In Need of
New Running Shoes -
The Impact of Gender Composition on
Shopping Motivation and Buying Behavior
Lara Galka
Student ID: 11145439
Final Version – 27.01.2017
MSc Business Administration – Marketing
University of Amsterdam
1
Statement of Originality
This document is written by student Lara Galka, 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.
2
Table of Content
Statement of Originality ... 1 Table of Content ... 2 List of Tables ... 4 List of Figures ... 4 Acknowledgements ... 6 Abstract ... 7 1. Introduction ... 8 2. Literature Review ... 112.1. Group Size in the Shopping Context ... 11
2.2. Utilitarian vs Hedonic Shopping Motivation ... 12
2.3. Gender’s Effect on Shopping Motivation and Buying Behavior ... 13
2.4. Social Situation in the Shopping Context ... 15
2.5. Evaluation of Touchpoint Elements ... 17
2.5.1. Process Elements ... 18
2.5.2. Employee-Customer Interaction Elements ... 19
2.5.3. Product-Interaction Elements ... 20
2.6. Buying Behavior Influenced by Shopping Motivation and Gender ... 21
3. Conceptual Model ... 23
3
4.1. Data Collection Procedure ... 25
4.2. Sample ... 26
4.3. Measure Development ... 27
4.4. Variable Coding ... 28
5. Results ... 31
5.1. Descriptive Data of the Sample ... 31
5.2. Variables and Measurements ... 34
5.2.1. Reliability of Scales ... 34
5.2.2. Skewness & Kurtosis ... 34
5.2.3. Correlation Matrix ... 35
5.3. Hypothesis Testing ... 39
6. Discussion and Conclusion ... 55
6.1. Summary of the Results ... 55
6.2. Discussion and Conclusion of the Results ... 59
6.3. Theoretical Implications ... 63
6.4. Managerial Implications ... 64
6.5. Limitations and Future Research ... 65
Bibliography ... 67
Appendix ... 71
I. Relevant Measurement Scales ... 71
II. Factorial ANOVAs (Testing H1a-H1c) ... 73
4
IV. Linear Regression Analyses (Testing H3, H4 and H5) ... 81
V. Linear Regression Analyses (Testing H3a, H4a and H5a) ... 84
VI. Linear Regression Analysis (Testing H6) ... 87
VII. Linear Regression Analysis (Testing H6a) ... 89
VIII. Questionnaire (English Version) ... 91
List of Tables
Table 1: Summarized Hypotheses ... 23Table 2: Descriptive Data of the Sample (N=740) ... 32
Table 3: Descriptive Data of Gender Composition (N=744) ... 32
Table 4: Descriptive Data of Social Situation (N=744) ... 33
Table 5: Descriptive Data of Shopping Motivation (N=367) ... 33
Table 6: Cronbach’s α Scores for Construct Scales ... 34
Table 7: Skewness and Kurtosis Scores... 35
Table 8: Correlation Matrix ... 38
Table 9: Correlation Between Social Situation and DVs ... 39
Table 10: Tests of Between-Subjects Effects (Factorial ANOVA 2x2) ... 41
Table 11: Tests of Between-Subjects Effects (Factorial ANOVA 2x3) ... 43
Table 12: Linear Regression Model: Shopping Motivation Influenced by Social Situation ... 46
Table 13: Linear Regression Model: Process-Element Evaluation ... 46
Table 14: Linear Regression Model: Employee-Customer Interaction Elements Evaluation ... 47
Table 15: Linear Regression Model: Product-Interaction Evaluation ... 47
5 Table 17: Linear Regression: Employee-Customer Interaction Elements Evaluation Based on
Gender of Companion ... 49
Table 18: Linear Regression: Product Interaction Elements Evaluation Based on Gender of Companion ... 50
Table 19: Linear Regression Model: Money Spent in Store ... 51
Table 20: Linear Regression Model: Number of Items Purchased ... 52
Table 21: Linear Regression: Money Spent in Store Based on Gender of Companion ... 54
Table 22: Linear Regression: Number of Items Purchased Based on Gender of Companion .... 54
Table 23: Summary of the Results Based on Hypotheses ... 55
List of Figures
Figure 1: Conceptual Model... 23Figure 2: Estimated Marginal Means of Shopping Motivation under the Assumption of an Interaction Effect of Gender with Companion vs No Companion (2x2 Model) ... 41
Figure 3: Estimated Marginal Means of Shopping Motivation under the Assumption of an Interaction Effect of Gender with Gender Companion (2x3 Model) ... 42
Figure 4: Average Shopping Motivation by Gender of the Focal Shopper ... 44
Figure 5: Average Shopping Motivation by Gender of Companion... 44
Figure 6: Number of Items Purchased by the Respondents (Survey Version 1)... 51
Figure 7: Number of Items Purchased by the Respondents (Whole Dataset) ... 53
6
Acknowledgements
This Master Thesis is the final assignment for my MSc Business Administration/Marketing studies at the University of Amsterdam. It explores the impact of the gender of a shopping companion on the focal shopper’s shopping behavior.
Completing my studies, including this thesis, would not have been possible without the help of a number of people.
First of all, I would like to thank the Asics team and especially Martin Block and Jelle van Arendonk for making the whole study possible and providing us with everything needed to conduct it in an appropriate manner.
Furthermore, a special thanks to my thesis supervisor Dr. Alfred Zerres, who critically reviewed the thesis process and provided some good feedback on how to improve.
Thank you, Nastie Schoenmakers, Emma Kraanen and Lukas Janßen, for your unconditional support in good and bad times during the whole project. Sharing our struggles and achievements really kept me going.
Finally, a big thank you to all the Asics customers, who were willing to invest time and effort into answering our survey.
Lara Galka
Amsterdam, January 27th 2017
7
Abstract
There is a growing interest of consumer researchers in the shopping behavior of customers and their companions. Due to limited research it is of importance to investigate which role companion shoppers play in the purchase setting and how they influence the focal shopper. This article focuses on the gender of shopping companions and tests if there is an effect of gender of shopping companion on the focal shopper’s shopping behavior. Shopping behavior was measured in shopping motivation, evaluation of touchpoint elements and buying behavior. The author conducted a cross-sectional study over a period of two weeks in cooperation with sports brand Asics. 744 surveys were filled out by respondents in four European stores. The results demonstrate that there is a positive influence of shopping companions on focal shoppers. However, the gender of the shopping companion does not have significant influence on the focal shopper’s shopping behavior. Males in this study were more hedonically motivated than women, which contradicts prior research. Furthermore, utilitarian shopping motivation leads to more items purchased and higher amounts spent in the underlying research. This finding contradicts findings of former researchers, who found hedonic shopping motivation to have similar positive effects. A question that arises is if the sports retail environment provides a different shopping experience than apparel retail and if this is a possible explanation for the differences.
