively affect the physical self can be derived from the previously stated stigmatization of and prejudice against overweight individuals, and self-efficacy theory. Society often judges overweight individuals as ugly and lazy (Latner & Stunkard, 2003; Wardle, Volz, & Golding, 1995). As stated previously, the awareness of these negative prejudices was shown by Miller and Downey (1999) to depresses the self-esteem of overweight individuals. Again, considering the hierarchical nature of self-concept, the negative effect impact on self-esteem should be a result of a low physical self-concept. In particular, the prejudices of
11 being ugly and lazy should have a negative effect on the physical sub-domains of bodily attractiveness and physical abilities, which in turn decrease physical self-worth. Support for this arguing comes from Aşçı (2005), who presents findings that body fat percentage is negatively related to physical endurance and bodily attractiveness.
Furthermore, based on self-efficacy theory (Bandura, 1986) overweight status should have a negative impact on a person’s perception of the own physical skills and abilities. In other words, overweight should lower a person’s belief in the own ability to learn and perform different kind of sports or engage in physically enduring exercise. Deforche, De Bourdeaudhuij, and Tanghe (2006) support this notion as they report that overweight individuals score lower on perceived sport competences in comparison to non-overweight individuals.
Thus, given the previous findings on the negative link between body weight and the self (Friedman & Brownell, 1995; Miller & Downey, 1999), the stigmatization of overweight individuals and self-efficacy theory, I argue that due to the hierarchical nature of the self-concept, overweight should have a negative effect on the physical self-concept. Reason being that overweight negatively influences the physical self sub-domains that form the evaluation of physical self-worth. Therefore, I introduce the following hypothesis:
H1: Overweight customers hold a more negative physical self-concept than non-overweight customers.
2.1.4 Indirect Effect of Body Weight on Physical Self-Concept
Exercise motivation offers another angle on how body weight might affect the physical self. The concept of motivation has received a lot of attention in exercise psychology as it is an essential part in studying human’s physical-sport activity behavior (Pelletier et al., 1995). Vallerand and Thrill (1993, p. 18) define motivation as “the hypothetical construct used to describe the internal and/or external forces that produce the initiation, direction, intensity, and persistence of behavior”.
12 Contemporary research on exercise motivation is mainly based on the Self-Determination Theory. This theory argues that human behavior is driven by the need for autonomy, competence and relatedness (Ryan & Deci, 2000). The need for autonomy is fulfilled when a person feels free of external control and decisions are made by one’s own choice. Competence refers to challenging behavior that a person is able to master effectively, and relatedness comprises feelings of connection to others in a social context (Wilson & Rodgers, 2002)
According to Self-Determination Theory (Deci & Ryan, 2002), the concept of exercise motivation is best represented as a self-determination continuum. The continuum ranges from low to high levels of self-determination and includes the following types of motivation (ordered from low to high): amotivation, extrinsic motivation (external, introjected, identified, integrated regulation) and intrinsic motivation (Ryan & Deci, 2000). The central distinction between the different motives in the self-determination continuum is between those motivations that are autonomous (intrinsic, integrated, and identified regulation) and those that are controlled (introjected, and external regulation) (Deci & Ryan, 2008).
With autonomous motivations, individuals have identified themselves with an activity’s value and ideally integrated it into their self (Deci & Ryan, 2008). With intrinsic motivation, individuals are likely to experience feelings of enjoyment, fun and inherent satisfaction derived only by participating in a specific behavior. The individual perceives the engagement in the behavior in itself as rewarding (Ryan & Deci, 2000). With integrated regulation, motivation stems from the congruity of the target behavior with the individual’s goals, e.g. engaging in weight-training early in the morning in order to avoid conflict with other daily duties (Tenenbaum & Eklund, 2012). In contrast, with controlled motivations, individuals’ actions are driven by external contingencies of reward or punishment. External regulation, for example, as the least self-determined extrinsic motivation, refers to the motivation toward a target behavior that resides
13 outside of the individual and within the social environment. Engagement in the target behavior can stem from obtaining a reward or avoiding punishment (Pelletier et al., 1995). With amotivation, individuals neither identify with the behavior, nor do they see the contingencies between the behavior and the separable outcome, and thus are characterized by a lack of intrinsic and extrinsic motivation (Pelletier et al., 1995; Ryan & Deci, 2000).
