Consumers between supermarket shelves: The influence of inter-personal distance on consumer behavior

Journal of Retailing and Consumer Services 26 (2015) 104–114

Contents lists available at ScienceDirect

Journal of Retailing and Consumer Services journal homepage: www.elsevier.com/locate/jretconser

Consumers between supermarket shelves: The influence of inter-personal distance on consumer behavior Michael Luck n, Martin Benkenstein 1 Institute of Marketing and Services Research, University of Rostock, Faculty of Business and Social Sciences, Ulmenstrasse 69, 18051 Rostock, Germany

art ic l e i nf o

a b s t r a c t

Article history: Received 6 February 2015 Received in revised form 1 June 2015 Accepted 7 June 2015

Recent research underlines the important influence on consumer behavior of the mere presence of consumers on other consumers in retail businesses. While these results mainly relate to overall behavior, it is largely unknown how consumers behave in stores and how they react to customers present around them. This study examines the impact of an encounter between two consumers, directly in front of the same shelf row, on diverse consumer behavior. By means of two pilot studies we gained initial insights into the mutual influence. An image-based scenario study shows that the proximal distance is important for consumers' emotional state. In addition, the results indicate that these triggered negative and positive emotions mediate the effects of the distance on shopping satisfaction, aversive behavioral tendencies, willingness to buy, and number of alternatives considered. & 2015 Elsevier Ltd. All rights reserved.

Keywords: Mere presence Distance Consumer-to-consumer encounter In-store behavior

1. Introduction Up to 80% of all retail buying decisions are made directly at the point of sale (POS) and in front of the shelf (Pornpitakpan and Han, 2013). Research shows that consumers react – consciously or unconsciously – to the smallest changes at the POS and are influenced by the attributes of a consumption area (Baker et al., 1994). Hence, analysis of consumer behavior at the POS has been an important field of research for a number of years. From the early 1970s, research initially focused on the influence of physical surroundings, such as light, color, design, or music (Turley and Milliman, 2000). Besides physical attributes in retail business, shopping is usually an activity that takes place with other customers “who are in the service facility simultaneously with – and who are unacquainted with – a focal customer” (Brocato et al., p. 385), whereby customer-to-customer contact – primarily non-verbal – is unavoidable (Nicholls, 2010). This being the case, it is surprising that this important factor has largely been neglected in academic research (Brocato et al., 2012; Söderlund, 2011; Tombs and McCollKennedy, 2013). It is commonly known that the presence of other consumers in retail businesses can affect the perception of consumption (Grove and Fisk, 1997; McGrath and Otnes, 1995), because other n

Corresponding author. Fax: þ49 381 498 4378. E-mail addresses: [email protected] (M. Luck), [email protected] (M. Benkenstein). 1 Fax: þ49 0381 498 4378. http://dx.doi.org/10.1016/j.jretconser.2015.06.002 0969-6989/& 2015 Elsevier Ltd. All rights reserved.

customers can significantly influence their behavior (Söderlund, 2011; Tombs and McColl-Kennedy, 2003). Previous research shows that the influence of other consumers on buying behavior is multifaceted. According to Baker et al. (1994), this influence can be determined mainly by three dimensions. A limited number of studies have addressed the appearance (Argo et al., 2008; Brack and Benkenstein, 2012, 2014; Dahl et al., 2012), the behavior (Argo et al., 2006; Grove and Fisk, 1997; Martin, 2012; Söderlund, 2011), and the number of other consumers (Argo et al., 2005; Eroglu and Harrell, 1986; Hui and Bateson, 1991; Pan and Siemens, 2011; Uhrich, 2011; Uhrich and Luck, 2012) as a consequence for buying behavior. The results of these studies relate primarily to the overall shopping experience in a store. In contrast, there is a lack of research regarding specific consumer behavior within stores between the shelves where consumers look for products and where they ideally make a purchase decision (Hui et al., 2009; Sorensen, 2003). A few studies have shown that the mere presence of another consumer in the immediate vicinity can affect the emotions of fellow customers in a positive (Argo et al., 2005) or negative (Dahl et al., 2001) way in certain circumstances. Recent research suggests that the mere presence of other consumers in special areas of a shop increases interest in an offer but at the same time reduces the inclination to enter these areas (Hui et al., 2009). The fact that customers are not rigid objects and that they constantly move around in a store, means that between shelves is the place where they steadily encounter each other. To date, relatively little is known about which behavior is caused by these encounters. Initial insights in this research area are provided by the field

M. Luck, M. Benkenstein / Journal of Retailing and Consumer Services 26 (2015) 104–114

experiments of Martin (2012). This research examines how consumers respond to inappropriate behavior in the form of an inadvertent touch by other customers. However, little is known about how consumers react and behave to the mere presence of another customer in front of the same shelf. The emphasis of our contribution to this research gap is to investigate the influence of other consumers between the shelves to broaden the knowledge of this academic research field. Specifically, the goal is to examine which purchase-related behaviors will be influenced by the mere presence of the first other consumer in the same shelf row. In contrast to previous studies, we do not investigate the influence of the behavior of other customers. Instead, we focus on the consequences for the consumer's own behavior of encountering another consumer next to him or her. It has been shown in research on crowding that consumer density, which results from the number of other customers, is the most important component of crowding, exerting an influence on consumers (Michon et al., 2005). However, our study goes a step beyond crowding research as we examine the presence of only one other consumer and focus on the behavior of individuals who respond to the mere presence of the first other consumer in front of a shelf, without interacting with this other customer. Due to the availability of only a few and rudimentary research findings in this field, we conducted two consecutive exploratory pilot studies as a first step toward achieving this goal. In a second step, the resulting insights were embedded in a theoretical framework to shape our research hypotheses, which were then tested empirically. Our outcomes broaden knowledge of social presence in a commercial context. Our results underline the relevance of encounters of consumers between the shelves for consumer behavior. They also show that management should not underestimate and should pay more attention to the presence of other customers. These research findings provide the first substantial insights into the effects of social interaction in front of the shelf and reveal a plethora of other research opportunities. Even though this is only an initial step in this research field, a few management implications can be derived.

2. Pilot studies Due to the lack of research concerning social presence in the consumption situation directly in front of the shelf, this research initially combines two exploratory pilot studies. In our first pilot study, the behavior of consumers in situations with the social presence of another customer was observed using video-based observation. In a second study, we used the method of “shopping with consumers” (Lowrey et al., 2005) to gain a greater understanding of the behavior exhibited in the presence of other consumers, especially in front of the shelf. 2.1. Pilot study 1: Video-based observation In our first pilot study, we observed situations with and without the social presence of another consumer in a real consumption area. To exclude the influence of the researcher, who also represents a social presence, we employed video-based observation as a non-reactive process of inquiry. This method allows the accurate and objective capture of modes of behavior (Dodd et al., 1998). Video-based observation is particularly suitable for collecting real data on the subject area (Belk and Kozinets, 2005). Using video recordings of shelf rows on different weekdays, we were able to observe and capture the behavior of consumers in front of the shelf. Standardized observation reports were employed so that the observable behavior of consumers could be used for later analysis.