8
1. Introduction
Mark Zuckerberg, Chairman and CEO of social media platform Facebook once said:
“People influence people. Nothing influences people more than a recommendation from a trusted friend.”
This quote proves true in many life situations. Shopping for example, is an inherently social experience and shopping alone vs shopping in group can tremendously affect how the shopping journey is perceived (Borges, Chebat, & Babin, 2010). Shopping companions have an influence on time spent in store, number of items purchased and even purchase satisfaction (Hart & Dale, 2014; Sommer, Wynes, & Brinkley, 1992; Woodside & Sims, 1976). Furthermore, they have an influence on the shopping motivation (hedonic vs utilitarian value) of the customer in focus (Babin, Darden, & Griffin, 1994; Borges et al., 2010). Especially peers as a shopping companion influence shopping behavior as compared to family members (Luo, 2005).
Furthermore, given that there are evolutionary differences between men and women, gender also affects the way how people shop: Men, when shopping in companion purchase more and spend longer time in the store (Hart & Dale, 2014) and also influence the decision-making more than women do (Ward, 2006). As author Ken Poirot puts it:
“Men most often know what they want, yet they are not always sure how they feel. Women most often know how they feel, yet they may not always know what they want.”
Might this be an indication for men having more utilitarian shopping motivation while women having more hedonic shopping motivation? This seems to be the case, as hedonic consumption was proven to be higher for women compared to men (Tifferet & Herstein, 2012).
One major gap in the literature however remains: in how far does the gender of the shopping companion(s) have an influence on shopping motivation and consequently on the evaluation of
9 different touchpoints in store and actual buying behavior? Will a female shopping companion encourage purchases more or less than a male companion? And does it play a role if the shopping companion is a good friend or your mother?
Gender is a factor that economics researchers should examine more carefully in the future (Cheng, Chuang, Wang, & Kuo, 2013). Most of these studies focus on the individual when comparing gender differences and shopping behavior. It is therefore important to “investigate the influence of the companion shopper in the social environment and the roles these individuals play in interpersonal relationships within the purchase setting” (Lindsey-Mullikin & Munger, 2011, p. 12). It has not been tested yet in how far one or more same-sex or other-sex shopping companion(s) influence shopping motivation and eventually the evaluation of touchpoints in store and the actual buying behavior. It is also of interest in how far this is moderated by social situation. One major aim of this study is to find out in how far female shopping companions have a different influence on touchpoint evaluation and buying behavior than male shopping companions have. The evolving research question is as followed:
“Is there an effect of the gender of shopping companion on shopping motivation, the evaluation of touchpoint elements and actual buying behavior?”
This study makes the following three contributions to marketing literature and practice: First, it will be tested if there is an influence of shopping motivation on evaluation of touchpoint elements and actual buying behavior, which is measured in money spent and number of items purchased. The current study was conducted in a sports retail environment and therefore provides a different scene than a normal fashion retailer. Second, the direct effect of gender of shopping companion on shopping motivation, the evaluation of the different touchpoints and actual buying behavior will be tested. Current research focuses on the effects of the focal shopper’s gender, but does not take the gender of shopping companions into consideration. Third, based on the
10 findings of this study, recommendations can be given on how to best target different gender groups in the shopping context.
To test my research question, four researchers including myself conducted a field study in four European retail stores of sports brand Asics. The on-location quantitative study aimed at finding out if and how the shopping motivation, touchpoint evaluation and buying behavior of customers is influenced by male as compared to female shopping companions.
11
2. Literature Review
2.1. Group Size in the Shopping Context
Most things in life are perceived differently when doing them alone as compared to doing them in a group. The same seems to apply in the shopping context. If a person is shopping alone or in a group, has a significant influence on the overall shopping experience and behavior. Shopping alone might be preferred over shopping in a group when the consumer identifies highly with the shopping environment. This is due to the fact that the shopper’s attention cannot be directed fully at the favorable place to shop anymore but is now partly directed at the shopping companion, which might lead to distraction and therefore a less pleasant shopping experience (Borges et al., 2010). However, there is much more evidence for the positive effect of being accompanied while shopping in current literature: A companion may enhance the shopping experience by facilitating shared experiences and also offers support and assistance in decision making (Borges et al., 2010). Some customers even consider the shared shopping experience as an opportunity to enhance personal friendships (Yu & Bastin, 2010). Furthermore, the urge to purchase is greater in the presence of companions (Prus, 1993; Woodside & Sims, 1976), especially in that of peers compared to family members (Luo, 2005). A shopping companion is defined as “a person who joins a focal shopper during a shopping trip” (Borges et al., 2010). This urge even increases when the peer group is cohesive and the individual is highly susceptible to influence of others (Luo, 2005). Additionally, Sommer et al. (1992) showed that based on Social Facilitation Theory, which states that the presence of other people influences behavior; groups spend more time in the store and purchase larger loads than lone individuals. This is supported by Woodside & Sims (1976), who found sales gains and increases in areas visited in the store when shopping companions were present during the store visit.
12 Hart & Dale (2014) added to this by showing that also purchase satisfaction and attitude toward the act were positively influenced by joint retail experiences. The retail experience might be regarded as more positively when shopping companions are present as they may socially satisfy the focal shopper (Hart & Dale, 2014).
2.2. Utilitarian vs Hedonic Shopping Motivation
Often researched in the shopping and retail context are two stable high-order dimensions, namely utilitarian and hedonic motivation (Babin et al., 1994; Chang & Fang, 2012; Olsen & Skallerud, 2011). Utilitarian consumer behavior is described as “ergic, task-related, and rational”, whereas hedonic behavior is rather subjective, personal, fun and playful (Babin et al., 1994). Hedonic shoppers enjoy the shopping act itself and browsing through the store while encountering the different touchpoints. Hedonic shopping motivation was shown to more positively influence overall satisfaction with the retailer than utilitarian motivation (Jones, Reynolds, & Arnold, 2006). Furthermore, hedonic shoppers do not see shopping akin to a negative sense of "work” that has to be done (Babin et al., 1994), but rather have a need for fun, novelty, escapism and social interaction during the shopping experience (Yu & Bastin, 2010). Additional value is gained through accessibility, and product value (Olsen & Skallerud, 2011). Hedonic shopping motivation therefore represents the emotional and entertainment worth of the shopping experience and reflects the symbiotic and / or synergistic relationship between the consumers and their shopping reference group(s) (Olsen & Skallerud, 2011; Yu & Bastin, 2010). Based on the differences between hedonic and utilitarian shoppers, it can be argued that shopping motivation influences the evaluation of touchpoint elements, which will be explained in detail in 2.5. Furthermore, as researched by Yu & Bastin (2010), hedonic needs motivate consumers in their impulse buying behavior. Utilitarian shoppers in contrast, are only satisfied if the shopping chore is completed successfully and as of same importance, if done in a fastidious manner (Babin et al., 1994). In the underlying study, shopping motivation will be measured as
13 one construct, consisting of a differential between utilitarian and hedonic shopping motivation. It is assumed that the higher the respondent scores on shopping motivation, the more hedonically motivated he or she is.