Research on the consequences of the different types of motivation has shown mixed results. The more autonomous motivations have been shown to cause positive outcomes (Tenenbaum & Eklund, 2012). Carbonneau, Vallerand and Lafrenière (2012) present findings that intrinsic motivation leads to positive affective consequences (e.g. entertainment, proudness, curiosity). Deci and Ryan (2002) link intrinsic motivation to higher levels of general well-being. Furthermore, findings by Nien and Duda (2008) suggests that intrinsic motivation lead to the adoption of approach rather than avoidance orientations because of a high perception of perceived competence under intrinsic motivation. On the other hand, research on controlled motivations produced mixed results, with some finding negative outcomes and others finding no relation to outcomes (Tenenbaum & Eklund, 2012). Ntoumanis (2001) provide findings that indicate the negative consequences of amotivation. The author finds that amotivation is related to boredom, lower effort, and less future intention to engage in physical activity. The study also reports that only intrinsic motivation elicits positive outcomes. These results are in line with Vallerand (2001; 2007), who states that the outcomes are becoming less positive as motivations move down on the self-determination continuum towards amotivation.
Self-Determination Theory suggests a differential impact of the various forms of self-determined motivations on physical self-perceptions (Ryan & Deci, 2000). In line with more general findings, research suggests an increasingly positive relationship between the more self-determined motivations and the physical self-concept. In a longitudinal study, Martín-Albo, Núñez,
14 Domínguez, León, & Tomás (2012) examined the impact of intrinsic motivation on the different sub-domains of the physical self-concept (strength, endurance, attractiveness and competence). Their findings suggest a significant positive effect of intrinsic motivation on the physical self-concept with the impact increasing over time. Furthermore, their study gives support for previous results of a cross-sectional study by Georgiadis, Biddle and Chatzisarantis (2001), that also reported on the positive effect of intrinsic motivation on the physical self-concept. Amotivation and external regulation, on the other hand, were reported by Thøgersen-Ntoumani and Ntoumanis (2006) to negatively affect physical self-perceptions.
The underlying rationale for these findings can be derived from social physique anxiety and self-determination theory. Due to the scope of this research, I only examine the relationship between intrinsic motivation and the physical self-concept in the current study.
Social physique anxiety refers to the apprehension a person holds that one’s own body is negatively judged by other people (Hart, Leary, & Rejeski, 1989). With social physique anxiety, the attainment of physical self-worth is contingent on a socially construed body ideal (Thøgersen-Ntoumani & (Thøgersen-Ntoumanis, 2006). Contingent worth is particularly problematic for physical self-perceptions, since it increases perceived surveillance and body shame (Patrick, Neighbors, & Knee, 2004). Thus, individuals with high social physique anxiety are less likely to evaluate their physical self positively (P. Crocker et al., 2003).
Whereas controlled exercise motivations, e.g. exercising to enhance one’s physical attractiveness, were found by Crawford and Eklund (1994) to lead to high social physique anxiety, intrinsic motivation was reported to have a negative impact on social physique anxiety. With intrinsic exercise motivation, the enjoyment and pleasure derived from the activity reduces the concern about negative evaluations by others (Thogersen-Ntoumani & Ntoumanis, 2007). In
15 contrast, exercising to increase one’s own physical appearance or out of guilt, make these concerns only more salient (Thøgersen-Ntoumani & Ntoumanis, 2006).
Another construct that accounts for the relation of intrinsic exercise motivation and the physical self is determination theory. As stated previously, this theory proposes that self-esteem is built by pursuing actions that are autonomous, and evoke feelings of competence and relatedness (Deci & Ryan, 1995). Although self-determination theory is mainly applied to general esteem, I argue that the different determined motives should also affect physical self-worth. Especially, considering that feelings of competence are particularly crucial in the development of physical self-worth (Fox, 1997).
To understand how intrinsic motivation fosters perceived competences, two different dimensions of intrinsic motivation must be acknowledged, namely intrinsic motivation to know and intrinsic motivation to accomplish things. With intrinsic motivation to know, a person engages in exercise to learn and explore new skills and competences. Intrinsic motivation to accomplish things, on the other hand, leads a person to experience pleasure from exercising by improving the own skills and competences (Pelletier et al., 1995). Thus, both dimensions should increase feelings of perceived competence, especially in comparison to less self-determined motives where a person engages in exercise to avoid guilt or shame rather than to learn or improve the own skills and abilities. In line with this argumentation, Wilson & Rogers (2002) and Murcia, Camacho and Rodriguez (2008) present findings on the positive impact of intrinsic motivation on the physical self-concept in their respective studies, with the latter finding a particular strong impact of intrinsic motivation on sport competences.