105

The reports followed ethical policies. The anonymity of consumers was ensured through pixelating the footage. The focus of the data collection was only on exhibited and observable behavior (Belk and Kozinets, 2005). Within this pilot study, we solely gathered data on the duration of stay in front of the shelf in the particular row, as well as the removal and consideration of purchase alternatives from the consumers' perspective. The gathered data were only accessible to the researchers and were destroyed after the data analysis (Carrigan and Kirkup, 2000). As a key criterion for the applicability of a recorded situation, the consumer had to come to a stop within the row and select at least one product. Thus, we ensured that consumers who just walked by were excluded from the analysis. A further condition was the social presence of one other consumer throughout the whole period, with no other consumers walking by. Based on this latter condition, we collected 207 situations that could be evaluated. In these 207 cases, approximately 20% made a choice in the presence of another consumer. Due to the lack of existing knowledge in this research area, we were unable to make a priori assumptions concerning how consumers would react in a situation with other consumers. Therefore, at the beginning of our pilot study, we assumed an indefinite coherence between the behavior exhibited and the presence of other consumers, and simply expected a difference in the parameters. Using ANOVA, we examined the differences in means for the groups (social presence vs. no social presence) for the variables “duration of stay in the row” and “number of alternatives considered.” The duration of stay in the row is measured in seconds and is the time from entering to leaving the aisle. The number of product alternatives considered refers to the physical removal of the products from the shelf before a decision is made. The groups differed significantly in relation to these two variables. Consumers in the group with no social presence stayed significantly longer in the row concerned (Mno social presence ¼46 s) compared to those consumers who made their choice while other consumers were present (Msocial presence ¼29 s; F(1,204) ¼8.089, p o.01). Before consumers chose a certain product, they considered significantly fewer product alternatives with other consumers around (Msocial presence ¼1.15) than when no one was present in the same row (Mno social presence ¼1.39; F(1,205) ¼ 4.461, p o.05). In keeping with the assumption of Hui et al. (2009) that areas with other consumers present are increasingly avoided, the results of our first pilot study show a negative impact caused by the presence of other consumers in the same row. This finding is very interesting but it still cannot be explained and requires further investigation. 2.2. Pilot study 2: Shopping with consumers To deepen our research, we used the method of shopping with consumers at the POS. To gain further insights, we engaged a convenience sample of nine students and employees. These participants suited the study purpose well as no special knowledge was required (the purchase context involving everyday products) and furthermore because they represent usual consumers from the general public. Table 1 gives an overview of these participants' demographic data (Mage ¼26.11 (SD ¼3.65); male¼ 56%). The research method of shopping with consumers connects the observation and questioning of consumers in real consumption settings (Lowrey et al., 2005). It also presents an excellent opportunity to capture consumer behavior and to let the participants directly describe and explain their actual behavior (Zaltman, 2003). Using this procedure, it is possible to obtain more comprehensive information in terms of the thoughts and feelings experienced by consumers within these particular situations than can be gained from other methods (McGrath and Otnes, 1995).

106

M. Luck, M. Benkenstein / Journal of Retailing and Consumer Services 26 (2015) 104–114

Table 1 Demographic characteristics of participants. Informant pseudonym

Gender

Age

Profession

Christin David Emil Harry Isabell Kelvin Lisa Nina Paul

F M M M F M F F M

25 22 28 27 34 28 24 24 23

Employee Student Employee Student Employee Employee Student Employee Employee

Furthermore, the triangulation afforded by this procedure offers a solution to the methodological weaknesses of pure observation of behavior in our first pilot study. For the purpose of our study, we accompanied our participants when they were shopping in different supermarkets on different weekdays. Our participants were invited individually to an appointment on a shopping street. At the beginning, they were told that they would take part in a study on thoughts, experiences, and behaviors while shopping. Subsequently, the participants were asked to do their shopping. To minimize the influence of the presence of the researcher on the participants' behavior, they were accompanied at the maximum distance. Hence, the natural behavior of participants could be observed. When participants had finished their shopping, we conducted semi-structured interviews concerning their consumption experience. The accompanied shopping trips were conducted in a short time frame as consumers are generally only aware of their behavior to a limited extent and are not able to reflect on it correctly or without rationalization after a short period (Zaltman, 2003). The average duration of accompanied shopping was 7.15 min. This procedure ensured the minimum loss of information. By obtaining a description of the situation as experienced by the participants, we were able to collect relevant information (Silberer and Wang, 2010). When the participants referred to the social presence of other consumers, they were asked to describe their feelings, thoughts, and behaviors in that situation in more detail. Due to the specific objective of the research and the fact that many behaviors are subconscious, the point of theoretical saturation was reached relatively quickly, that is no new insights were generated. In this case, theoretical saturation was achieved after nine accompanied shopping trips and the data collection was thus complete. Upon approval from the participants, the interview recordings were transcribed in their entirety and were made accessible for further analysis. Using an inductive procedure, we conducted qualitative content analysis to identify relevant behaviors caused by the presence of other consumers. This analysis aimed to categorize the data and was undertaken according to the guidelines of McCracken (1988). This structured procedure allowed us to identify typical examples of general behavioral patterns with regard to our research question (Spiggle, 1994). The results show that customers are bothered by the presence of others where they want to make their choice of product. Evidently, the mere presence of another consumer is able to evoke negative emotions. For instance, Isabell said: “If they are not always where I am, I do not care about them. But if they – let’s say – directly look at the same place as me or want to do that, they annoy me.” Another important issue based on the interviews with our participants is the reaction to other consumers who are close to the desired products in the same row. Participants initially avoid a situation in which there is another consumer in front of the shelf and delay their choice to a later moment. For example, Kelvin

responded as follows: “If they are in my way? Most times I wait until they have gone away. Or, as there are two ends of a shelf, I first look at the other end, hoping that later I can take a look at the part of the shelf I initially wanted to look at.” Nevertheless, participants reported that if they really wanted to buy a product, they would get close to it while other consumers were still present. For instance, Harry told us that he would first look for something different, but if he were seriously interested in buying a certain product, he would move closer to another consumer. Participants reported different behaviors that are highly relevant to businesses. For example, Christin said she usually hastens her choice in the presence of another person: “I would do the things I have to do, but maybe I would hurry a bit more than usual, I would not leave, I would still take the things […] I want. Nevertheless, the presence of another consumer in the same row may lead to some products in certain areas not being considered. This behavioral reaction is particularly important for businesses as only perceptible and clearly visible products are bought by consumers. Kelvin stated: “I just passed by, I did not look at the shelf she was standing in front of at that moment.” To summarize, the qualitative content analysis revealed four different behavioral reactions that are negative for the retailer and can be triggered by the presence of another person in front of a shelf at the POS:

   

experiencing negative emotions delaying the purchase hurrying when choosing a product ignoring the offered goods/non-consumption

2.3. Possible explanation of the behaviors identified The results of our two pilot studies underline that the presence of other customers is highly relevant to buying behavior. In summary, we observed that the presence of another person in front of a shelf evoked predominantly aversive behavioral tendencies in the customer. The behavioral reactions indicated correlate with behaviors that are usually caused by a large number of other consumers in a shop and occur in perceived crowding. However, it is a surprise that in our case only one other consumer is enough to trigger these aforementioned behavioral changes. A possible approach to explaining these unexpected patterns is the fact that exhibited behavior can be influenced significantly, not only by the available space within the whole retail outlet, but also by the interpersonal distance between two people (Worchel and Teddlie, 1976). Typically, consumers are in the same consumption area with a few other consumers. As consumers move, the distances between them change. If consumers are in one place and are looking for the same or a similar product, they inevitably come very close to each other. Therefore, interpersonal distance can be highly relevant in the retail context. For this reason, the theory of personal space is a good explanatory approach to the behavioral reactions we found during our pilot studies. In the following section, we review the literature on the theory of personal space and draw on the theoretical findings and the results from our pilot studies to develop hypotheses related to our retail context. Subsequently, these are tested empirically in our main study.