2.3. Gender’s Effect on Shopping Motivation and Buying Behavior
Not only group shopping itself, but also the gender of the focal shopper and that of shopping companions influences the shopping experience. Gender is one of the most common forms of segmentation used by researchers. It is highly affecting the perceived shopping experience of individuals as males and females are socialized differently. The different socialization plays an important role in the retail context (Noble, Griffith, & Adjei, 2006). As stated in evolutionary psychology, gender differences are rooted in genetic variations that arose millions of years ago through natural selection (Buss, 1995). In the shopping context for example, women have higher levels of brand commitment, hedonic consumption and impulsive buying as compared to men (Tifferet & Herstein, 2012) and enjoy shopping more as they consider it a recreational activity (Kruger & Byker, 2009). This can be explained as “over the millennia, a genetically-based tendency to enjoy behaviors relevant for gathering [e.g. hedonic consumption and impulse buying] would have been selected for in women” (Tifferet & Herstein, 2012, p. 179). Additionally, women are more prone to uniqueness, assortment seeking, social interaction, and browsing whereas men attach higher importance to information attainment and convenience, are less likely to browse and rather enter a store with the mission to buy (Noble et al., 2006; Underhill, 2009). Therefore, it can be argued that men engage more in utilitarian shopping behavior than women. “Males and females were also found to be significantly different with respect to affective process components” (Coley & Burgess, 2003, p. 282). Women tend to shop from an affective mindset and score higher on irresistible urge to buy, positive buying emotion and mood management, which is ‘engaging in impulse buying as a way to manage ones mood.’ Therefore it can be argued that touchpoints in store have a higher possible influence on female’s
14 (spontaneous) buying behavior than they do have on men. Furthermore, women more frequently make impulsive purchases in apparel-related product categories, whereas this is the case for men in technology and entertainment categories (Coley & Burgess, 2003). Differences were also shown for cognitive process components (Coley & Burgess, 2003). Women are more likely to engage in cognitive deliberation and unplanned buying.
But why should the gender of the shopping companion have an influence on the focal shopper and his or her buying behavior? The companion shopper is present during the shopping experience and provides the last input before a purchase takes place. This can significantly influence the actions of the focal shopper (Lindsey-Mullikin & Munger, 2011). Especially shopping with a companion of the opposite gender might affect the focal shopper’s buying behavior. It is shown that shoppers rather exhibit impulse purchase behavior when shopping with an opposite gender companion (Cheng et al., 2013). The influence of jointness of shopping experiences was found to have a stronger influence on men than on women. It had a positive influence for males on amount paid, attitude toward the act, and re-patronage intentions. Amount paid and attitude toward the act are positively related to hedonic shopping motivation, which is normally more descriptive for women. It can therefore be argued that males, when shopping in companion, score higher on hedonic shopping motivation as compared to shopping alone. This effect might even be stronger when they are accompanied by a female, as females are more likely to be hedonic shoppers in general. I therefore argue:
H1a: Males score higher on hedonic shopping motivation when being accompanied than when shopping alone.
H1b: Males score higher on hedonic shopping motivation when being accompanied by a female as compared to being accompanied by another male.
15 For females, this effect was reversed, indicating a negative influence of shopping companion (Hart & Dale, 2014). Gender effects in spouses’ joint decisions in the shopping context have been found by Ward (2006): “When men and women express strong preference intensities for product choices that differ from each other, men are more likely to ultimately gain the product choice they preferred in the joint decision than are women” (Ward, 2006, p. 122). This effect was proven across as well as within product category and might be an indication for a higher influence of male shopping behavior when a male and female shop together. As mentioned before, shopping companion has a negative influence for women (Hart & Dale, 2014). However, a female shopping companion might not have the same negative effect as a male companion as she is more likely to engage in hedonic shopping behavior herself. However, if a female is accompanied by a male, who scores higher on utilitarian shopping motivation than herself, the female’s hedonic motivation might be mitigated. Therefore I argue:
H1c: Females score lower on hedonic shopping motivation when accompanied by a male as compared to being accompanied by another female.
In the current study, it is important to take the effect of gender of companion into consideration. This is due to the assumption that a companion of the opposite gender might change the natural tendency of women to be hedonically motivated and of men to be utilitarian-motivated shoppers. Therefore, being accompanied by a male might make women less hedonic shoppers, whereas being accompanied by a female might make men more hedonic shoppers.
2.4. Social Situation in the Shopping Context
Not only the presence of shopping companions, but also the relation of the focal shopper to those, influences the shopping experience. Borges et al. (2010) found that shopping orientation varies with the type of shopping companion. Babin et al. (1994) made a link between the two shopping motivations and shopping group size and found that utilitarian shopping value was
16 slightly lower when shopping with a friend as compared to shopping alone or shopping with a family member. “Shoppers express significantly more affect and hedonic shopping value when shopping with a friend relative to shopping alone or with a family member” (Borges et al., 2010, p. 298). Shopping alone (or even with a family member) in contrast, can be seen as a utilitarian chore to get shopping done in a fast way whereas shopping with a friend is a more hedonic and playful experience and even leads to longer time spent in the store (Sommer et al., 1992). Family members make spontaneous purchases less likely (Borges et al., 2010) and shopping alone has a more positive influence on the attention on the task to be performed than shopping with companion (especially peers) (Baron, Moore, & Sanders, 1978). Shopping with family members can be seen as fostering a sense of responsibility and discourage wastefulness and extravagance (Borges et al., 2010). Shopping with a spouse in particular was also reported to have rather negative consequences on the shopping experience as compared to shopping with a peer. 88 percent of married couples report disagreeing with each other during the buying process (Spiro, 1983).
In contrast to that, peers positively influence hedonic shopping value and the urge to purchase, however, have the potential to reduce shopping effectiveness by making utilitarian cues less salient (Borges et al., 2010). They also encourage impulsive buying as peers tolerate and award spontaneity more than family members do (Luo, 2005). In order to test those effects in the current sports retail environment and replicate findings of researchers in the field, it is argued that the effects of gender composition as described above are moderated by the social situation in the shopping context:
H2: Social situation moderates the impact of gender composition on shopping motivation. Being accompanied by peers has a positive effect on hedonic shopping motivation as compared to being accompanied by family.