Furthermore, intrinsic motivation should be more likely to satisfy the need for autonomy, since the person has identified himself with the activity’s value and integrated it into his self (Deci
16 & Ryan, 2008). Subsequently, fulfilling the need for autonomy should under intrinsic motivation should increase physical self-worth.
Based on the previous literature review, and in particular the concept of social physique anxiety and self-determination theory, I argue that intrinsic exercise motivation reduces social physique anxiety, increases a person’s perceived physical skills and abilities, satisfies the need for autonomy and as a result has a positive effect on physical self-worth. Therefore, I introduce the following hypothesis:
H2: There is a positive relation between intrinsic motivation and physical self-concept.
Several studies have shown that overweight influences the degree of self-determined exercise motivation. For example, Gillison, Standage and Skevington (2006) find that overweight leads to less self-determined exercise motivation. Similar results are presented by Hwang and Kim (2013), who report that overweight and obese adolescents are less likely to be intrinsically motivated and instead scored higher on amotivation and external regulation in comparison to their normal-weight counterparts.
The rationale for these findings can be derived from differences in the enjoyment of exercise between overweight and non-overweight individuals. Overweight status negatively affects enjoyment, pleasure and fun a person feels during exercise (Deforche, De Bourdeaudhuij, & Tanghe, 2006; Ekkekakis & Lind, 2006). To understand why this effect occurs, one must acknowledge that feelings of competence are a crucial factor for experiencing enjoyment and pleasure in exercise situations (Deci & Ryan, 1995). Considering that overweight individuals score lower on perceived sport competences in comparison to non-overweight individuals (Deforche et al., 2006), overweight status should lead to less enjoyment in exercise situations and thus, should decrease intrinsic motivation.
17 Furthermore, overweight individuals often feel pressured by external sources (e.g. by friends, family or society) to exercise (Hwang & Kim, 2013). Internal motives (i.e. exercising for its own sake and the pleasure derived from it) are bound to suffer as these external reasons decrease a person’s perceptions of autonomy. As a result, individuals are less likely to identify themselves with the value of the activity and will not integrate it into their self (Deci & Ryan, 2008), which is the premise of intrinsic motivation.
Although the previously stated positive effect of intrinsic exercise motivation should be the same for non-overweight and overweight individuals, based on the previous literature, I argue that for overweight individuals the positive effect of intrinsic motivation is less likely to occur. Reason being that because of their lower perceived sport competences and their greater lack of perceived control, overweight individuals are less likely to experience pleasure and enjoyment from exercise. Thus, overweight individuals are more likely to be driven by less self-determined motives (i.e. external regulation or amotivation) rather than by intrinsic motives. Thus, I introduce the following hypothesis:
H3: Overweight customers score lower on intrinsic motivation than non-overweight customers Finally, based on the previous two hypotheses, I posit that body weight has a negative indirect effect on physical self-concept through intrinsic motivation. More specifically, in comparison to non-overweight customers, overweight customers will be less intrinsically motivated and as a result score lower on the physical self-concept. Therefore, I introduce the following hypothesis:
H4: Overweight customers score lower on physical self-concept than non-overweight customers because of the mediating effect of intrinsic motivation.
18 2.2 How the Self-Concept affects Customer Behavior
2.2.1 Self-Concept in Customer Behavior
The self-concept has received quite some attention in the customer behavior literature (Hosany & Martin, 2012) and plays an important role in understanding customers actions’ (Belk, 1988; Sirgy, 1985). This stream of research focuses on the influence of the self-concept on the consumption of brands, products and services (Sirgy, 1982).
Brands have their own personality and image (D. A. Aaker, 1996), which is created by a variety of different factors such as marketers’ actions (e.g. advertising), pricing, images of the typical user, and other psychological associations (Grubb & Grathwohl, 1967). People consume brands not only for functional, but also for those symbolic associations (Levy, 1959). With many products, symbolic benefits are even the driving force behind purchase decisions as they allow for expression (D. A. Aaker, 1996) and enable customers to manifest and enhance their self-concept (Hosany & Martin, 2012; Sirgy, 1985).
