• No results found

Tapping atmospherics for the retail experience

N/A
N/A
Protected

Academic year: 2021

Share "Tapping atmospherics for the retail experience"

Copied!
112
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

University of Amsterdam – Amsterdam Business School MSc. in Business Administration – Marketing Track

27th of January 2017 Master’s Thesis – Final version Nastie Schoenmakers 11122005

Supervisor: dhr. dr. A. Zerres

(2)

Statement of originality

This document is written by Nastie Schoenmakers, who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document are 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.

(3)

ACKNOWLEDGEMENT

This Master’s Thesis is the final step to complete my MSc. in Business Administration at the University of Amsterdam. This research revolves around retail atmosphere and would not have been possible without the help of a number of people.

First of all, I would like to express my thanks to dhr. dr. Alfred Zerres who supervised this entire thesis from beginning to end. It was always particularly helpful to discuss difficulties in the thesis project throughout. Moreover, the critical feedback given strengthened this thesis.

Furthermore, I would like to thank the whole Asics team and in particular Martin Block. Thank you for making this project possible and believing in us from the start. Moreover, I would like to thank, Lara Galka, Emma Kraanen and especially Lukas Janßen, who have also written their Master’s Thesis in corporation with Asics. This project would not have been the same without you.

Moreover, I would like to express my gratitude to Eveline Jansen who was at all times understanding and able to help. Lastly, special thanks to my parents and brother for always supporting and believing in me.

(4)

Table of content

ABSTRACT 6

INTRODUCTION 7

LITERATURE REVIEW 9

RETAIL ATMOSPHERE 9

AFFECT,EMOTION AND MOOD 11

EFFECTS OF RETAIL ATMOSPHERE ON CONSUMER’S EMOTION 13

LINKING THE RETAIL ATMOSPHERE TO STRATEGY 14

STRUCTURE 15

DATA AND METHOD STUDY 1 17

INTRODUCTION STUDY 1A AND B 17

RESEARCH DESIGN 17

DATA COLLECTION 19

RESULTS STUDY 1 21

SAMPLE DESCRIPTION 21

DESCRIPTIVE RETAIL ATMOSPHERE ELEMENTS 22

CONCEPTUALIZATION 26

EXPLORATORY TESTING OF THE CONCEPTUALIZATION 26

DATA AND METHOD STUDY 2 28

RESEARCH DESIGN AND DATA COLLECTION PROCEDURE 28

SAMPLE 30

MEASURE DEVELOPMENT 31

ANALYTICAL STRATEGY 31

RESULTS STUDY 2 33

SAMPLE DESCRIPTION 33

PREPARING THE DATA 34

CORRELATION AND RELIABILITY ANALYSIS 35

FACTOR ANALYSIS 38

DIRECT EFFECTS OF 16RETAIL ELEMENTS 39

TESTING THE COMPLETE CONCEPTUALIZATION 42

DISCUSSION 56 DISCUSSION OF RESULTS 56 THEORETICAL CONTRIBUTIONS 61 PRACTICAL IMPLICATIONS 62 LIMITATIONS 64 FUTURE RESEARCH 65 CONCLUSION 67 REFERENCES 69

(5)

APPENDICES 74

APPENDIX A:ATMOSPHERIC RETAIL ELEMENTS,TURLEY AND MILIMAN (2000) 74

APPENDIX B:QUESTIONNAIRE STUDY 1 75

APPENDIX C:FREQUENCY TABLES STUDY 1 83

APPENDIX D:RESULTS STUDY 1A AND B 85

APPENDIX E:APPROACH GUIDELINE 88

APPENDIX F:STUDY 2 SUB-QUESTIONNAIRE 1 90

APPENDIX G:FREQUENCY TABLES STUDY 2 96

APPENDIX H:NORMALITY CHECKS 98

APPENDIX I:FACTOR ANALYSIS 99

APPENDIX J:REGRESSIONS 102

APPENDIX K:PROCESS MODEL 6 108

APPENDIX L:DIRECT,TOTAL INDIRECT AND TOTAL EFFECTS 109

(6)