17
2.5. Evaluation of Touchpoint Elements
Customer touchpoints are “points of human, product, service, communication, spatial, and electronic interaction collectively constituting the interface between an enterprise and its customers over the course of customers’ experience cycles” (Dhebar, 2013). In other words, touchpoints include all contacts a customer makes with the company during his or her shopping journey (Stein & Ramaseshan, 2016). As there are numerous different touchpoints, a classification of those into groups provides a better overview. Some authors call these “service clues”, including the technical performance of the service (functional clues), the tangibles associated with a specific service (mechanic clues) and the behavior and appearance of service providers (humanic clues) (Berry, Wall, & Carbone, 2006). Stein and Ramaseshan (2016), dividing more precisely, identify seven distinct elements of customer experience touchpoints, namely atmospheric, technological, communicative, process, employee-customer interaction, customer-customer interaction and product interaction. In the underlying research, the focus will be on three of those: process-, employee-customer interaction- and product-interaction elements. The first two were selected as they include some sort of human interaction, which is of interest in a study measuring the impact of social influence. Product interaction was chosen as it was proven before that gender has an influence on the evaluation of assortment in store (Noble et al., 2006). First of all, the effect of shopping motivation on the evaluation of those three touchpoint elements will be tested. Additionally, it will be tested if there is a direct effect of the gender of the shopping companion on the evaluation of the three touchpoint elements. As the presence or absence of a shopping companion has high influence on purchase decisions and shopping behavior, and as there is limited research in that field, this paper aims at extending existing knowledge by testing the influence of shopping companion’s gender on the evaluation of those three touchpoint elements.
18
2.5.1. Process Elements
As defined by Stein and Ramaseshan (2016), process elements are “actions and steps the customers need to take to achieve a particular outcome with the retailer” (p. 12). Those include the waiting time and service process at the fitting rooms and cash-desk as well as the overall navigation through the store, in other words: How easy it is to locate merchandise, fitting rooms, and the cash-desk etc.? It can be argued that process elements are especially important for shoppers with utilitarian shopping motivation as they want to finish the shopping experience as quickly as possible and do not want to deal with any delays in the shopping process. Therefore, if they are slightly dissatisfied, they might be more likely to evaluate process elements negatively. On the other hand, hedonic shoppers might attach high importance to process elements as it makes a substantial part of their shopping experience. However, in case of a rather dissatisfying experience with process elements (e.g. long waiting times at the register or fitting room), hedonic shoppers might not be as affected as utilitarian shoppers. This is based on the assumption that hedonic shoppers do not have the urge to finish the shopping in a fastidious manner as compared to utilitarian shoppers (Babin et al., 1994). Therefore,
H3: Utilitarian shopping motivation negatively influences the evaluation of process elements as compared to hedonic shopping motivation.
Furthermore, as mentioned above, men attach higher importance to information attainment and convenience, are less likely to browse and rather enter a store with the mission to buy (Noble et al., 2006; Underhill, 2009). Therefore, if two men are shopping together, they might be driven by the aim to finish their shopping trip and would rate slight problems with process elements rather negatively. Furthermore, men have high influence on women when it comes to shopping behavior (Ward, 2006). Their goal-driven shopping behavior might rub off on the female they accompany. In that case, the female focal shopper might aim at finishing the shopping trip in a fastidious manner and would be more likely to rate any delays in process elements more
19 negatively than when shopping with another female who is not that goal-oriented. Therefore I argue that:
H3a: Being accompanied by a male negatively influences the evaluation of process elements as compared to being accompanied by a female.
2.5.2. Employee-Customer Interaction Elements
Employee-customer interaction elements are all direct and indirect interactions customers have with employees in the store (Stein & Ramaseshan, 2016). Those include in how far the customer considers the employee friendly, helpful and competent and the service personalized.
Shopping motivation was proven to influence social interaction during a shop visit. Personal interaction by store employees is likely to contribute to provide hedonic shopping value (Yu & Bastin, 2010) whereas it is rather negatively perceived by shoppers with utilitarian shopping motivation as they do not want to be distracted in their mission to purchase. Especially utilitarian shoppers who enter the store with the aim to purchase might be hesitant to engage in lengthy sales talks with the employees. They might seek for advice in order to find items in a fastidious manner. This however also poses the risk for less knowledgeable employees to receive lower ratings particularly from utilitarian shoppers who do not want their shopping trip to be delayed or unsuccessful. Hedonic shoppers on the other hand enjoy the full shopping trip as a kind of “recreation”, which also includes being in contact with store employees. Therefore I argue:
H4: Hedonic shopping motivation positively influences the evaluation of employee-customer interaction elements as compared to utilitarian shopping motivation.
Apart from shopping motivation, the gender of the shopping companion is hypothesized to influence the evaluation of employee-customer interaction elements. The presence of a shopping companion can have a substantial influence on how the employee-customer interaction is perceived. For example, the greater the perceived expertise attached to the
20 salesman, the greater the effect of the presence of shopping companions on customer purchasing behavior (Woodside & Sims, 1976). It might be argued that if a skilled salesman can convince two or more people who are shopping together to buy a product, doubts of costumers decrease and they become more likely to make a purchase. But even if the salesman is a non-expert, the presence of a shopping companion increases the likelihood to buy as compared to not being accompanied (Woodside & Sims, 1976). Furthermore, the gender of the shopper has an influence on the level of social interaction with the store employees. Females are more prone to social interaction while shopping (Noble et al., 2006). It can therefore be claimed that female shopping companions encourage social interaction with the sales staff and therefore, as based on Woodside & Sims’ (1976) findings increase the likelihood to purchase, unaffected by the expertise of the salesman. Additionally, a female shopping companion might increase the evaluation of the employee-customer interaction touchpoint as females are more social-interaction seeking than men in general. I therefore argue that:
H4a: Being accompanied by a female positively influences the evaluation of employee-customer interaction elements as compared to being accompanied by a male.
2.5.3. Product-Interaction Elements
Product-interaction elements are defined as “direct and indirect interactions customers have with the core tangible or intangible product offered by the retailer” (Stein & Ramaseshan, 2016, p. 12). Within this touchpoint, especially the product assortment is of interest, as it was shown by previous research that gender has an influence on assortment-seeking. Women are more prone to assortment-seeking (Noble et al., 2006) and expect a wider range of products when shopping. It is therefore likely that a female shopping companion will more negatively influence the evaluation of product-interaction elements when the assortment is narrow than a male shopping companion will. As shown by Yu & Bastin (2010), the shopper’s perceived importance of product assortment is positively related to utilitarian shopping whereas not to hedonic shopping. These
21 findings contradict with the assumption that women as hedonic shoppers are more prone to assortment-seeking. However, having utilitarian shopping motivation, customers visit the store in order to purchase something they are looking for. Finding a small assortment and having to visit another store to complete the shopping might be especially annoying for utilitarian shoppers, as they do not perceive shopping as a recreational activity. Therefore, two different effects based on shopping motivation and gender of shopping companion are hypothesized:
H5: Utilitarian shopping motivation negatively influences the evaluation of product-interaction elements as compared to hedonic shopping motivation.