Like self-concept research in the social sciences, research in the field of customer behavior initially conceptualized self-concept as a unidimensional construct (Bellenger, Steinberg, & Stanton, 1976; Grubb & Hupp, 1968). More recent marketing literature though also advocates the view on the diversity and multidimensionality of the self-concept (Mittal, 2006; Sirgy & Su, 2000). However, different from the self-concept in the social sciences, marketing literature mostly conceptualizes the concept in terms of the four dimensions of actual concept, ideal self-concept, social self-concept and ideal social self-concept (Sirgy & Su, 2000). Actual self-concept refers to how customers see themselves and ideal self-concept to how they would like to see themselves. How customers believe that they are seen by significant others is referred to as the social self-concept and the ideal social self-concept reflects how customers would like to be seen by significant others (Belch, 1978; Malhotra, 1988; Sirgy, 1982; Sirgy & Su, 2000). In their
19 overview of self-concept studies in customer behavior, Hosany and Martin (2012) note that most empirical research only examines the dimensions of the actual and ideal self. Due to the limited scope of this research, only the actual and ideal self will be examined in this study.
Another well-regarded conceptualization of the customer self comes from Mittal (2006). In his conceptual essay, Mittal (2006, p. 552) propose that a set of six components (“(a) our bodies; (b) our values and character; (c) our success and competence, (d) our social roles, (e) our traits, and, […] (f) our possessions”) makes up the self. These components show conceptual similarities to the concept by Shavelson et al. (1976) and suggest the importance of the physical self-concept in customer behavior. Mittal (2006) also provides four aspects of the body that have been used by customers to define themselves, namely body shape, weight, look and facial appearance.
A person’s self-concept does not impact customer behavior directly, but rather through its alignment with the image of a brand or product. This matching process is referred to as self-congruity (Sirgy, 1982). In other words, self-self-congruity (often also self-congruence, self-image congruence, and image congruence) refers to the mental process of matching the customers’ self-concept with a certain brand, product or brand user (Sirgy et al., 1997; Sirgy et al., 2000). Brand user image refers to the perception of the stereotypical user of that particular brand (Sirgy, 1986). Because of the four-different kind of self-concept dimensions (actual, ideal, social, ideal social), Sirgy et al. (2000) argue that self-congruity should be distinguished into four respective types of self-congruity, namely actual self-congruity, ideal self-congruity, social self-congruity and ideal social self-congruity. Whereas actual self-congruity refers to how well the customer’s actual self-concept is aligned with the brand/product/service, ideal self-congruity refers to the degree of fit between the customer’s ideal self-concept and the brand/product/service (Sirgy et al., 2000).
20 There are two major streams of research on self-congruity. Whereas one research stream examines the relationship between self-congruity and pre-purchase evaluations, another, more recent stream focuses on the relationship between self-congruity and post-purchase evaluations. Research on pre-purchase evaluations illustrates the impact of self-congruity on purchase intentions (Hong & Zinkhan, 1995; Adcroft, Teckman, Hee Kwak, & Kang, 2009), product preference (Ericksen, 1997; Hong & Zinkhan, 1995), brand preference (Sirgy et al., 1997) and product choice (Malhotra, 1988; Sirgy et al., 1997). Moreover, studies on post-consumption evaluations show the impact of self-congruity on satisfaction (Ekinci & Riley, 2003; Jamal & Al-Marri, 2007; Sirgy et al., 1997), perceived product quality (Adcroft et al., 2009; He & Mukherjee, 2007), brand loyalty (Kressmann et al., 2006) and brand attitudes (Ibrahim & Najjar, 2008).
Despite the importance of self-congruity in customer behavior, the concept has been largely disregarded in retailing (Sirgy et al., 2000; Chebat et al., 2009). However, Sirgy et al. (2000) developed an integrative model to examine the influence of self-congruity in the retail environment. As literature in this area is still scarce, the integrative model by Sirgy et al. (2000) will serve as the foundation for the research at hand.
2.3 How Self-Congruity affects Store Evaluation
2.3.1 Effect of the Physical Self-Concept on Self-Congruity in a sport apparel store
Sirgy et al. (2000) argue that a retailer uses multiple store cues to create store images that reflect the needs of their target customer. As a result of the retailer’s efforts, customers will form a specific image of the store’s typical customers.
The authors mention to two different kind of cues, some being uncontrollable and others being controllable. Whereas uncontrollable cues reflect personal characteristics of the customer, the latter refer to the four Ps of the marketing mix. Product-related cues lead customers to make inferences about the typical store clientele as a direct function of the assortment quality and the
21 different products of the store. Place-related cues, on the other hand, refer to the color, lightning, and music of the store, as well as the to the location of the store. Price-related cues drives customers to make inferences about the typical store clientele as a direct function of the pricing of the products (e.g. high prices for status-conscious customers). Lastly, promotion-related cues lead customers to make inferences about the typical store clientele as a direct function of a retailer’s in-store communication. The authors give the example of an upscale department store consisting of in-store advertising reflecting economic status (Sirgy et al., 2000).