Abstract

This exploratory research aimed to take a more global approach towards the retail atmosphere. In particular, this research intended to explore the most influential antecedents of retail atmosphere and their subsequent effects on the overall perceived atmosphere, consumer’s emotion and an important marketing objective, brand image. The ultimate goal of study 1 was to yield a scale containing the most influential retail elements which (i) is able to capture the customer’s perceptions with regard to these retail elements and (ii) is reasonably short in order to be used in the main study and beyond this research in practical research settings. The scale contains 16 retail elements, which subsequently were used in study 2. This study addressed the question whether these elements have an effect on brand image directly and indirectly through both the overall perceived atmosphere and consumer’s emotion. Only the direct effect of colour schemes (+) on brand image was significant. However, the following retail elements had a significant indirect effect on brand image mediated through both the overall perceived atmosphere and consumer’s emotion: products sold in the store (+), exterior signs (-), employee characteristics (+), entrance (+), product displays (+), racks and cases on which products are displayed (+), grouping of products (-) and colour schemes (+).

Keywords: retail atmosphere, overall perceived atmosphere, consumer’s emotion, brand image, retail strategy.

(7)

Introduction

Retail research has always been of the utmost importance in the marketing field, both academically and for practitioners it stands at the top of their agenda (Grewal, Levy, & Kumar, 2009). Retailers are lifting the in-store experience into a rather exciting and entertaining experience in order to get shoppers through the door. The store has changed from a transaction place to a place used more and more as a showroom, museum and fulfilment centre. Retailers are increasingly making shopping fun and entertaining in order to induce store visits. Shoppers do not love a store merely because they love the merchandise it carries. They love a store because it touches them personally and emotionally. In the recent years it has become apparent that shoppers focus on the total customer experience (Danziger, 2006). How a store makes consumers feel is getting more and more important. The overall perceived atmosphere in a retail store is an important dynamic in this process (Verhoef et al., 2009).

The current literature has often taken a micro approach towards the retail atmosphere. It mostly focuses on minor changes in the retail atmosphere and the subsequent effects on consumer behaviour (Turley & Chebat, 2002). Even though this research stream is proven to be extremely valuable, as minor changes in the retail atmosphere are capable of changing consumer’s behaviour (Turley & Milliman, 2000). The literature lacks a more holistic and global approach for the research on retail atmosphere itself (Mattila & Wirtz, 2001) and for the research linking the retail atmosphere to strategic objectives of a firm (Turley & Chebat, 2002).

This research aims to address the absence of the more global overview within the retail atmosphere literature and the linkage to strategy. First of all, the lack of the more holistic approach on retail atmosphere is addressed with study 1. Study 1 aims to identify which retail elements are the (most) influential antecedents of “atmosphere” and to yield a scale which measures the consumer’s perception with regards to these elements. This study takes the well-cited experimental evidence review article of Turley and Milliman (2000) as a starting point in which they have gathered an extensive list of 57 retail elements that make up the retail atmosphere. Subsequently,

(8)

the absence of the global approach towards the linkage of the retail atmosphere to strategic objectives of a firm in literature is addressed with study 2. The fundamental goal of study 2 is to explore what the effects are of the most influential retail elements on brand image directly and indirectly through the overall perceived atmosphere and consumer’s emotion. The overall perceived atmosphere is taken into account because the overall perceived atmosphere is the most holistic response a person can have towards the retail environment (Rayburn & Voss, 2013). Consumer’s emotion is taken into account because they are widely acknowledged to be influenced by retail atmospherics (Burns & Neisner, 2006). Moreover, emotions are central to the actions of consumers (Bagozzi, Gopinath, & Nyer, 1999). Study 2 explores whether the retail atmosphere has an effect on an important marketing objective. This is highly relevant because a well-managed brand image is able to predict consumer behaviour (Marks & Olson, 1981), increase brand equity (Keller, 2003; Sweldens, Van Osselaer, & Janiszewski, 2010) and shape a company’s financial performance (Kim, Gon Kim, & An, 2003). The present research therefore intends to make contributions to the theory and practice of retail atmosphere.