M. Luck, M. Benkenstein / Journal of Retailing and Consumer Services 26 (2015) 104–114

107

3. Main study

front of the shelf.

3.1. Theoretical basis and development of hypotheses

3.1.2. The mediating role of emotions on consumer behavioral intentions In a supermarket there are many different components of customer behavior critical to the performance of a retail company, such as intention to leave the situation, willingness to buy, inclusion of information, and satisfaction with the consumption situation. Crucial for behavior, and not only in consumption situations, are perceived emotions (Bagozzi et al., 1999). Mehrabian and Russell (1974) have developed an approach to the explanation of behavior. In their basic assumptions of the stimulus–organism– response (S–O–R) model, the emotions triggered play an essential role as a mediator in any situation for the behavior exhibited by a human being. Donovan and Rossiter (1982) have successfully transferred the model to the consumption context. Therefore, in what follows, we also assume that changed emotions resulting from a violation of personal space mediate consumers' behavioral tendencies. As mentioned in Section 3.1.1, people feel uncomfortable and show aversive behaviors when the distance between them and another person is too small (Felipe and Sommer, 1966). Furthermore, extreme proximity to another person and a violation of personal space leads to an increasing need to avoid such situations or to leave them as quickly as possible (Hayduk, 1983). Studies indicate that the duration of stay decreases in such a situation (Becker, 1973; Middlemist et al., 1976). In general, these reactions can be characterized as flight behavior or as aversive behavioral reactions that increase the distance to the other person (Altman, 1975; Hayduk, 1983). We transfer these insights from the social sciences and assume that a distance between customers in front of a shelf that is too small causes a higher expression of aversive behavioral tendencies. Thus, our hypothesis states: H2. A consumer will expresses more aversive behavioral tendencies when another consumer is very close to a desired product in front of the shelf. This relationship is mediated by shoppers' emotions. A company is interested in customers receiving and processing as much information from the consumer environment as possible. This information includes the perception of products, pricing information, and other promotional activities. Studies related to the presence of other consumers have shown dependencies between the number of other consumers present and the amount of information received and processed (Eroglu and Harrell, 1986; Harrell and Hutt, 1976). It can be assumed that consumers who experience greater pleasure have a higher intention to explore the consumer area than customers with a less pleasurable experience (Ridgway et al., 1990). Following the logic of the theory of personal space, we assume that a violation of personal space by another consumer in the same row leads to a decrease in the consideration of product alternatives, and we therefore phrase our hypothesis as follows: H3. A consumer will consider fewer product alternatives when another consumer is very close to a desired product in front of the shelf. This relationship is mediated by shoppers' emotions. Observational studies in a café chosen as the consumption context show that on arriving, customers avoid positioning themselves in close proximity to other customers already present as long as there are unoccupied tables. In such a situation they prefer the greatest possible distance from other guests (Becker, 1973; Tombs and McColl-Kennedy, 2010). Can a certain distance not be achieved, for example because the tables are too small, people avoid sitting down and leave without buying anything (Manzo, 2005). Accordingly, it can be assumed that an awkward situation arises from a violation of personal space and it can be posited that a negative effect on the intention to buy will also occur in a retail environment. Hence, we hypothesize as follows:

3.1.1. Proximity to other unknown consumers A theory from the social sciences that is concerned with different distances between people is the theory of personal space. In the 1960s, this was significantly shaped by the scientific investigations of Hall (1966) and Sommer (1969). During that period, several studies were conducted that concentrated on the behavioral consequences of different distances between two people in sociofugal situations. Sociofugal situations are situations in which people primarily avoid social interaction. Examples of such spaces are libraries, railway stations, and other waiting areas (Osmond, 1957), where there seems to be an unwritten rule that social isolation is supported and accepted by all individuals present. Here, social interaction is not encouraged and contact between strangers is considered unexpected and unwelcome (Sommer, 1967). The majority of studies on the effect of distance between two individuals have been conducted in these places. Observations show that extreme proximity to another person in the public arena leads to a violation of personal space. Subsequent studies have consistently shown that extreme proximity causes predominantly negative emotions and aversive behavioral tendencies (Altman, 1975; Felipe and Sommer, 1966). Grounded in these investigations, Hall (1966) defines personal space as “a small protective sphere or bubble that an organism maintains between itself and others” (p. 119). Personal space can also be defined as a zone that an unknown subject – for instance another consumer – cannot enter without a reaction by the organism (Sommer, 1969). It can be assumed that most behaviors in the context of distance from another person occur unconsciously and proceed subconsciously. Although studies of distance have received considerable attention in social psychology, the impact on consumer behavior in the commercial context has hitherto been neglected. A retail business can be considered a sociofugal place, where customers do not necessarily expect contact with other people. Actions or interactions among acquaintances occur within the closest acceptable distance of four to seven feet (Hall, 1966). If there is more than one consumer in the same aisle, many actions occur within that special distance, potentially closer depending on the size of the aisle. This means that parts of a consumer's personal space are often affected by another consumer in very close proximity and who is usually a stranger. According to the theory of personal space, behavioral adaptations by consumers are expected in this context. To our knowledge, the theory of personal space has been translated to the commercial context only through one empirical study (Argo et al., 2005). The shortest distance recorded was approximately two feet from another consumer; in this case, there was no real violation of personal space and a positive effect on the consumers' emotions could be noticed. However, research findings in social psychology show that when people who do not know each other come very close within public areas, they react with increased negative feelings, which can generally be described as discomfort (Altman, 1975). Thus, it can be assumed that in a consumption situation also, especially in a row with another consumer present, the feeling of discomfort (negative emotions) is stronger when the distance is too close. Conversely, it can be assumed that comfort (positive emotions) is affected by a violation of personal space in a negative way. If there is no violation of personal space in the row, the discomfort experienced will be less strong. Hence, we hypothesize as follows: H1. A consumer will experience more negative and less positive emotions when another consumer is very close to a desired product in