A study by Guzmán, Abimbola, Brengman and Willems (2009) gives support for the proposition by Sirgy et al. (2000) regarding the influence of cues on store image. The authors report that the store image is primarily determined by the store environment, store design, and other factors including the merchandise sold and the sales personnel.
The current study tries to examine how the physical self-concept affects self-congruity in a sport apparel store. Naturally, a sport apparel store should evoke strong stereotypic images of a person with an attractive body, that is competent in different kind of sports, and able to engage in heavy, enduring physical activity. Although, to my best knowledge, research on this topic is non-existent, the rationale for this argumentation can be transferred from Sirgy et al. (2000). The authors argue that, as stated previously, retailers position their stores in a certain way by using cues in the store environment to evoke a specific image (e.g. an upscale store uses specific in-store cues to evoke images of upscale shoppers). In the same vein, the current study argues that a sport apparel brand tries to manifest the perception of being a brand for athletic and sporty customers by, for example, using promotion-related cues (i.e. showing in-store videos of the generalized target customer engaging in physical activity, e.g. running up a hill).
The argument that a sport apparel store should evoke strong stereotypic images of a person with an attractive body, that is competent in different kind of sports etc., becomes more apparent
22 when one considers the following statement about a new campaign from a well-known sport apparel brand. “With this campaign, [we] aim to inspire and motivate athletes to reach their goals. We know success in sport comes from the human pursuit to take performance to a whole new level. Through this effort, [we] celebrate the passion that connects professionals with everyday warriors, and shares that spirit across the world” (Miles, 2016). Considering that such campaigns often are advertised in a brand’s retail store, these cues should contribute to the previously described image. Furthermore, for a sport apparel store this image should be particularly strong if the brand is known to be favored by professional athletes, considering that the physical self-concept for elite athletes is higher than for nonelite groups (Marsh, Hey, Roche, & Perry, 1997). Although it is difficult to know for sure whether other people perceive these differences, it can be assumed that people do believe that professional athletes are more competent in the different physical domains, and thus have a higher physical self-concept.
Considering that actual self-congruity refers to how well the customer’s actual self-concept is aligned with the image of the typical store patron (Sirgy et al., 2000), I argue that a customer with a lower physical self-concept will experience less actual self-congruity in a sport apparel store, where the typical store patron is imagined to have a positive physical self-concept. In other words, a person that does not believe in his physical appearance and/or abilities should feel discrepancies, and thus incongruity, when comparing himself to a person that is believed to feel strongly about these physical self-concept facets. Therefore, I introduce the following hypothesis:
H5a: The greater (lower) the physical self-concept of a customer, the greater (lower) the actual self-congruity in a sport apparel store.
To this date, there is no research on what influences ideal self-congruity in a sport retail store. However, I argue that whether a customer will experience ideal self-congruity in a sport apparel store should in part depend on how much importance the person places on the physical
23 self-concept. A person may perceive himself to lack the ability to perform different kind of sports. However, it might be the case that the customer still places high importance on the physical self-concept. i.e. values being competent in sports, physically strong and enduring, as well as attractive, and thus wants to get closer to that kind of perception.
Considering that ideal self-congruity refers to how well the customer’s ideal self-concept is aligned with the image of the typical store patron (Sirgy et al., 2000), I argue that the importance a customer places on the physical self-concept will determine whether this customer perceives ideal self-congruity in a sport apparel store. Again, this argumentation follows the rationale of Sirgy et al. (2000), who gives the example of a working-class person that wants to become more upscale, and therefore patronizes an upscale department store, leading the person to perceive ideal self-congruity in that kind of store.
Thus, based on the previous sections, I hypothesize that a person that places high importance on the own physical self and aspires to be like the typical store patron of a sport apparel store, should feel high ideal self-congruity in a sport apparel store. Therefore, I introduce the following hypothesis:
H5b: The greater (lower) the importance of the physical self-concept for a customer, the greater (lower) the ideal self-congruity in a sport apparel store.
2.3.2 Effect of Self-Congruity on Store Evaluation
As previously stated, research on the impact of self-congruity in the field of retailing is still scarce to this point. Thus, there is not much evidence yet, giving support to an effect of self-congruity on store evaluation. Still, the underlying motives that drive consumers to seek self-congruity in pre- and post-purchase evaluations can also be applied to the retailing setting.