This structure of this research is as follows. First, the relevant literature is discussed. Second, the data and method together with the results of study 1 are debated. This study focuses on the antecedents of retail atmosphere and is needed to develop a conceptualization which explores the relationship between the most descriptive retail elements and brand image, possibly through the overall perceived atmosphere and consumer’s emotion. Third, with the conceptualization in mind the data, method and results of study 2 are presented. Fourth, the discussion of the results is presented and finally, the conclusion of this research is offered.

(9)

Literature review

Retail atmosphere

An extensive research body on retail atmosphere is available. One aspect that the scholars who research the subject have in common is the division of atmosphere in different categories. Kotler (1973) theoretically describes atmosphere as the air surrounding a sphere and adds that the term is also used to define the quality of the surroundings. Elaborating on that, the term atmospherics is used to describe the conscious designing of a space to create certain effects via an emotional reaction induced by the atmosphere in customers. Atmosphere is seized through the senses; sight, sound, scent and touch, and therefore Kotler categorizes atmosphere using the following four sensory terms; visual, aural, olfactory and tactile. The visual dimensions of atmosphere include colour, brightness, size and shapes. The aural dimensions of atmosphere are volume and pitch. The olfactory or aromatic dimensions are scent and freshness. Last, the tactile dimensions of the atmosphere are softness, smoothness and temperature. As Kotler already specified in 1973, the atmosphere is becoming more important for firms to create a differential advantage (Kotler, 1973).

Bitner (1992) explores the impact of physical surroundings on customers and employees and divides the atmosphere in three elements: ambient conditions, space and function, and signs, symbols and artifacts. The first atmosphere element of Bitner, ambient conditions, relates to the senses and can be compared to Kotler’s sensory terms of atmosphere. These are taken together by Bitner and named the ambient conditions. These include general characteristics of the environment (visual), noise and music (aural), scent (olfactory) and temperature (tactile). Next to this, Bitner differentiates between space and function and signs, symbols and artifacts. The space and function relates to the outline of the store or layout, equipment and furnishing. The signs, symbols and artifacts include the signage, personal artifacts and style of the decor. Bitner states the importance of aligning the needs of the ultimate user and the requirements of functional units to ensure strategy advantage from the atmosphere.

(10)

Berman and Evans (1995) depict that the exterior of the store is included in the atmosphere. Therefore, they paint that the atmospheric stimuli contain four categories, namely the exterior of the store, the general interior, the layout and design variables, and the point-of-purchase and decoration variables. Here the general interior relates mostly to Kotler’s (1973) four sensory terms and the ambient conditions of Bitner (1992). The layout and design variables can be compared to the second atmospheric element of Bitner; the space and function. Moreover, the symbol and artifact element of Bitner is comparable to the decoration category of Berman and Evans. Berman and Evans (1995) add the exterior of the store, because in their opinion the experience of a customer starts already outside of the retail store. Therefore, it should be taken into account when reviewing the atmosphere.

Turley and Milliman (2000) reviewed the experimental evidence of atmospheric effects on shopping behaviour. They build further upon the four categories of Berman and Evans, and depict a fifth category should be added, namely the human variables. The five categories are the following. 1) The exterior variables include the entrance, exterior display windows and height of the building among others. 2) The general interior variables incorporate colour schemes, lightening, music, scents, temperature and cleanliness together with other variables. 3) The layout and design variables are made up of the placement of merchandise, traffic flow, furniture and waiting ques. 4) The point-of-purchase and decoration variables include signs and cards, artwork, product displays and price displays. 5) The human variables include employee characteristics, crowding in the store, customer characteristics and privacy. The complete list, containing 57 retail elements can be found in Appendix A.

Turley and Miliman’s (2000) article will be taken as the starting point in this exploratory research for several reasons (i) the division of the atmosphere in five basic variables creates an organized and logic structure, (ii) it provides a comprehensive list of atmospheric retail elements which fall under those five atmospheric variables, and (iii) the complete list addresses the atmosphere as consumers do – in a holistic way.