H5a: Being accompanied by a female negatively influences the evaluation of product-interaction elements as compared to being accompanied by a male.
2.6. Buying Behavior Influenced by Shopping Motivation and Gender
For the effect of shopping motivation on buying behavior, two opposing things could be hypothesized: On one hand, it might be argued that utilitarian shopping motivation positively influences buying behavior (defined as money spent and number of items purchased) as utilitarian shoppers enter the store with the mission to buy. However, even if utilitarian motivation mostly leads to a purchase when the required product is available, hedonic shopping motivation might be a trigger for shoppers to buy more. Hedonic shoppers enjoy the shopping act itself and like browsing through the store while encountering the different touchpoints. This browsing behavior might lead to the discovery of items that would otherwise not have been purchased. Furthermore, impulsive buying was found to be positively correlated with hedonic shopping value (Tifferet & Herstein, 2012; Yu & Bastin, 2010). It can be argued that when being hedonically motivated, the customer would end up buying more than when looking for one (or more) specific item(s) only. Furthermore, the actual purchase itself can produce hedonic value and might even
22 serve as the climax of the shopping process (Babin et al., 1994). Therefore it can be argued that:
H6: Hedonic shopping motivation positively influences buying behavior (defined as money spent and number of item purchased) as compared to utilitarian shopping motivation.
Especially women are more likely to purchase impulsively. There are four main reasons for that as identified by Tifferet & Herstein (2012): First, impulse buying is related to hedonic shopping, which women are more likely to engage in. Second, women suffer more from negative emotions (e.g. anxiety, depression) than men. Those negative emotions were found to have a positive link with impulse buying. Third, consumers with a higher need to touch (women) are more susceptible to impulse buying as they have once physically encountered the product (Peck & Childers, 2006). Finally, based on nature, men are not equipped with behavioral traits, which lead to impulse buying, whereas women are. A female shopping companion might therefore encourage the focal shopper to buy things he or she might not have planned to buy in case of shopping alone or in company with a male. It might therefore be argued that:
H6a: Being accompanied by a female positively influences buying behavior (defined as money spent and number of item purchased) as compared to being accompanied by a male.
23
3. Conceptual Model
Figure 1 summarizes the theoretical framework and the hypotheses that have been derived.
Figure 1: Conceptual Model
Table 1: Summarized Hypotheses
H1a
H1b
H1c
Males score higher on hedonic shopping motivation when being accompanied than when shopping alone.
Males score higher on hedonic shopping motivation when being accompanied by a female as compared to being accompanied by another male.
Females score lower on hedonic shopping motivation when accompanied by a male as compared to being accompanied by another female.
H2 Social situation moderates the impact of gender composition on shopping motivation. Being accompanied by peers has a positive effect on hedonic shopping motivation as compared to being accompanied by family.
H3
H3a
Utilitarian shopping motivation negatively influences the evaluation of process elements as compared to hedonic shopping motivation.
Being accompanied by a male negatively influences the evaluation of process elements as compared to being accompanied by a female.
24 H4
H4a
Hedonic shopping motivation positively influences the evaluation of employee-customer interaction elements as compared to utilitarian shopping motivation.
Being accompanied by a female positively influences the evaluation of employee-customer interaction elements as compared to being accompanied by a male.
H5
H5a
Utilitarian shopping motivation negatively influences the evaluation of product-interaction elements as compared to hedonic shopping motivation.
Being accompanied by a female negatively influences the evaluation of product-interaction elements as compared to being accompanied by a male.
H6
H6a
Hedonic shopping motivation positively influences buying behavior (defined as money spent and number of item purchased) as compared to utilitarian shopping motivation. Being accompanied by a female positively influences buying behavior (defined as money spent and number of item purchased) as compared to being accompanied by a male.
25
4. Research Method
4.1. Data Collection Procedure
This quantitative research was conducted using a cross-sectional survey design. Together with three other researchers, data was collected in four Asics stores within Europe (Amsterdam, Barcelona, Brussels and London). The data collection in Barcelona took place 21.11.16 - 26.11.16 and in London 21.11.16 – 27.11.16. In Amsterdam (29.11.16 – 05.12.16) and Brussels (28.11.16 – 03.12.16) data was collected the week after. Differences in data collection days are based on the opening times in the four stores. A test run was conducted in the Amsterdam store one week prior to the first round of data collection. Based on this test run, the questionnaire was adapted according to customer feedback. The original survey was designed in English; in order to overcome language barriers, the survey was translated into Dutch, French, German and Spanish.
The survey was designed in Qualtrics, covering all customer experience topics relevant for Asics and the individual topics of the four researchers. This resulted in an extended list of questions, which was split in two in order to be user-friendly and not too time-consuming for the respondents in store. During data collection, one of the two survey-versions was randomly selected by Qualtrics for each respondent. SPSS was used for descriptive statistics and analyses.
The real-time survey on location provided a realistic shopping situation, which consumers were able to directly recall in order to answer the survey. In the beginning of each survey, a decision had to be made on including an observation or not. If an observation took place (3% of the cases), respondents were followed through the store in order to measure the times they spent on different touchpoints (e.g. employee contact, shoe wall, fitting room etc.). After finishing their shopping journey, they were approached and asked if they would like to participate in the
26 survey. The survey was handed to them on a tablet as it contained personal questions, which the respondents had to fill out unobserved. However, the researcher always stayed close to the respondent to clarify questions or for further explanation.
When approaching the customers, the researcher made clear that participating in the study would help him/her personally with graduating and that the survey was not designed by Asics. As an incentive, respondents directly received some candy and were informed that they could win a pair of running shoes when entering their e-mail address at the end of the survey.
In the survey version which is relevant for this study (Survey Version 1), the first questions aimed at identifying which touchpoints the customers noticed or used and how they rated them. In Brussels, those touchpoints included technological elements not available in the other stores (e.g. Shoe Selector, Community Feed etc.). Questions about the overall perceived store atmosphere, the shopping motivation, process-elements, employee-customer elements and product-interaction elements followed in that order. Furthermore, a brand-related question was included, followed by self-concept and body-concept questions, which were based on another researcher’s study. Finally, respondents were asked about their attitude towards Asics lifestyle products and their demographics.