To understand how actual self-congruity might affect the in-store experience, one must consider that individuals feel protective of their personal identity (actual self-concept).
Cognitive-24 consistency theories (Lecky, 1945; Heider, 1946; Festinger, 1957) argue that the actual self is driven by the need for self-consistency. Thus, individuals try to act in ways that are consistent with their own beliefs and identities (Kressmann et al., 2006). Staying true to one’s beliefs and identity by, for example, consuming brands, products, which have a personality that is congruent with one’s actual self-perception, should lead to the reinforcement of the own self-concept (Sirgy, 1982) and feelings of confidence (Hosany & Martin, 2012). Conversely, acting in ways inconsistent with a person’s own beliefs and values (i.e. consuming brands that have a personality incongruent with one’s actual self-perception), produces dissonance, a state of tension and discomfort (Sirgy, 1985). Similarly, self-verification theory (Swarm Jr, 1983) argues that individuals are motivated to confirm and sustain their actual self-concept. Thus, they look for experiences that reinforce their actual self, and avoid experiences that compromise their actual self-concept (Swann, Stein-Seroussi, & Giesler, 1992).
A few early studies have indicated a relationship between actual self-congruity and store preferences (Bellenger et al., 1976; Stern, Bush, & Hair, 1977). A more current study that further supports the previous argumentation comes from Willems, Swinnen, Janssens and Brengman (2011). The authors present findings that customers patronize stores that are congruent with their actual self, e.g. agreeable customers are found to patronize agreeable fashion stores, whereas open-minded and extraverted customers patronize innovative fashion stores.
Applied to the current research, this would mean that a customer who sees himself as being competent in sports, physically strong and enduring etc. (actual self-concept), should evaluate the in-store experience in a sport apparel store more positive, where the typical user of the store is perceived as being athletic and sporty (store image). By visiting this kind of store, the customer satisfies his need for consistency and thus reinforces his actual self-concept. In contrast, a customer should evaluate the in-store experience more negatively, when he sees himself lacking the physical
25 abilities and appearance of the typical store customer. Reason being that in this case the image of the store is not aligned with the actual self-concept the customer holds, which would lead the customer to perceive dissonances. Therefore, based on self-consistency and self-verification theory, I propose the following hypothesis:
H6a: The greater (lower) the actual self-congruity of a customer in a sport apparel store, the greater (lower) the evaluation of the store atmosphere.
To understand how ideal self-congruity might affect the in-store experience, one must consider that a person can also be driven by the need for self-esteem (Sirgy, 1985; Sirgy et al., 2000). This need leads individuals to find ways to enhance their own self-concept (Sedikides & Strube, 1997). Individuals have an ideal image of how they would like to be and behavior consistent with this ideal image (e.g. like consuming a brand that reflects that ideal image), helps them to enhance their self-concept. Kressmann et al. (2006) argues that this is because the behavior enables customers to decrease discrepancies between their actual and ideal self-perception. In other words, customers infer that the consumption of a product, brand etc. that is congruent with one’s ideal self-perception, should lead customers to get closer to their ideal self-concept (Sirgy, 1982; Sirgy & Su, 2000). Like in the previously mentioned example of Sirgy et al. (2000) about the working-class person that aims to become more upper-working-class, because he does not like his self-perception of being lower class, and thus patronizes an upscale department store to get closer to that ideal image. Support for relationship between ideal self-congruity and store evaluation comes from Ibrahim and Najjar (2008), who report that ideal self-congruity lead customers to form an initial positive attitude towards the store.
Based on the previous section, I argue that because of a person’s need for self-enhancement (Sirgy, 1985; Sirgy et al., 2000), the in-store experience of a sport apparel store should be evaluated more positively, when the image of the store helps the customer to approach his ideal self-concept.
26 A customer that perceives himself to be lacking sport competences etc., but does aspire to be athletic and sporty, should evaluate a sport apparel store more positively, when the typical store patron is perceived to be competent in different kind of sports, physically enduring and strong, as well as attractive. Reason being that patronizing such a store should lead customers to get closer to his aspired image. Therefore, I introduce the following hypothesis:
H6b: The greater (lower) the ideal self-congruity of a customer in a sport apparel store, the greater (lower) the evaluation of the store atmosphere.
2.4 How Body Weight affects Store Evaluation
So far, I presented hypotheses that state the proposed relationships between the individual variables, i.e. the effect of body weight on physical self-concept, intrinsic motivation on physical self-concept etc. However, the focus of the current study is to test a) if and b) to what extent the effect of body weight on the evaluation of a sport apparel store is mediated through the motivation to engage in exercise, the physical concept of the customer, and the construct of actual self-congruity.