(11)

Overall Perceived Atmosphere

It is important to make a distinction between the intended and perceived atmosphere. The perceived atmosphere of the same retail store can vary for different consumers (Kotler, 1973). The overall perceived atmosphere builds on the understanding that people holistically process the atmosphere from the perceptions they hold about it (Rayburn & Voss, 2013). Matilla and Wirtz (2001) argue that past studies have mostly focused on one or two specific stimuli (e.g. music, colour or scent) in the atmosphere. Research has failed to address how all stimuli in a retail store interact. This is highly important because, the atmosphere is perceived and processed by consumers not piece by piece, but rather in a complete and holistic way (Bitner, 1992; Jang & Namkung, 2009; Mattila & Wirtz, 2001).

Rayburn and Voss (2013) describe the perceived overall atmosphere as the overall impression of the retail’s store atmosphere as a pleasant or unpleasant place to shop. It is said to be the most holistic response a person can have towards the retail environment. Important to note is that the perceived atmosphere and experienced mood or emotions are concepts that are related, however not the same. Vogels (2008) confirms this with the performed experiment that aims to quantify the perceived atmosphere with expressions used to describe the atmosphere of an environment. The final list of 38 terms describing perceived atmosphere underlie 4 dimensions, namely: coziness, liveliness, tenseness and detachment. The concept of perceived atmosphere and emotion differ because, emotion, unlike atmosphere is an affective state or feeling. Atmosphere on the other hand is a subjective impression of the environment and does not directly relate to the actual effect on mood (Vogels, 2008).

Affect, Emotion and Mood

Important to clarify are the related, however distinctive, concepts of affect, emotion and mood, which in the past were often wrongly used interchangeably. Affect is the main concept that can be defined as a feeling state most apparent in emotion and mood, however always available to

(12)

consciousness (Ekkekakis, 2012). Affect does not need to be directed at anything, however it is caused by something (e.g. the weather or an event) and includes a sense of pleasure or displeasure, tension or relaxation and depression or excitement (Russell & Barrett, 1999). Emotion is always concerned with an explicit object, such as a person, an event or a thing and can be in the past, present, future, real or even imaginary, at which the emotion is directed to (Solomon, 1976). Emotions include one being afraid of, angry with, in love with and has pity for, and have a cognitive component to it (Russell & Barrett, 1999). Moods in general last longer than emotions and are about the affective states that are about nothing exact or about everything and could be about the world in general (Frijda, 2009). Thus, the causation of a specific mood may not always be easily identified (Ekkekakis, 2012).

Emotions are intended, based on specific objects and lean toward actions (Frijda, 1993) Moreover, Bagozzi, Gopinath and Nyer (1999) describe emotions as central to the actions of consumers. Therefore, emotions are interesting to study in relation to the retail environment.

Consumer’s Emotion

Emotions can be measured with different approaches and are studied widely in relation to consumer behaviour (Laros & Steenkamp, 2005). Mehrabian and Russell (1974) measure emotions alongside three dimensions with the PAD-model (pleasure, arousal and dominance). It is shown that human judgments on diverse stimuli, including retail stores, can be described in terms of these three dimensions. Pleasure, arousal and dominance are identified as affective responses. Pleasure is associated with positive or negative emotions. It ranges from unpleasant to pleasant and can be worded as the range between unhappy and happy. Arousal denotes the activity level both mentally and physically. This dimension ranges from calm to exited and among others is directly related to the amount of information in the environment. Dominance indicates the control or lack of control over the situation and is related to the extent to which the consumer feels free to handle in the way they please (Mehrabian & Russell, 1974).