4.2. Sample
Data was collected via random sampling as every customer in the Asics store had the same chance of being selected. In total 1,054 customers were approached during the two weeks of data collection, of which 744 (71%) agreed on participating in the survey. 372 (50%) of those were answering the survey version applicable for the underlying study (Survey Version 1).
27
4.3. Measure Development
In order to maximize the level of reliability and internal validity, measurement scales were borrowed from previous studies, of which some were adapted to the retail shopping context. Next to measuring the demographics of the focal shopper and the shopping companion, the social situation and observed buying behavior, there are four main measurement constructs of interest in the underlying study. Three of them are measured on a seven-point Likert scale anchored by “disagree strongly” (1), “neither disagree nor agree” (4) and “agree strongly” (7). One makes use of a semantic differential scale with seven points.
Scale items for shopping motivation are adapted from previous research discussing and testing hedonic (recreation) versus utilitarian (task-fulfillment) shopping motivation (Wagner & Rudolph, 2010). Wagner & Rudolph based their hedonic (Jacoby, Szybillo, & Berning, 1976; Kaltcheva, Weitz, Kaltcheva, & Weitz, 2006; Mano, 1999) and utilitarian (Kaltcheva et al., 2006; Roy & Tai, 2003; Tatzel, 1982) measures on prior research in the field. Cronbach’s alpha in the original scale was reported to be .90 for the items testing hedonic motivation and .94 for the items testing utilitarian motivation. An example of an item of this construct is: “I wanted to go shopping for its own sake” for hedonic motivation and “I was searching for a specific product” for utilitarian motivation. The adapted scale excluded two of the eight items and made use of a semantic-differential design. Cronbach’s alpha for the adapted 6-item scale was reported to be .76.
Scale items for process elements are adapted from previous research discussing shopping convenience (Seiders, Grewal, & Godfrey, 2005). Items were reduced from seven to five items and Cronbach’s alpha was reported to be .89 for transaction convenience and .84 for benefit convenience. One example of an item measuring process elements is: “It is easy to find the products I am looking for at Asics.” Cronbach’s alpha of the adapted 5-item scale was reported to be .82.
28 Employee-customer interaction elements are measured with the service quality inferences scale by Baker, Grewal & Parasuraman (1994). This scale is composed of five items, of which four were used in the underlying research. Amongst them is: “This store's employees would be willing to help customers.” An alpha of .84 was reported for the version of the scale used by the authors. Cronbach’s alpha in the underlying study was reported to be .82, which is coinciding with Baker et al. (1994).
Product-interaction elements are measured with the product assortment scale by Srinivasan, Anderson, and Ponnavolu (2002). The scale consists of four items and was adapted as the original context was the customers’ perception of a broad and deep product assortment in the e-retail environment. One of the items used in the current study is: “This store provides a "one-stop shop" for my shopping.” An alpha of .81 was reported for the scale used by the authors. Cronbach’s alpha in the underlying study was lower with .67, indicating that the scale might be a better fit for the e-retail environment as reported by Srinivasan et al. (2002).
For detailed construct items, see appendix I. For the full questionnaire (in English) see appendix VIII.
4.4. Variable Coding
In total, there are nine variables used in the analysis of the underlying research. Gender and
Gender Companion are independent variables. Social Situation is used as a moderator and Shopping Motivation acts as a mediator. Dependent variables are the evaluation of Process Elements, Employee-Customer Interaction Elements and Product-interaction Elements.
Furthermore, Buying Behavior acts as a dependent variable, which is separated into Amount
29 Gender received a categorical coding of 0 = male and 1 = female. To test H1a, it was only of interest to test if a shopping companion of any gender influences shopping motivation. Therefore, “Gender companion” was coded as 0 = shopping alone and 1 = companion male, companion female and companion mixed, leaving two groups for analysis. To test H1b and H1c, only those respondents, shopping in companion were included in the analysis. Therefore, a new variable was created, where companion male = 0, companion female = 1, companion mixed = 2.
In order to test H3a, H4a, H5a and H6a, a linear regression was performed. Therefore, the variable “shopping companion” was recoded into two dummy variables (Dummy_CompFemale) and (Dummy_CompMixed). “Companion male” was used as the base category.
The variable “Social Situation” was summarized in one new, hand-coded variable including the categories “family” = 0 and “peers” = 1 as only this differentiation was important for analysis. Answers “family” and “spouse” were recoded into “family”, whereas “boyfriend/girlfriend”, “friend”, “business partner” and “acquaintance” was recoded into “peers”. “Others” was excluded for analysis.
Shopping motivation as well as the three dependent variables process elements, employee-customer interaction elements and product-interaction elements were measured on a 7-point Likert scale including a semantic differential for shopping motivation. For all of those scales, the mean was taken, leaving each respondent with an individual average.
For shopping motivation, the higher the respondent scored on the semantic differential, the higher he or she scored on hedonic shopping motivation as 1 indicated total utilitarian shopping motivation whereas 7 indicated total hedonic shopping motivation.
For process elements, the higher the respondents scored on the 7-point scale, the more likely he or she was to be satisfied with the process elements in store. The same applies for the evaluation of employee-customer interaction elements and product-interaction elements.
30 The variable “amount spent” measures the amount the respondent spent in store. It was originally measured in a string in the dataset and therefore had to be re-coded by hand to create a variable with scale numbers. All amounts were rounded to full numbers.
Number of items purchased was measured ordinal, with 0 indicating no item purchased and 17 indicating more than 15 purchased items.
31
5. Results
This chapter provides descriptive information about the used variables in the underlying research. Furthermore, correlations between the variables are established and the hypotheses are tested.
5.1. Descriptive Data of the Sample
In total, 1054 customers were approached in the four Asics stores, of which 744 (71%) agreed on participating in the survey. As the survey was divided into two versions, 372 (50%) of those were answering the survey version applicable for the underlying study. In the total dataset 4 responses were excluded due to missing data. The distribution between men and women is 54.6% males and 44.2% female (1.2% did not want to specify). The average age is 53.7 years and most respondents reside in the four countries were the research took place (Spain 24.3%; Belgium 21.2%; Netherlands 16.2%; UK 10.8%). However, in the age category, three respondents indicated that they were one year old, which might have slightly biased the age distribution. Most respondents (77%) indicated that they had a higher educational background (Bachelor and higher). This might be explained with the positive relationship between higher education and health status (Sifferlin, 2012). Higher educated people pay on average more attention to their health, which results in increased sporting behavior. Therefore it can be argued that they visit sport stores more often than lower educated people. However, even if the level of education was predominantly high, the average income did not follow that pattern. 30% of the respondents did not want to disclose their income, 41% indicated an income lower than 50,000€, leaving 29% for incomes higher than 50,000€ per year.