More specifically, as stated in the introduction, I argue that overweight customers evaluate the in-store experience in a sport apparel store more negatively in comparison to non-overweight customers. The rationale for this statement is summarized in the following. Overweight customers hold a more negative physical self-concept in comparison to non-overweight customer due to the negative direct effect of body weight, and the negative indirect effect of body weight through the intrinsic motivation. Thus, with a lower physical self-concept, overweight customers will perceive less actual self-congruity in a sport apparel store, where the typical customer is being perceived as having a high physical self-concept (i.e. competent in sports, physically strong etc.). Based on individuals need for self-consistency, the store evaluation should be more negative as the discrepancies between the self-perception of the consumer and the image of the typical brand user
27 might lead to dissonances. Therefore, based on the previous literature review and hypotheses, I introduce the final two hypotheses:
H7: Body weight negatively influences a customer’s evaluation of the store atmosphere through physical self-concept and actual self-congruity.
H8: Body weight negatively influences a customer’s evaluation of the store atmosphere through intrinsic motivation, physical self-concept and actual self-congruity.
Another aim of the study is to examine the role of ideal self-congruity in the setting of a sport apparel store. As argued in the introduction, overweight status might not necessarily have to lead to a more negative store evaluation. It might be the case that the customer is not satisfied with the own physical self-concept, yet aspires to be more like the typical customer of a sport apparel store, i.e. competent in different kind of sports, physically enduring and so forth. As discussed previously, in this situation, patronizing a sport apparel store with such an image might lead the customer to get closer to his aspired image.
In this literature review, argumentation is given on why the importance of the physical self-concept should affect ideal self-congruity, as well as on why ideal self-congruity should have an impact on the in-store experience. Unfortunately, to my best knowledge, there is no research yet on how body weight might affect the importance of the physical self-concept. While the relationship between body weight and the importance of the physical self-concept is not hypothesized in this research, the current study still explores a model in which the importance of the physical self-concept and ideal self-congruity mediate the effect of body weight on the evaluation of a sport apparel store. As the aim of the current study is to examine differences in the store evaluation between overweight and non-overweight customers, the relation between body weight, importance of the physical self-concept, ideal self-congruity and store evaluation should be considered.
28 Figure 1 shows the research model which summarizes the proposed directs relationships. However, the research model does not portray the proposed indirect effects. The first model proposes that body weight influences store atmosphere indirectly through 1) physical self-concept and actual congruity, and 2) intrinsic motivation, physical concept, and actual self-congruity. The second model conceptualizes an indirect effect of body weight on store atmosphere through importance physical self-concept and ideal self-congruity.
Figure 1: Research Model
Note: Dashed lines are not hypothesized, but examined in the analysis
3. Pre-Study: Development of shortened Exercise Motivation scale
3.1 Aim of Pre-Study
The aim of the first study is to develop a shortened version for two dimensions of the modified Behavioral Regulation in Exercise Questionnaire-2 (Markland & Tobin, 2004), which in its original form consists of 19 items to measure motivation to engage in physical activity. After a discussion with the research team and the project supervisor, the decision was made to shorten the scale, because of concerns regarding the length of the final questionnaire.
The modified BREQ-2 scale is divided into five different behavioral regulation stages, namely amotivation, external regulation, introjected regulation, identified regulation, and intrinsic regulation. Even though not all stages are of interest for this study, the three to four items per stage for the dimensions of interest could have still caused problems in terms of the questionnaire length.
29 Especially since participants often do not have enough time to complete long inventories and thus too many items per variable might lead respondents to skip questions or respond randomly (Ninot, Fortes, & Delignières, 2006).
The reduction of items is based on the results of an exploratory factor analysis and a previous validation of the scale (Murcia, Gimeno, & Camacho, 2007). The validation of the scale is implemented at the beginning of the second study by conducting a confirmatory factor analysis. The results of the first study are presented in the following sections.