108

M. Luck, M. Benkenstein / Journal of Retailing and Consumer Services 26 (2015) 104–114

H4. A consumer will be less willing to buy a product when another consumer is very close to a desired product in front of the shelf. This relationship is mediated by shoppers’ emotions. A consumer's consumption satisfaction is a central determinant of future behavior (Donovan and Rossiter, 1982; He et al., 2012). If one person comes closer to another in a social situation, this behavior is taken as a violation of norms and expectations (Grove and Fisk, 1997; Sommer, 1969). This perception leads to an increasing dissatisfaction with the shopping experience as a whole (Brocato et al., 2012). Empirical studies show that satisfaction is affected positively by triggered positive emotions and negatively by triggered negative emotions (Mano and Oliver, 1993). We therefore assume that purchase satisfaction with a specific shopping situation is influenced negatively by a perceived violation of personal space, and we hypothesize as follows: H5. A consumer will experience lower satisfaction with the consumption experience when another consumer is very close to a desired product in front of the shelf. This relationship is mediated by shoppers' emotions. 3.2. Research design, method and participants To test our hypotheses, we used image-supported scenarios of a consumption context in a one-factor between-subjects design (violation of personal space vs. no violation of personal space). The participants were 120 undergraduates at a large public university (Mage ¼22.54 (SD ¼2.64); male¼68%). 3.2.1. Objective of study and development of the stimulus material The different areas of personal space cannot be considered a rigid and fixed construct as they are constantly adapted in relation to environmental conditions (Hayduk, 1983). Depending on the situation, culture, gender, and personality characteristics, these areas can differ in width. Moreover, analyses show that small interpersonal distances will be perceived as less adverse if people know each other (Rustemli, 1992). Based on these conditions, in our study we focus on a retail setting in which consumers are in front of a shelf and do not know the other customers present. To manipulate social presence, several studies have shown the advantages of using image material to supplement text scenarios as stimulus material in laboratory experiments (Argo and Main, 2008; Uhrich, 2011). On this basis, we used image-supported scenarios to simulate the presence of other customers. The adapted cartoon method is based on the study of He et al. (2012) in which social presence was successfully manipulated with simple comic pictures. With the use of cartoons, the potential influences of the appearance and the behavior of the other consumer can be kept neutral and constant, which is necessary here as the mere presence of another consumer and the violation of personal space are central to our study. We prepared our stimulus material with a cartoon camera to capture a consumption situation that is as real as possible. Every image showed the same section of a row of shelving in a supermarket. This approach enabled us to control potential disrupting variables related to the shop's surroundings in all the situations shown. To simulate proximity to a consumer, the image of another consumer was added to the initial cartoon. We repeated this process several times to develop a total of 29 cartoons showing different distances between the two consumers and the relevant part of the shelf with the intended product. To ensure that one of these scenarios was seen as a violation of personal space, eight participants were asked in a pretest to categorize the images as distant or close. All images that could not clearly be assigned by all participants were excluded from further procedures. Subsequently, five marketing academics familiar with the research topic evaluated the remaining 15 images and chose one image representing

very close proximity as well as one representing distant social presence. As the focus of the study was not on analyzing the size of personal space but on the influence of violation in a consumption context, we chose the closest distance without touching the other person because people feel particularly uncomfortable at a distance of less than one foot from another person (Hayduk, 1983). The pictures used in the study are presented in Appendix Fig. A. 3.2.2. Procedure and measures The participants were randomly assigned to one of two groups: “violation of personal space” and “no violation of personal space.” At the beginning, we told participants they would take part in a survey on thoughts, experiences, and behaviors while shopping in common retail businesses, and that they would start with a written scenario before receiving images of the situation described. They were asked to put themselves in that particular situation as much as possible. The written scenario described a common consumption situation in a supermarket in which participants spontaneously want to buy snacks and are not familiar with the offer. They were asked to go through the supermarket and find the right shelf using the signs. First, participants saw the shelf with the other consumer from the side. To avoid other interpretations of the situation, participants were instructed to enter the row immediately to make their choice of snacks. To illustrate the distance from the other customer, participants then saw one version of the stimulus material with a frontal view of the shelf and another consumer, close or distant. Afterwards, the participants were asked to complete a questionnaire. Discomfort and comfort were measured by positive and negative emotions. Bagozzi et al. (1999) suggest that subcategories of emotions have to be captured using at least three items. Therefore, we adopted items from the PANAS scales and measured positive and negative emotions (Watson et al., 1988) on a seven-point scale (1 ¼not at all, 7 ¼very much). To measure the participants' aversive behavioral intentions, we adopted three items from Mehrabian and Russell (1974), measured on a seven-point scale from 1¼ not at all to 7 ¼very much. The consumers' willingness to buy was adapted from Wood and Scheer (1996) and was covered with a single item (1 ¼very unlikely, 7 ¼very likely). As an important determinant, satisfaction with the consumption experience is of considerable importance. This variable was measured using items of the scale developed by Eroglu and Machleit (1990) adapted to our specific situation. Because satisfaction in the situation in front of the shelf was pivotal, inappropriate items were omitted. The two remaining items “I would enjoy shopping in this row” and “I am satisfied with the shopping experience in this situation” were measured on a seven-point scale (1 ¼strongly disagree, 7¼ strongly agree). To determine the consumers' informationseeking behavior, the number of alternatives considered was measured with a single item (“In this situation, how many snacks would you look at before making a decision?”). This item was also measured on a seven-point scale (1 ¼ none, 7 ¼very many). To check the manipulation of our independent variable, participants were asked to recall whether there were other people in the scenario depicted and to recall the number of other consumers present. As the aim of our study was to analyze the effects of a violation of personal space, this manipulation also had to be verified. To determine the incipient violation of personal space, participants were asked to mark in a picture how close other consumers could get without making them feel uncomfortable (Altman, 1975; Hayduk, 1983; see Appendix Fig. B). Afterwards, participants gave information on demographic variables and were asked to make a guess concerning the research purpose. Finally, we thanked them for their participation.

M. Luck, M. Benkenstein / Journal of Retailing and Consumer Services 26 (2015) 104–114

3.3. Results 3.3.1. Check of manipulations At the beginning of our analysis we used ANOVA to test whether the presence of the other customer was perceived in the same way in both situations. The number of other consumers did not differ significantly between the distant and close groups (F (1,112) ¼2.279, p 4.10; Mclose ¼.90, Mfar ¼1.02). The manipulation of social presence was successful. Further analyses showed that 90% of the participants stated correctly that they were not alone in that situation. To analyze the influence of social presence, individuals have to be aware of the presence of the other person. Therefore, those participants who did not respond correctly were excluded from further analysis (Dahl et al., 2001; He et al., 2012). In a second step, we tested the violation of personal space with a 2 (“violation” vs. “no violation”)  2 (perceived “violation” vs. “no violation”) chi-square test (χ² (2)¼62.235, p o.001). The manipulation of the violation of personal space was successful. Further analyses showed that in 19% of the cases of a violation of personal space and in 2% of the no violation cases the manipulation was not successful. Hence, these cases were excluded from further analysis. Other participants (n ¼ 7) were excluded as they identified the research purpose correctly. The resulting 83 valid cases were included in the following analyses. 3.3.2. Data analysis To examine our hypotheses, we used Smart PLS 2.0 software (Ringle et al., 2005), which applies partial least squares (PLS) analysis. This method was chosen for its capacity to examine interdependence between a single categorical independent variable and more than one dependent variable. Using PLS, it is possible to analyze a small sample within a more complex model and therefore hypothesis testing can be done with a high level of statistical power (Hair et al., 2012; Iacobucci, 2010). The minimum sampling size for models with reflective variables is determined by using the rule of thumb, that the minimum size is 10 times the largest number of paths leading to endogenous latent variables (Barclay et al., 1995). At 83 participants, our sample is greater than the minimum and exceeds the recommended number of 50 (Iacobucci, 2010). Therefore the use of PLS is appropriate and is highly consistent with our research objectives. 3.3.2.1 Measurement model. Prior to analyzing the data, we evaluated the reliability and the validity of the measurement model. The reliability of the measurement model is determined by using loadings of individual items and should reach at least 0.7. Furthermore, each construct's composite reliability should also exceed a value of 0.7. Another criterion for judging the model is the use of average variance extracted (AVE), which has to achieve a value of 0.5 to demonstrate sufficient convergent validity. Table 2 provides an overview of these quality criteria. In addition, the constructs used must be sufficiently different from each other. Using discriminant analysis in conjunction with the Fornell– Larcker criterion, the latent variables must share more variance with their indicators than with other variables (Henseler et al., 2009). The result of the discriminant analysis is shown in Table 3. In summary, all indicators used and latent variables satisfy the aforementioned criteria. 3.3.2.2 Structural model. To test our hypotheses, the nonparametric resampling bootstrapping procedure (5000 resamples/95% confidence interval) is used for the statistical evaluation of the path coefficient (Henseler et al., 2009; Preacher and Hayes, 2008). To analyze hypothesis H1, we consider paths a1 and a2. The results in Table 4 confirm our hypothesis. A violation of personal space in front of a shelf results in an increase in negative and a decrease in