The distribution of the 744 surveys answered in the four different research locations is as follows: Barcelona 28.2%, Amsterdam 25.4%, Brussels 23.8%, and London 22.6%. These numbers indicate quite equal distribution between the four cities.
32
Table 2: Descriptive Data of the Sample (N=740)
Frequency in % Gender Male Female Not disclosed 404 327 9 54.6 44.2 1.2
Age M = 53.7, SD = 13.5, Min 1, Max 79
< 30 30-40 41-50 > 50 34 87 158 461 4.6 11.8 21.4 62.3 Country of residence Spain Belgium Netherlands UK Others 180 157 119 80 204 24.3 21.2 16.2 10.8 27.5 Education Lower Higher Not disclosed 144 570 26 19.5 77 3.5 Survey Version 1 (applicable for this research)
2
372 372
50 50
Out of the 744 respondents, 56% were shopping with a companion, whereas 44% were not. Of those, shopping with a companion, 39.1% were shopping with a male companion, 47% with a female companion and 13.9% with both a male and a female.
Table 3: Descriptive Data of Gender Composition (N=744)
Frequency in % Gender Composition Alone
Shopping with Companion
Shopping Companion Male Shopping Companion Female
Shopping Companions Male and Female
327 417 163 196 58 44 56 39.1 47 13.9
When it comes to social situation, 324 respondents indicated that they were shopping alone, which implies that 3 respondents misunderstood the question. Multiple answers were allowed for this question as a companion could be e.g. “family” and “spouse” at the same time. Family and
33 spouse both belong to the group “family” as mentioned before. The other four options are grouped as “peers”.
Table 4: Descriptive Data of Social Situation (N=744)
Frequency in %
Social Situation Alone
Group Family Family Spouse Group Peers Boyfriend/Girlfriend Friend Business Partner Acquaintance Other 324 130 101 106 90 19 11 8 43.5 17.5 13.6 14.2 12.1 2.6 1.5 1.1
When shopping motivation was divided into two categories (utilitarian vs hedonic), it turned out that 55.3% were rather utilitarian shoppers (average values 1-3), whereas 35.7% were rather hedonic (average values 5-7). 9% were neither one nor the other (average value 4). This data was based on a dataset of 367 respondents (the 372 respondents who received Version 1 minus 5 missing values).
Table 5: Descriptive Data of Shopping Motivation (N=367)
Frequency in % Shopping Motivation Utilitarian
Hedonic None 203 131 33 55.3 35.7 9
34
5.2. Variables and Measurements
5.2.1. Reliability of Scales
The four main constructs mentioned in 4.3. were used in the underlying study: Shopping Motivation (SM), Process Elements (PE), Employee-Customer Interaction Elements (ECIE) and Product-Interaction Elements (PIE). One counter-indicative item in the product-interaction elements construct was recoded. All constructs except product-interaction elements showed sufficient reliability in the first instance (Cronbach’s α > .70). Within employee-customer interaction elements the item “Employees of this store are not too busy to respond to customers' requests promptly” was removed, increasing the reliability from .82 to .91. Within product-interaction elements, the originally counter-indicative item “The choice of products in this store is limited” was removed, leading to an increase in Cronbach’s α from .67 to .81. The resulting strong Cronbach’s α scores for all constructs allowed for the items in the constructs to be averaged. For further details see Table 6 below. Differences in N are due to non-forced answers in the survey, which were excluded in the analysis.
Table 6: Cronbach’s α Scores for Construct Scales
Cronbach’s α N of items M SD SM (N=327) .76 3 (semantic differential) 3.56 1.78 PE (N=248) .82 5 5.91 .96 ECIE (N=337) .91 3 6.26 .95 PIE (N=368) .81 3 5.06 1.24 SM = shopping motivation PE = process elements
ECIE = employee-customer interaction elements PIE = product-interaction elements
5.2.2. Skewness & Kurtosis
To test for skewness and kurtosis, the averaged scales of the four constructs were used. None of the constructs is normally distributed. Especially process elements and employee-customer interaction elements are negatively skewed, indicating asymmetrical distribution with a long tail
35 to the left. Most values in those constructs are centered to the right of the distribution graph which implies that process elements and employee-customer interaction elements are especially positively evaluated by the respondents. Either, respondents did really perceive those elements as outstanding or they were biased clicking to the right of the scale in order to speed up the answering process. In that case however, product-interaction elements should also be more negatively skewed as this construct was placed later in the questionnaire. For further analysis, it is assumed that respondents were especially satisfied with process elements and employee-customer interaction elements, as indicated by the negative skewness. Those two constructs also have a strong leptokurtic distribution, as kurtosis values are > 2. However, according to Tabachnick & Fidell (1996), skewness will not make a substantive difference in the analysis, when the sample is large. The same applies for kurtosis. As the current sample includes (>) 364 respondents, the risk of distorted results is minimized.
Table 7: Skewness and Kurtosis Scores
Skewness Std. Error Kurtosis Std. Error
SM (N=367)* .250 .127 -1.072 .254
PE (N=365)* -1.231 .128 2.789 .255
ECIE (N=364)* -2.110 .128 5.997 .255
PIE (N=368)* -.630 .127 .073 .254
* = cases in which some items were missing, were averaged based on the data of available items
SM = shopping motivation PE = process elements
ECIE = employee-customer interaction elements PIE = product-interaction elements
5.2.3. Correlation Matrix
Table 8 provides an overview of the correlations between the variables. The table includes independent variables (gender, gender companion), the mediating variable (shopping motivation), dependent variables (process elements, employee-customer interaction elements, product-interaction elements) and control variables (age and education level). In cases of
non-36 normal distribution, a Spearman Correlation is usually used, as this test does not make any assumptions about the distribution. However, as both a Pearson and Spearman test were conducted on the current dataset and no major differences were revealed, Pearson Correlation was reported in the Correlation Matrix in Table 8. The analysis presents several significant and relevant (r = higher or close to .3) correlations between independent, moderating and dependent variables, which are described in the following.
The positive relation between gender and gender companion male suggests that when the gender category changes from 0 to 1 (0 = male, 1 = female), the variable gender companion male increases (r = .320 p < .01). At the same time, gender has a negative relation with gender companion female (r = -.317 p < .01). This suggests that female shoppers in the current study are more likely to be accompanied by a male than a female while shopping. However, this relationship is only weak.
The strong negative correlation between gender companion male and gender companion female (r = -.755 p < .01) makes sense, as the values of the two variables can only increase at the same time in case of “gender companion mixed”.