3.2 Data Collection, Sample and Procedure
For the data collection, a questionnaire was set up using the research platform Qualtrics. The survey was conducted using convenience sampling, which is a form of non-probability sampling (Field, 2013). The questionnaire was mainly distributed via social media channels and message boards. As an incentive for participation a €50 shopping voucher was raffled among the participants. In total, 105 valid responses were collected. Thus, the sample size of Study 1 fulfills the general recommendation for the minimum sample size (MacCallum, Widaman, Zhang, & Hong, 1999). The finale sample consisted of 60 female (57,1%) and 45 male (42,9%) respondents. The age of the respondents ranged between and 18 and 58 years, with a mean distribution of 25,7 years of age. The instrument used for the questionnaire was the modified BREQ-2, which is comprised of five subscales. The scale comprises intrinsic regulation (4 items, e.g. I exercise because it is fun), identified regulation (3 items, e.g. I value the benefits of exercise), introjected regulation (4 items, e.g. I feel guilty when I don’t exercise), external regulation (4 items, e.g. I exercise because people say I should), and amotivation (4 items, e.g. I can’t see why I should bother exercising). By using a five-point Likert Scale (1 = Very Untrue for me, 5 = Very true for me), respondents were to rate the 19 items of the BREQ-2 scale.
30 Participants conducting the questionnaire received the following instruction: “Why Do You
Engage in Exercise? Using the following scale, please indicate to what extent each of the following items is true for you. Note that there are no right or wrong answers and no trick questions. [I] simply want to know how you feel about exercise. Your responses will be held in confidence and used only for [my] research purposes.” (Tenenbaum & Eklund, 2012).
3.3 Data Analysis and Results
A maximum likelihood (ML) factor analysis was conducted on the 19 items using SPSS Statistics 24. This method was chosen, as the maximum likelihood approach has “a more formal statistical foundation than the principal factors methods and thus provides more capabilities for statistical inference […]” according to Fabrigar, Wegener, MacCallum and Strahan (1999). Oblique rotation (direct oblimin) was chosen as the form of rotation as it allows the factors to correlate (Field, 2013), which is the case with the BREQ-2 (Markland & Tobin, 2004). Based on the literature (Markland & Tobin, 2004; Murcia et al., 2007), a fixed number of five factors was selected for extraction. The Kaiser-Meyer-Olkin criterion for sampling adequacy had a value of = .80, which is “meritorious” according to Hutcheson & Sofroniou (1999), and verifies the adequacy of the sample for factor analysis. The significance of Bartlett’s Test of Sphericity (p<.001) gives further support for the sample adequacy as it indicates that the inter-item correlations are sufficiently large for factor analysis. Lastly, the correlation matrix was examined. Fields (2013) suggests that any value greater than 0.9 for the correlations coefficients indicates problems of multicollinearity. After scanning the correlation matrix, no signs of multicollinearity were detected.
In an initial analysis three items were shown to have strong cross-loadings (>.30 on multiple factors). These items were removed from the analysis. In the next step, the factor analysis was rerun and now showed a clean pattern matrix without any strong cross-loadings. Table 1 shows the
31 factor loadings after rotation. In contrast with previous theory, the analysis of the eigenvalues for each factor suggests a four-factor solution.
Table 1: Summary of exploratory factor analysis for BREQ-2
1 2 3 4 5 Factor
I think it is important to make the effort to exercise regularly.
Identified I value the benefits of exercise. .65
(I don’t see why I should have to exercise.) -.65
It’s important to me to exercise regularly. .56
(I can’t see why I should bother exercising.) -.51
I exercise because it is fun. -.97
Intrinsic I enjoy my exercise sessions. -.76
I find exercise a pleasurable activity. -.69
I take part in exercise because my friends/family/partner say I should.
External I exercise because other people say I should. .77
I feel under pressure from my friends or family to exercise.
I feel like ashamed when I miss an exercise session.
Introjected I feel like a failure when I haven’t exercised in
I feel guilty when I don’t exercise. .64
I don’t see the point in exercising. .61
Amotivation I think exercising is a waste of time. .45
Eigenvalues 5.51 2.27 1.75 1.30 .89
% of variance 34.43 14.16 10.94 8.12 5.56
Cumulative % of variance 34.43 48.59 59.53 67.65 73.21
Note: Factor loadings <.40 were removed from this table.
Four factors had eigenvalues over Kaiser’s criterion of 1. In total, these factors explained 67.65 percent of the variance, also supporting a four-factor solution. The scree plot showed ambiguous results, suggesting either a four or five factor solution. A closer inspection of the different factors shows that amotivation loads in part on the first factor (identified regulation), leading to this ambiguous results. Markland and Tobin (2004) noted that “a potential concern […] is that the identification and amotivation subscales could be assessing a single construct”. However, a further analysis conducted by those authors (2004) suggests that identified regulation and motivation assess different constructs. In general, Factor 2 (intrinsic regulation), factor 3 (external