(13)

Russell theory and eliminate the dominance dimension. They found that the pleasure and arousal are sufficient to characterize the emotional or affective responses in a wide range of environments, including the retail environment. The dominance dimension is found to require a cognitive understanding of the environment and situation and consequently not solely related to affective responses (Russell & Pratt, 1980). The pleasure and arousal dimensions are found to interact suggesting when aroused in pleasant situations this intensifies the approach behaviour in that (retail) environment and vice versa for a high arousal degree in unpleasant environments where it will strengthen the avoidance behaviour of that environment (Donovan & Rossiter, 1982).

The grouping of emotions in positive and negative affect is another popular conceptualization. Positive Affect (PA) is the degree to which a person feels enthusiastic, active and alert. On the other hand, Negative Affect (NA) is about distress and an unpleasurable engagement. To measure the underlying mood factors a brief 20-item scale is developed called the PANAS scales (Watson, Clark, & Tellegen, 1988).

Much later, Laros and Steenkamp (2005) argue that even though positive and negative affect is widely used in consumer behaviour, important distinctions are lost with this measurement. Therefore, a hierarchical model of consumer emotion is proposed integrating the different various research streams. The first layer of the hierarchy is still twofold, positive and negative affect, under this layer four general emotions are placed. For positive affect contentment, happiness, love and pride and for negative affect anger, fear, sadness and shame. Below the second layer a third layer is placed consisting of 42 more specific emotions falling within the before mentioned eight general emotions (Laros & Steenkamp, 2005).

Effects of Retail Atmosphere on Consumer’s Emotion

Many studies have previously researched the effects of the retail environment on consumer’s emotion and the subsequent influence on consumer purchasing behaviour (Bitner, 1992; Donovan & Rossiter, 1982; Mattila & Wirtz, 2001). A retail environment consists of retail elements, as

(14)

mentioned before, which are holistically perceived as the store atmospherics. Store atmospherics are capable of affecting consumers both cognitively and emotionally (Burns & Neisner, 2006).

Scholars in the field of retail atmospheric have mostly focused on specific atmospheric elements (Turley & Milliman, 2000). Seen here is that atmospheric stimuli are manipulated, many scholars have focused on variables such as music (Areni, 2003; Hul, Dube, & Chebat, 1997; Mattila & Wirtz, 2001), scent (Mattila & Wirtz, 2001; Michon, Chebat, & Turley, 2005; Spangenberg, Crowley, & Henderson, 1996), colour (Bellizzi & Hite, 1992; Crowley, 1993) and lightening (Areni & Kim, 1994; Baker, Grewal, & Parasuraman, 1994). Turley and Milliman (2000) provide an overview on the experimental evidence of the atmospheric effects on shopping behaviour. They found that the general interior variables of a store atmosphere were among the ones that were most widely studied.

Research has shown that positive perceptions of a retail environment have an upwards influence on positive emotions, patronage intentions and on the desire to remain longer in a retail setting (Andreu, Bigné, Chumpitaz, & Swaen, 2006). Moreover, positive emotions results in a higher desire to stay in the store (Babin & Darden, 1995; Donovan & Rossiter, 1982), faster decision times (Isen, 1989), buying of more items (Sherman & Smith, 1987), spent more money on articles customer’s just liked instead of needed (Spies, Hesse, & Loesch, 1997) and have a higher satisfaction with the retailer (Babin & Darden, 1996; Machleit & Eroglu, 2000).

Linking the Retail Atmosphere to Strategy

As seen before, there is written abundantly on the effects of retail atmosphere on emotions, time spent in store and satisfaction. On the other hand, the literature on using the retail atmosphere as a strategic tool is scarce even though this subject directly corresponds with managerial interests (Turley & Chebat, 2002). Turley and Chebat (2002) state that the absence of the discussion of issues like linking retail atmosphere with corporate and marketing objectives are hardly ever

(15)

discussed in literature. This research therefore takes a more global approach and attempts to link the retail atmosphere with an important marketing or corporate objective, namely brand image.