109

positive emotions. To analyze our hypothesized multiple mediations, we follow the recommendations of Preacher and Hayes (2008). The central condition is a significant indirect effect (a  b) to indicate mediation. These effects are summarized in Table 5. To specify the mediation, the direct effects of the independent variables on the dependent variables are calculated without the mediators at the beginning of the analysis. Then, the same effects are calculated with the respective mediation. This approach is based on the examination of mediators by Baron and Kenny (1986). Accordingly, the conditions for the detection of the mediator are sufficiently fulfilled as soon as a previously significant effect of the independent variable on the dependent variable is eliminated or reduced by the integration of the significant paths via the mediator. In this sense, the mediator explains how external physical conditions affect the dependent variable (Baron and Kenny, 1986). Fig. 1a shows the total effects of a violation of personal space in a consumer setting on satisfaction with the shopping situation (c1), willingness to buy (c3), number of product alternatives considered (c4), and aversive behavioral tendencies (c2). As seen in Fig. 1b, significant effects between the independent variables and the dependent variables are displayed when the mediators are taken into account. The method of analysis applied generated a corresponding t-value for determining the statistical significance of the path coefficients by the bootstrapping procedure. According to the t-value with the corresponding degrees of freedom, all direct effects (cj′) are no longer significant. The results in Table 5 show that the effects of emotions on the dependent variables (bj) are significant. By calculating the product of the coefficient of the positive emotions (a1) and negative emotions (a2) with the relevant coefficient of the dependent variables (bj), the size of the indirect effect is calculated (a1bj þ a2bj). Using the calculation of bootstrap confidence intervals, it is possible to evaluate the mediation chains in terms of their significance. A consistent picture emerges for all dependent variables. Both emotions act as a mediator across all central dependent variables and reduce or increase the corresponding behavioral tendencies. Therefore, the results support hypotheses H2–H5. These findings mean that a violation of personal space effectively leads to a significant change in consumer behavior due to perceived decreases in positive emotions and increases in negative emotions. After accounting for the mediators, the direct effects are reduced in size and significance. The direct effect of violation of personal space on avoidance (c2′¼.08; t ¼.95; p ¼.17), willingness to buy (c3′¼  .02; t¼.23; p ¼.41), and number of product alternatives considered (c4′¼  .15; t ¼1.36; p ¼.09) are no longer significant. Thus, these behavioral intentions are totally mediated by emotions. As shown by the results in Table 5 for satisfaction, according to the results of the t-statistic, the relationship is no longer significant (c1′¼ .14, t¼ 1.65, p 4 .05). In contrast, the corresponding bootstrap interval indicates a significant effect. Preacher and Hayes (2008) characterize bootstrapping as the most powerful method; hence we assume a significant association. For this reason, the relationship between a violation of personal space and satisfaction is partially mediated through emotions (a1b1 þa2b2). Analyzing the covariates shows no effects for gender and age in that relationship.

4. Discussion 4.1. Summary The intense competition in the retail sector makes it necessary to identify all factors that influence the consumer in any way.

110

M. Luck, M. Benkenstein / Journal of Retailing and Consumer Services 26 (2015) 104–114

Table 2 Measurement model. Construct/item

Loading Composite reliability (CR) Average variance extracted (AVE)

Negative emotions Afraid Nervous Ashamed Hostile Positive emotions Interested Excited Enthusiastic Avoidance How much would you try to leave or get out of this situation? How much would you try to avoid any looking around or exploration in this situation? How much is this a situation in which you might try to avoid other people, avoid having to talk to them? Satisfaction with the consumption experience I would enjoy shopping in this row I am satisfied with the shopping experience in this situation Willingness to buy How likely is it that you would buy something in this situation? Number of product alternatives In this situation, how many snacks would you look at before making a decision?

0.89

0.66

0.86

0.68

0.86

0.68

0.83

0.71

n.a

n.a

n.a

n.a

0.86 0.90 0.75 0.73 0.83 0.85 0.78 0.90 0.80 0.78

0.77 0.91

n.a.: not applicable.

Table 3 Discrimant validity.

(1) (2) (3) (4) (5) (6)

Avoidance Negative emotions Positive emotions Number of product alternatives Satisfaction Willingness to buy

(1)

(2)

(3)

(4)

(5)

(6)

0.83 0.70  0.29  0.38  0.28  0.51

0.81  0.20  0.26  0.43  0.47

0.82 0.29 0.52 0.32

n.a. 0.25 0.26

0.84 0.22

n.a.

n.a.: not applicable. Table 4 Effects of violation of personal space on positive and negative emotions.

Positive emotions Negative emotions n

Coefficient

t-Value (bootstrap)

p

 0.23 0.24

2.25 2.97

0.014n 0.002nn

po .05. p o.01 (based on t(4999), one-tailed test).

nn

Academics have as yet paid little attention to the interpersonal influence of other consumers in front of the shelf. The aim of our study was to address this research gap and to investigate the effects of an encounter between two unknown customers on consumer behavior. In the first stage of our research, we conducted two exploratory pilot studies and identified behaviors relevant to retailers that are caused by the presence of other consumers. Subsequently, we stated our hypotheses on the basis of these results and adopted the theory of personal space. We empirically tested our hypotheses with an image-supported scenario study. The results highlight the importance of the first other customer between the shelves for following customers. Our study is the first to transfer the theory of personal space successfully to the consumption context. In accordance with the theory of personal space, our results show that customers who are too close to each other in the same row – but not in front of the same products – experience increased discomfort and can therefore change their behavior, which is relevant to the retail industry. A guarantee of the success of a retail company is to have satisfied

customers who linger for as long as possible, look around extensively, and finally buy something. However, the outcomes of this study demonstrate that encounters between the shelves, especially violations of personal space, will tend to be avoided and consumers will want to leave such situations very quickly. As shown in Fig. 1a, a violation of personal space can initially induce a negative influence on purchase-relevant behaviors (c1–c4). This behavior can be interpreted as an evasive reaction or coping. Taking into consideration emotions as mediators (see Table 5), these direct effects are reduced (c1′–c4′). In addition, the results show a substantial increase in the explained variance with the inclusion of mediators (Fig. 1b). Moreover it can be shown that a violation of personal space affects positive emotions and negative emotions with the same strength and in the opposite direction. Interestingly, the outcomes show that these negative and positive feelings affect behavioral tendencies to varying degrees. On closer examination, negative emotions mediate stronger willingness to buy and aversive behavioral tendencies. On the other hand, positive emotions mediate satisfaction and number of product alternatives considered in a stronger manner. This effect is much less pronounced than for negative emotions, but a stronger direct effect persists. Furthermore, we observe a greater effect from the multiple mediations of positive and negative emotions on the latter behavioral tendencies. This suggests that despite the significant reduction of well-being, a violation of personal space has a great impact on satisfaction and the consideration of various product alternatives. As can be seen in the analysis of mediation, satisfaction is not fully mediated by emotions. It can be assumed that other mediators besides emotions are responsible for changes in satisfaction in such a consumer situation. Researchers can take this result as an avenue for further investigation. Previous studies that have focused on the social presence of a few other consumers in a consumption situation (Argo et al., 2005; He et al., 2012) are extended by our research in that a consumption situation is significantly influenced by proximity to another consumer. Combined with the results of research on the positive effects of a moderate number of other consumers, our results indicate that the distance between those present is a determinant of the consumer behavior exhibited at the POS. Thus, the mere social presence of another consumer does not necessarily have a positive effect as soon as the distance is reduced beyond a critical limit. Our