Different than expected, shopping motivation only has significant correlations with the control variables age and educational level. When shopping motivation increases (shoppers score higher on hedonic shopping value), age is also more likely to increase or the other way around. In other words, older shoppers seem to score higher on hedonic shopping value than younger shoppers. However, this relationship is very weak (r = .157 p <.01). Furthermore, there is a weak negative relation between shopping motivation and educational level (r = -.162 p <.01), meaning that if respondents had higher levels of education, they were more likely to be utilitarian shoppers.
37 The variable “process elements” is positively correlated with the two other dependent variables employee-customer interaction elements and product-interaction elements. If a respondent’s evaluation of process elements increases, he or she is also more like to rate employee-customer interaction elements higher. This relationship is moderate (r = .538 p < .01). Evaluation of process elements is moderately correlated with product-interaction elements (r = .394 p <.01). There is also a positive correlation between the evaluation of product-interaction elements and employee-customer interaction elements (r = .347 p < .01). These findings indicate that if customers are more satisfied with one touchpoint in the store, they are also more likely to perceive and rate the others accordingly. Especially the correlation between process elements and employee-customer interaction elements stands out, which makes sense as employees are involved in many of the process element processes (e.g. giving product advice, assisting customers during check-out and fitting). But also the relation between process elements and product-interaction elements is positive, indicating that positive experiences while searching for products and finding a wide selection of those is translating to a positive evaluation of process elements or vice versa. The positive relation between product-interaction elements and employee-customer interaction elements might be an indicator for the positive effects of employees when trying to get an overview over a large assortment.
Furthermore, there is a very weak negative correlation between employee-customer interaction elements and age (r = -.109 p < .05). This finding suggests that older shoppers are less likely to rate employee-customer interaction elements high than younger shoppers are. An explanation could be that older shoppers have higher expectations when it comes to customer service.
Finally, educational level has very weak negative correlations with product-interaction elements (r = -.182 p < .01) and age (-.133 p < .05). This indicates that higher educated customers rate product-interaction elements lower. As mentioned above, higher educated customers were more likely to be utilitarian shoppers. In that case, they are more likely to enter a shop with a mission
38 to buy. If the product they are looking for is not available and they cannot finish their shopping trip in the store, they might be more likely to rate product-interaction elements lower, which is in accordance with H5. Furthermore, the weak negative correlation with age indicates that older customers are on average less educated than younger customers. There are current trends that young educated people pay special attention to their health. Being young and well educated is therefore a main driver for sporting, which motivates people to visit a sports shop in the first place. This could be a reason for that effect.
Table 8: Correlation Matrix
SM = shopping motivation PE = process elements
ECIE = employee-customer interaction elements PIE = product-interaction elements
The correlation effects of type of shopping companion on the four dependent variables (shopping motivation, process elements, employee-customer interaction elements, product-interaction elements) are separately summarized in Table 9.
There is a significant negative relation between Shopping Alone and shopping motivation (r = -.109 p < .05) as well as between Shopping Alone and product-interaction elements (r = -.119 p <
39 .05). This indicates that people who are shopping alone are rather utilitarian shoppers, which is also supported by literature. Furthermore, lone shoppers also have the tendency to rate all touchpoints lower than the other seven groups. However, only the negative correlation with product-interaction elements is significant. These two findings, when taken together, are in accordance with H5, which hypothesizes that utilitarian shopping motivation negatively influences the evaluation of product-interaction elements. Another significant finding is the weak positive relation between Family and product-interaction elements (r = .120 p < .05). Shopping with family members seems to increase the evaluation of product-interaction elements, which is against assumptions, as shopping with family members was hypothesized to increase utilitarian shopping motivation. All other relations to companions (except “Other – employee-customer interaction elements” (r = -.103 p < .05)) do not have significant correlations with shopping motivation or the evaluation of touchpoints.
Table 9: Correlation Between Social Situation and DVs
5.3. Hypothesis Testing
In the following, H1-H6 including sub-hypotheses are tested (see Figure 1). H1a-H1c hypothesize the effect of companions and gender composition on shopping motivation. In H2, the variable social situation is added to the model. H3-H6 then test the effect of shopping motivation on the three dependent variables and actual shopping behavior. Finally, H3a-H6a test the effect of the gender of a shopping companion on the evaluation of the different touchpoints and actual buying behavior.
40 H1a – H1c
To test H1a, all data, where respondents were answering Survey Version 1, was included for analysis (372 respondents). Of those, 363 respondents did not have any missing data. A factorial ANOVA using a 2 (male vs female) x 2 (alone vs companion) design was conducted to test H1a. Shopping motivation was measured with a semantic differential, where higher values indicate more hedonic shopping motivation. There was a significant effect of a shopping companion compared to no companion on shopping motivation, F (1, 359) = 4.013, p < .05, η² = .011. This effect was supported by a significant simple contrast test, testing the difference between level 1 (shopping alone) vs level 2 (being accompanied). However, the effect of gender of the focal shopper on shopping motivation was not significant F (1, 359) = 1.599, p = .207, η² = .004. Even if there was no significant effect, there is a contrast to former research, where women scored higher on hedonic shopping motivation. In the current research, males (M = 3.37, SD = 1.89) scored slightly higher than females on hedonic shopping motivation when there was no effect of a shopping companion. In other words, they scored higher when they were shopping alone, as compared to females shopping alone (M = 3.08, SD = 1.89). The interaction effect of gender and no companion vs companion on shopping motivation was non-significant, F (3, 355) = .024, p = .877, η² = .000.
41
Table 10: Tests of Between-Subjects Effects (Factorial ANOVA 2x2)
Source Type III Sum of Squares
df Mean
Square
F Sig. Partial Eta Squared Corrected Model 18.866a 3 6.289 1.735 .159 .014 Intercept 4076.475 1 4076.475 1124.804 .000 .758 Gender 5.795 1 5.795 1.599 .207 .004 Alone vs Companion 14.545 1 14.545 4.013 .046* .011 Gender * Alone vs Companion .087 1 .087 .024 .877 .000 Error 1301.075 359 3.624 Total 5686.556 363 Corrected Total 1319.941 362 a
. R Squared = .014 (Adjusted R Squared = .006) *. Statistical significance: p < .05
As seen in Figure 2, there is a significant difference in shopping motivation between shopping alone vs shopping in companion. Even if there was no significant difference between men and women, there was a significant difference between shopping alone as compared to shopping in company, F (1, 359) = 4.013, p = .046, η² = .011. For both men and women, hedonic shopping motivation was significantly higher when shopping in companion. In H1a it was hypothesized that males score higher on
hedonic shopping motivation when shopping in company as compared to shopping alone. No
Figure 2: Estimated Marginal Means of Shopping Motivation under the Assumption of an Interaction Effect of Gender with Companion vs No Companion (2x2 Model)