Brand Image

Communicating a brand image to targeted consumers has long been a fundamental part of the marketing plan (Gardner & Levy, 1955). From a brand or company’s perspective a well-managed brand image is important for at least three reasons. First, attitudes toward the brand are helpful in predicting consumer behaviour (Marks & Olson, 1981). Second, the activities conducted to induce positive brand attitudes can increase the brand equity (Keller, 2003; Sweldens et al., 2010). Thirdly, in the long run the brand image consumers’ hold can shape a company’s financial performance (Kim et al., 2003).

Brand image and brand awareness together make up the knowledge consumers have about a brand. Brand knowledge is theorized as a point in memory to which multiple associations of the brand are linked. Brand image can be defined as perceptions consumers’ hold about a brand as replicated by the brand associations held in memory. Brand associations contain the meaning brands hold for consumers. Types, favorability, strength and uniqueness of brand associations influence the brand image. There can be distinguished between three different types of brand associations: (i) attributes that characterize a product or service, (ii) benefits or personal values attached to a product or service and (iii) brand attitudes or the overall evaluation of the brand.

Structure

This research has an exploratory nature. It first aims to identify which retail elements are the most influential antecedents of the retail atmosphere. It subsequently uses these elements in a scale. This scale is used to explore the effects these elements have on brand image directly and indirectly, through the overall perceived atmosphere and consumer’s emotion.

The remainder of this research is structured as follows. Study 1 is designed to identify the retail elements that are the most influential antecedents of retail atmosphere. The data, method and

(16)

results of this study are presented right after this chapter. Hereafter, the conceptualization is presented. The conceptualization is used as a basis to explore the effects the identified retail elements from study 1 have on brand image. This is done with study 2, which follows after the conceptualization with the data, method and result section. Thereafter, the discussion of the results are presented together with theoretical and practical implications, limitation of this research and suggestions for future research. To end with, the conclusion of this research is presented.

(17)

Data and Method Study 1

Introduction Study 1a and b

This research builds on the comprehensive list of atmosphere retail elements provided in the well-cited article from Turley and Milliman. This article is used as input for the scale development process in this study. As mentioned before, the total retail atmosphere consists of five variables (Turley & Milliman, 2000). These five variables contain in total 57 retail elements, whereof 50 are applicable for this particular study. Seven out of 57 elements of this list are excluded, as they do not represent retail elements of a typical store. The overview of the retail element list can be found in Appendix A. The first two studies aim to reduce Turley and Milliman’s (2000) 50 elements to a more manageable number of items. The fundamental goal of this study is to identify the retail elements with the greatest influence on retail atmosphere and to yield a scale which (i) measure the customer’s perceptions with regard to the identified elements and (ii) is reasonably short to be used in the main study of this research and beyond this study in practical research settings. In-store surveys are an example hereof. In order to get the desired response rate these surveys need to be reasonably short. Moreover, long questionnaires might induce respondents to skip questions or respond casually (Ninot, Fortes, & Delignières, 2006).

Research Design

Methodologically, this study follows the approach applied in Aaker’s (1997) research on Dimensions of Brand Personality. In this research a set of personality traits was reduced from 309 traits to a more practicable number of 114 traits that were identified as most descriptive for a brand personality. In order to do this subjects rated how descriptive the 309 traits were for a brand’s personality on a 7-point Likert scale (1 = not at all descriptive, 7 = extremely descriptive). To isolate the most descriptive personality traits for a brand a cut-off rate of 6 (very descriptive) was

(18)

selected. This cut-off rate included “only” 114 personality traits in the main study of Aaker’s (1997) article.

Study 1 intends to identify the most influential retail elements from the extensive list of Turley and Milliman (2000) which make up the retail atmosphere. This is done by the means of two studies, study 1a and b. Both studies are conducted using a cross-sectional survey design and administered in Qualtrics1, in English. The studies are mainly similar in their regard however have one important distinction. Study 1a follows the division of the retail atmosphere in the five atmospheric variables as suggested in literature – external, interior, layout and design, point of purchase and decoration, and human variables. Consequently, study 1a asks the respondent to rate how descriptive they find the respective retail element (e.g. colour schemes) for the atmospheric variable it belongs to (the interior variable). Study 1b ignores the distinction of the retail atmosphere in the five atmospheric variables. It asks the respondent to rate how descriptive they find the retail element (e.g. colour schemes or the entrance) for the overall retail atmosphere. Both studies ask this in the manner described for all 50 retail elements on a 7-point Likert scale (1 = not at all descriptive, 7 = extremely descriptive). The retail elements in both studies are randomized.