M. Luck, M. Benkenstein / Journal of Retailing and Consumer Services 26 (2015) 104–114

111

Table 5 Results for total, direct and indirect effects of personal space violation in a consumer setting. Dependent variable

Coefficient

t-Value (bootstrap)

p (one-tailed)

Satisfaction (c1) R²¼ 0.40/Q²¼ 0.24 (c1′) a1b1 a2b2 a1b1 þa2b2 Avoidance (c2) R²¼ 0.52/Q²¼ 0.33 (c2′) a1b3 a2b4 a1b3 þa2b4 Willingness to buy (c3) R²¼ 0.27/Q² ¼0.23 (c3′) a1b5 a2b6 a1b5 þa2b6 Number of product alternatives (c4) R²¼ 0.14/Q²¼ 0.06 (c4′) a1b7 a2b8 a1b7 þa2b8

 0.32

3.53

n

 0.14  0.10  0.07  0.17 0.27

1.65

n.s.

0.051

2.73

nn

0.004

0.08 0.03 0.16 0.19  0.18

0.95

n.s

0.173

1.72

nnn

0.044

0.23

n.s.

0.409

2.29

nnn

0.012

1.36

n.s.

0.088

Percentile 95% confidence interval Lower

 0.02  0.05  0.10  0.15  0.24  0.15  0.05  0.04  0.09

Upper

o 0.000  0.2858  0.1750  0.1549  0.2762

 0.0036  0.0266  0.0089  0.0724

sig sig sig sig

 0.0584 0.0003 0.0568 0.0883

0.2102 0.0830 0.2508 0.2807

n.s sig sig sig

 0.1852  0.1161  0.1818  0.2520

0.1417  0.0028  0.0287  0.0587

n.s sig sig sig

 0.3219  0.1223  0.0957  0.1695

0.0347  0.0011  0.0062  0.0344

n.s sig sig sig

n.s. not significant (based on t(4999), one-tailed test). n

po .001. p o.01. p o .05.

nn

nnn

Satisfaction

R²=0.10

c1=-0.32***

c2=0.27**

Avoidance

R²=0.08

Willingness to buy

R²=0.03

Violation of personal space c3=-0.18*

c4=-0.24*

Number of product alternatives

R²=0.06

*p<0.05; **p<0.01; ***p<0.001 (based on t(4999), one-tailed test)

a1=-0.23*

Positive emotions (interest)

b1=0.42***

Satisfaction b3=-0.14*

Avoidance

R²=0.51

Willingness to buy

R²=0.27

b4=0.65***

Violation of personal space

b5=0.22* b6=-0.42***

a2=0.24**

R²=0.40

b2=-0.31**

Negative emotions (nervousness)

b7=0.22* b8=-0.18*

Number of product alternatives

*p<0.05; **p<0.01; ***p<0.001; (based on t(4999), one-tailed test)

Fig. 1. (a) Model with total effects. (b) Model with multiple mediation design.

R²=0.14

112

M. Luck, M. Benkenstein / Journal of Retailing and Consumer Services 26 (2015) 104–114

Violation of Personal Space

No Violation of Personal Space

Fig. A1. Simulation of social presence and violation of personal space.

Fig. B1. Determining incipient violation of participants' personal space.

research work further extends insights into the effect of an unintended touch between two consumers (Martin, 2012), and reveals that not only an accidental touch by other customers, but simply a violation of personal space has a negative influence. Overall, we can show that a violation of personal space is enough to exert a negative effect on feelings and undesirable behavioral tendencies for retailers. 4.2. Limitations and further research Our paper is another contribution to the investigation of social influence at the POS and gives a number of leads for future research. As in other research, our work has some limitations, detailed here, which can be used as a starting point for future research. The first limitation is caused by the use of the image-supported scenario in our main study. Due to the advantages of this method, the results are high in internal validity but associated therewith they are lower in external validity. Future research can validate our results by conducting experimental studies in a real consumption setting. As in many other studies, our second limitation relates to the choice of students as participants. Further research can demonstrate the generalizability of our results by using participants from other populations. The third limitation relates to the fact that our empirical studies were conducted in a Western context.

Previous research shows that socially accepted interpersonal distances between people are perceived and evaluated differently by members of different cultures (Rustemli, 1992). For instance, Britons prefer greater distances from others than the French or South Africans. In contrast, Chinese citizens have adapted to proximity due to their large population (Hall, 1966). Another limitation is the restriction of our focus to a certain product category. We chose snacks because the influence of social presence is considered neither to be obvious nor inconceivable with this product category (Belk, 1974). Relevance and familiarity with a certain product may exert a moderating effect in this context. Future research should add potential moderators to verify our results and above all to offer assured recommendations for other assortments. In situations in which service personnel have to restock shelves during opening hours, critical interpersonal distances from consumers in a row are unavoidable. In our research, we solely considered the presence of other consumers and we can demonstrate that a violation of personal space in terms of spatial proximity significantly affects the behavior exhibited. Whether these effects are also evoked by the mere presence of service staff will also be of particular importance in future studies. Studies on consumers' perceptions of products are predominantly conducted in a laboratory and without the presence of other consumers. As such, the considerable effect of social presence is not considered at all (Nicholls, 2010). Our research demonstrates that the perception of products at the POS is influenced by social presence. Pure laboratory studies on product perception, product evaluation, and purchase decision without the manipulation of social presence could therefore be invalid and any management implications may be misleading. Future research should consider synchronously the physical and social components of a consumption area. Regardless of the considerable advantages offered by PLS, it is a soft modeling approach. Nevertheless, as we have applied the recommended quality criteria for the model characteristics, the model is robust and valid. The measurement of willingness to buy and the number of product alternatives considered using single items must be viewed with caution. PLS is not limited to the use of single items, so further studies should use scales with more than one item to reflect consistency to a greater extent (Hair et al., 2012). 4.3. Management implications Understanding the effects of social presence and proximity to other consumers has the potential to afford managers the opportunity to adopt strategic measures to guide and control consumers. An essential insight derived from this research is the fact that perception in the situation in front of the products is significantly influenced by distance from other consumers. The area of this influence can be estimated at about six feet (two feet on each side, left and right, and two feet directly in front of the consumer). On the basis of these insights, different management implications can be deduced. As one implication, the management should be particularly concerned about the fact that consumers should not come too close to each other when doing their shopping. For example, the management should ensure that advertised goods are not solely located in one row to prevent a gathering of consumers. This is especially relevant when the number of consumers in a sales area is very small to moderately sized, as consumers tolerate a violation of their personal space more readily when they knowingly enter situations with many other consumers (Becker, 1973). Furthermore, our results indicate that consumers should be encouraged not to stay too long in one place to make space for

M. Luck, M. Benkenstein / Journal of Retailing and Consumer Services 26 (2015) 104–114

others. Hence, the management should aim to ensure constant movement through the shop by providing permanent stimuli. If consumers occupy certain aisles and shelf sections due to long decision-making processes, the management needs to ease consumers' selection decisions. Prompt decision making by consumers clears the occupied places more quickly for the next consumer. There are several measures that can enable the management to regulate the duration of decision making. One design measure would be a targeted arrangement of products in different areas in a retail business. Products that require a longer process of decision making should be placed in less frequented areas. Thus, consumers do not occupy this area to the disadvantage of other consumers and they can make their decisions without being disturbed. As another design measure, the retail business should not organize product offers and positioning according to the available space but according to consumer frequency in the particular shelf sections. Due to a violation of personal space, fewer product alternatives are considered and consequently less information will be gathered. Hence, the management should aim to ensure good product presentation in general and an appropriate number of product alternatives. The ideal presentation of relevant purchase information enables quick decision making for the consumer. Moreover to counteract possible losses of sales due to customers lingering longer in one space, the management should provide a secondary placement on the shop floor for products that are bought with high frequency and need only little shelf space.