Both studies start the questionnaire with the question to recall a recent shopping experience where they visited a fashion or sporting goods retail store. It is asked to take this experience into account for the remainder of the questionnaire. This recall is strengthened by asking the respondent to close their eyes for a few seconds and try to recall the trip from the moment they arrive at the store until they leave the store. Moreover, a few questions about this experience are asked. To ensure the respondents understand what the retail atmosphere entails a description is given in both studies. Besides this explanation also an explanation of the five atmospheric variables is given in study 1a. Hereafter, the main questions follow as outlined in the previous paragraph. Lastly, both studies end with several demographic questions. The complete questionnaire of both study 1a and b is presented in Appendix B.

(19)

The goal of the distinction between study 1a and b is explained here. First of all, with conducting two studies it is possible to discover communalities between retail elements that are descriptive for the atmospheric variable it belongs to (study 1a) and the overall retail atmosphere (study 1b). Thus, by conduction two studies it is possible to find out which aspects remain the same when the way it is asked for varies. Second, because of the exploratory nature of this research it is important to work in a thorough manner. Hence, conducting two different variations of the study is done because it is necessary to be sure to select the right retail elements to carry forward in the main study.

Pilot test study 1. A pilot test with 13 participants was done for study 1a. The survey was evaluated per question with all participants individually. This was helpful in order to find out whether the questions were clear and interpreted as the researcher meant them. An important adjustment according to the feedback was made. 9 out of 13 test respondents had difficulty with the term descriptive. As this word is used in the main question of the questionnaire this was of the utmost importance to address. It was decided to include an explanation of descriptive. This looks as follows. “In the next questions you will be asked to rate various retail elements on how descriptive they are for you for the above mentioned atmospheric variables (study 1a) / the overall retail atmosphere (study 1b). Descriptive can be explained as the extent to which something is able to describe and characterize another thing.” Hereafter the questionnaire was discussed once more with three test persons. All deficiencies were addressed and the questionnaire was found to be clear and ready to distribute. As study 1a and b were fairly similar the enhancements made in 1a were also applicable to study 1b.

Data Collection

As said, both questionnaires were administered in Qualtrics. However, the distributions of the two studies differ. Study 1a was distributed via social networking sites and therefore a non-probability convenience sampling technique. The data collection of study 1a stretched over a total of 18 days.

(20)

The incentive for participating was a €50,- online voucher which was raffled among the participants who filled in their e-mail address at the end of the questionnaire. Study 1b was distributed via Amazon Mechanical Turk2 and therefore is a non-probability convenience sampling technique as well. Respondents were paid 0.50 U.S. Dollars per completed questionnaire. The data collection of study 1b stretched over a total of 12 days.

2

(21)

Results Study 1

Sample Description

Specifics of the sample of both studies will be discussed here. In total study 1a received 313 responses whereof 31,9% is uncompleted. After removing the unfinished responses, 213 usable survey responses are left. A bit over half of the respondents are male (53.5%). The ages of the respondents range from 15 to 63 years old. The average age of the respondents is 27 years old (N=213, mean=27, mode=24). Respondents from 27 different nationalities answered the questionnaire. Most respondents are from the Netherlands (71.4%), México (7%), Belgium (3.3%), Germany (3.3%) and Spain (2.3%). Most respondents obtained a Bachelor’s degree of applied sciences (29.6%), hereafter a University Bachelor’s degree (28.6%). Of all respondents a bit less than half (47.4%) was a student at the time they filled in the questionnaire. Today’s students are greatly involved in consumer activities from a young age and therefore, with their vast consumer experience, are a suitable sample for this study (Paul, 2001). Data on income is sensitive and 18.8% chose to not disclose this information. Both less than €10,000 (22.1%) and between €10,000 and €19,999 (22.1%) was earned by most respondents.