Appendix See Fig. A1, B1.

References Altman, Irwin, 1975. The Environment and Social Behavior: Privacy, Personal Space, Territory, Crowding. Brooks/Cole, Monterey, CA. Argo, Jennifer J., Dahl, Darren W., Manchanda, Rajesh V., 2005. The influence of a mere social presence in a retail context. J. Consum. Res. 32 (2), 207–212. http: //dx.doi.org/10.1086/432230. Argo, Jennifer J., Dahl, Darren W., Morales, Andrea C., 2006. Consumer contamination: how consumers react to products touched by others. J. Mark. 70 (2), 81–94. http://dx.doi.org/10.1509/jmkg.70.2.81. Argo, Jennifer J., Dahl, Darren W., Morales, Andrea C., 2008. Positive consumer contagion: responses to attractive others in a retail context. J. Mark. Res. 45 (6), 690–701. http://dx.doi.org/10.1509/jmkr.45.6.690. Argo, Jennifer J., Main, Kelley J., 2008. Stigma by association in coupon redemption: looking cheap because of others. J. Consum. Res. 35 (4), 559–572. http://dx.doi. org/10.1086/591102. Bagozzi, Richard P., Gopinath, Mahesh, Nyer, Prashanth U., 1999. The role of emotions in marketing. J. Acad. Mark. Sci. 27 (2), 184–206. http://dx.doi.org/10.1177/ 0092070399272005. Baker, Julie, Grewal, Dhruv, Parasuraman, Ananthanarayanan, 1994. The influence of store environment on quality inferences and store image. J. Acad. Mark. Sci. 22 (4), 328–339. http://dx.doi.org/10.1177/0092070394224002. Barclay, Donald W., Higgins, Christopher A., Thompson, Ronald, 1995. The partial least squares approach to causal modeling: personal computer adoption and use as illustration. Technol. Stud. 2 (2), 285–309. Baron, Reuben M., Kenny, David A., 1986. The moderator–mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J. Personal. Soc. Psychol. 51 (6), 1173–1182. http://dx.doi.org/ 10.1037/0022-3514.51.6.1173. Becker, Franklin D., 1973. Study of spatial markers. J. Personal. Soc. Psychol. 26 (3), 439–445. http://dx.doi.org/10.1037/h0034442. Belk, Russell W., 1974. An exploratory assessment of situational effects in buyer behavior. J. Mark. Res. 11 (2), 156–163. http://dx.doi.org/10.2307/3150553. Belk, Russell W., Kozinets, Robert V., 2005. Videography in marketing and consumer research. Qual. Mark. Res.: Int. J. 8 (2), 128–141. http://dx.doi.org/ 10.1108/13522750510592418. Brack, Anna D., Benkenstein, Martin, 2012. The effects of overall similarity regarding the customer-to-customer-relationship in a service context. J. Retail. Consum. Serv. 19 (5), 501–510. http://dx.doi.org/10.1016/j. jretconser.2012.06.006.

113

Brack, Anna D., Benkenstein, Martin, 2014. Responses to other similar customers in a service setting – analyzing the moderating role of perceived performance risk. J. Serv. Mark. 28 (2), 138–146. http://dx.doi.org/10.1108/JSM-05-2012-0089. Brocato, E. Deanne, Voorhees, Clay M., Baker, Julie, 2012. Understanding the influence of cues from other customers in the service experience: a scale development and validation. J. Retail. 88 (3), 384–399. http://dx.doi.org/10.1016/j. jretai.2012.01.006. Carrigan, Marylyn, Kirkup, Malcolm H., 2000. Video surveillance research in retailing: ethical issues. Int. J. Retail Distrib. Manag. 28 (11), 470–480. http://dx. doi.org/10.1108/09590550010356831. Dahl, Darren W., Argo, Jennifer J., Morales, Andrea C., 2012. Social information in the retail environment: the importance of consumption alignment, referent identity, and self-esteem. J. Consum. Res. 38 (5), 860–871. http://dx.doi.org/ 10.1086/660918. Dahl, Darren W., Manchanda, Rajesh V., Argo, Jennifer J., 2001. Embarrassment in consumer purchase: the roles of social presence and purchase familiarity. J. Consum. Res. 28 (3), 473–481. http://dx.doi.org/10.1086/323734. Dodd, Christopher A., Clarke, Ian, Kirkup, Malcolm H., 1998. Camera observations of customer behaviour in fashion retailing: methodological propositions. Int. J. Retail Distrib. Manag. 26 (8), 311–317. http://dx.doi.org/10.1108/ 09590559810231779. Donovan, Robert J., Rossiter, John R., 1982. Store atmosphere: an environmental psychology approach. J. Retail. 58 (1), 34–57. http://dx.doi.org/10.1016/00224359(94)90037-X. Eroglu, Sevgin A., Harrell, Gilbert D., 1986. Retail crowding: theoretical and strategic implications. J. Retail. 62 (4), 346–363. Eroglu, Sevgin A., Machleit, Karen A., 1990. An empirical study of retail crowding: antecedents and consequences. J. Retail. 66 (2), 201–221. Felipe, Nancy J., Sommer, Robert, 1966. Invasion of personal space. Soc. Probl. 14 (2), 206–214. Grove, Stephen J., Fisk, Raymond P., 1997. The impact of other customers on service experiences: a critical incident examination of “getting along”. J. Retail. 73 (1), 63–85. http://dx.doi.org/10.1016/S0022-4359(97)90015-4. Hair, Joe F., Sarstedt, Marko, Ringle, Christian M., Mena, Jeannette A., 2012. An assessment of the use of partial least squares structural equation modeling in marketing research. J. Acad. Mark. Sci. 40 (3), 414–433. http://dx.doi.org/ 10.1007/s11747-011-0261-6. Hall, Edward T., 1966. The Hidden Dimension. Anchor Books, Garden City, NY. Harrell, Gilbert D., Hutt, Michael D., 1976. Buyer behavior under conditions of crowding: an initial framework. Adv. Consum. Res. 3, 36–39. Hayduk, Leslie A., 1983. Personal space: where we now stand. Psychol. Bull. 94 (2), 293–335. http://dx.doi.org/10.1037/0033-2909.94.2.293. He, Yi, Chen, Qimei, Alden, Dana L., 2012. Consumption in the public eye: the influence of social presence on service experience. J. Bus. Res. 65 (3), 302–310. http://dx.doi.org/10.1016/j.jbusres.2011.03.014. Henseler, J.örg, Ringle, Christian M., Sinkovics, Rudolf R., 2009. The use of partial least squares path modeling in international marketing. Adv. Int. Mark. 20, 277–319. http://dx.doi.org/10.1108/S1474-7979(2009)0000020014. Hui, Michael K., Bateson, John E., 1991. Perceived control and the effects of crowding and consumer choice on the service experience. J. Consum. Res. 18 (2), 174–184. Hui, Sam K., Bradlow, Eric T., Fader, Peter S., 2009. Testing behavioral hypotheses using an integrated model of grocery store shopping path and purchase behavior. J. Consum. Res. 36 (3), 478–493. http://dx.doi.org/10.1086/599046. Iacobucci, Dawn, 2010. Structural equations modeling: fit indices, sample size, and advanced topics. J. Consum. Psychol. 20 (1), 90–98. http://dx.doi.org/10.1016/j. jcps.2009.09.003. Lowrey, Tina M., Otnes, Cele C., McGrath, Mary A., 2005. Shopping with consumers: reflections and innovations. Qual. Mark. Res.: Int. J. 8 (2), 176–188. http://dx.doi. org/10.1108/13522750510592445. Mano, Haim, Oliver, Richard L., 1993. Assessing the dimensionality and structure of the consumption experience: evaluation, feeling, and satisfaction. J. Consum. Res. 20 (3), 451–466. Manzo, John, 2005. Social control and the management of “personal” space in shopping malls. Sp. Cult. 8 (1), 83–97. http://dx.doi.org/10.1177/ 1206331204265991. Martin, Brett A.S., 2012. A stranger's touch: effects of accidental interpersonal touch on consumer evaluations and shopping time. J. Consum. Res. 39 (1), 174–184. http://dx.doi.org/10.1086/662038. McCracken, Grant, 1988. The Long Interview. Sage., Newbury Park, CA. McGrath, Mary A., Otnes, Cele C., 1995. Unacquainted influencers: when strangers interact in the retail setting. J. Bus. Res. 32 (2), 261–272. http://dx.doi.org/ 10.1016/0148-2963(94)00051-F. Mehrabian, Albert, Russell, James A., 1974. An Approach to Environmental Psychology, Cambridge, MA. MIT Press. Michon, Richard, Chebat, Jean-Charles, Turley, Lou W., 2005. Mall atmospherics: the interaction effects of the mall environment on shopping behavior. J. Bus. Res. 58 (5), 576–583. http://dx.doi.org/10.1016/j.jbusres.2003.07.004. Middlemist, R., Dennis, Eric S., Knowles, Matter, Charles F., 1976. Personal space invasions in the lavatory: suggestive evidence for arousal. J. Personal. Soc. Psychol. 33 (5), 541–546. http://dx.doi.org/10.1037/0022-3514.33.5.541. Nicholls, Richard, 2010. New directions for customer-to-customer interaction research. J. Serv. Mark. 24 (1), 87–97. http://dx.doi.org/10.1108/ 08876041011017916. Osmond, Humphry, 1957. Function as the basis of psychiatric ward design. Psychiatr. Serv. 8 (4), 23–29. http://dx.doi.org/10.1176/ps.8.4.23.

114

M. Luck, M. Benkenstein / Journal of Retailing and Consumer Services 26 (2015) 104–114

Pan, Yue, Siemens, Jennifer C., 2011. The differential effects of retail density: an investigation of goods versus service settings. J. Bus. Res. 64 (2), 105–112. http: //dx.doi.org/10.1016/j.jbusres.2010.02.011. Pornpitakpan, Chanthika, Han, Jie H., 2013. The effect of culture and salespersons' retail service quality on impulse buying. Australas. Mark. J. 21 (2), 85–93. http: //dx.doi.org/10.1016/j.ausmj.2013.02.005. Preacher, Kristopher J., Hayes, Andrew F., 2008. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behav. Res. Methods 40 (3), 879–891. http://dx.doi.org/10.3758/ BRM.40.3.879. Ridgway, Nancy M., Dawson, Scott A., Bloch, Peter H., 1990. Pleasure and arousal in the marketplace: interpersonal differences in approach-avoidance responses. Mark. Lett. 1 (2), 139–147. http://dx.doi.org/10.1007/BF00435297. Ringle, Christian M., Wende, Sven, Will, Alexander, 2005. SmartPLS 2.0 (beta). University of Hamburg, Hamburg, Germany. Rustemli, Ahmet, 1992. Crowding effects of density and interpersonal distance. J. Soc. Psychol. 132 (1), 51–58. Silberer, Günter, Wang, Hsiau-Han, 2010. Shopping with consumers as a research method. J. Mark. Trends 1 (2), 43–50. Söderlund, Magnus, 2011. Other customers in the retail environment and their impact on the customer’s evaluations of the retailer. J. Retail. Consum. Serv. 18 (3), 174–182. http://dx.doi.org/10.1016/j.jretconser.2010.09.006. Sommer, Robert, 1967. Sociofugal space. Am. J. Sociol. 72 (6), 654–660. http://dx. doi.org/10.1086/224402. Sommer, Robert, 1969. Personal Space: The Behavioral Basis of Design. PrenticeHall, Englewood Cliffs, NJ. Sorensen, Herb, 2003. The science of shopping. Mark. Res. 15 (3), 30–35. Spiggle, Susan, 1994. Analysis and interpretation of qualitative data in consumer research. J. Consum. Res. 21 (3), 491–503. http://dx.doi.org/10.1086/209413. Tombs, Alastair G., McColl-Kennedy, Janet R., 2003. Social-servicescape conceptual model. Mark. Theory 3 (4), 447–475. http://dx.doi.org/10.1177/

1470593103040785. Tombs, Alastair G., McColl-Kennedy, Janet R., 2010. Social and spatial influence of customers on other customers in the social-servicescape. Australas. Mark. J. 18 (3), 120–131. http://dx.doi.org/10.1016/j.ausmj.2010.04.001. Tombs, Alastair G., McColl-Kennedy, Janet R., 2013. Third party customers infecting other customers for better or for worse. Psychol. Mark. 30 (3), 277–292. http: //dx.doi.org/10.1002/mar.20604. Turley, Lou W., Milliman, Ronald E., 2000. Atmospheric effects on shopping behavior: a review of the experimental evidence. J. Bus. Res. 49 (2), 193–211. http: //dx.doi.org/10.1016/S0148-2963(99)00010-7. Uhrich, Sebastian, 2011. Explaining non-linear customer density effects on shoppers' emotions and behavioral intentions in a retail context: the mediating role of perceived control. J. Retail. Consum. Serv. 18 (5), 405–413. http://dx.doi.org/ 10.1016/j.jretconser.2011.06.002. Uhrich, Sebastian, Luck, Michael, 2012. Not too many but also not too few: exploring the explanatory mechanisms for the negative effects of low customer density in retail settings. Qual. Mark. Res.: Int. J. 15 (3), 290–308. http://dx.doi. org/10.1108/13522751211231996. Watson, David, Clark, Lee A., Tellegen, Auke, 1988. Development and validation of brief measures of positive and negative affect: the PANAS scales. J. Personal. Soc. Psychol. 54 (6), 1063–1070. http://dx.doi.org/10.1037/0022-3514.54.6.1063. Wood, C.M., Scheer, Lisa K., 1996. Incorporating perceived risk into models of consumer deal assessment and purchase intent. Adv. Consum. Res. 23, 399–404. Worchel, Stephen, Teddlie, Charles, 1976. The experience of crowding: a two-factor theory. J. Personal. Soc. Psychol. 34 (1), 30–40. http://dx.doi.org/10.1037/00223514.34.1.30. Zaltman, Gerald, 2003. How Customers Think: Essential Insights into the Mind of the Market, Boston. Harvard Business Press, MA.