Study 1b received 234 responses whereof 16.7% was unfinished. The uncompleted responses are manually deleted; hereafter 195 usable responses are left. From the total responses in this study 55.9% are male and 1% did not want to disclose their gender. The respondents range between 18 and 72 years old (N=195, mean=34.1, mode=31) Respondents from 10 different nationalities answered the questionnaire, whereof 94.9% respondents are from the United States of America. Most respondents obtained a University Bachelor’s degree (36.9%), hereafter most respondents attended some college without obtaining a degree (23.1%). The majority of respondents are working as paid employees (69.2%). Only 1.5% of respondents preferred not to disclose data on their income, most respondents earned between $20,000 and $29,999 (16,4%),

(22)

hereafter most respondents earned between $30,000 and 39,999 (14.4%). Several frequency tables for both studies are presented in Appendix C.

Descriptive Retail Atmosphere Elements

In order to find out which retail elements are most descriptive for either the respective atmospheric variable (study 1a) or the overall retail atmosphere (study 1b) the responses are analysed with SPSS Statistics 21. In order to find the retail elements, which are rated as most descriptive, the elements are sorted descending on the mean. The complete list can be found in Appendix D.

In order to isolate the most descriptive retail elements there must be decided upon a cut-off rate. This can be done in several ways. For instance, Aaker (1997) reduces a list of personality traits from 309 to 114 traits that are descriptive for brand personality. This is done by deciding on a cut-off rate of 6 which is “very descriptive”, on the same 7-point Likert scale (1= not at all descriptive, 7= extremely descriptive) as used in this study. When looking at this study, the most descriptive retail element for both the respective atmospheric variable (study 1a) and the overall retail atmosphere (study 1b) is products sold in the store. It is rated with a descriptiveness mean of 5.58 and 5.55 in study 1a and b respectively. Aaker (1997) decided on a cut-off rate of 6 (very descriptive), however this is not possible in this study as the highest rated retail element has a descriptiveness level of 5.58 and 5.55 (5= descriptive – 6= very descriptive). Therefore, I took another approach to reduce the 50 retail elements to a more manageable number. When plotting the number of retail elements against the descriptiveness level in a graph multiple elbows were found. The figures for both studies are presented on the next page. Study 1a shows two elbows at 16 and 3 retail elements, which have a 4.75 and 5.25 “level of descriptiveness” respectively. Study 1b shows three elbows that are seen at 11, 7 and 1 retail element(s), which have a 4.5, 4.75 and 5.25 “level of descriptiveness”. The elbows are pointed out in the figure with an arrow. After the elbows the number of retail elements level off. The decision for the precise cut-off rate is made with two principles in mind (i) it is necessary to include enough retail elements to yield a comprehensive

(23)

scale, however the scale also (ii) needs to be reasonably short in order to be used in the main study and beyond this research in practical settings. For these reasons, the cut-off rate is decided at the first elbow of both studies. This is at 16 retail elements for study 1a and at 11 retail elements for study 1b. This corresponds to a level of descriptiveness for study 1a of 4.75 and of 4.5 for study 1b. As can be seen, for both studies the cut-off rate is between 4 (somewhat descriptive) and 5 (descriptive).

Figure 1: Study 1a: Number of retail elements with descriptiveness cut-off rate (N=213)

Figure 2: Study 1b: Number of retail elements with descriptiveness cut-off rate (N=195)

(1= Not at all descriptive, 2= Not descriptive, 3= Somewhat not descriptive, 4= Somewhat descriptive, 5= Descriptive, 6= Very descriptive and 7= Extremely descriptive).

The most descriptive retail elements, until the cut-off rate (16 and 11 retail elements for study 1a and b respectively), are presented in table 1 on the next page, sorted descending on the mean. The retail elements in grey are below the cut-off rate (4.5) of study 1b however, are presented in order to make a better comparison between the two studies.

0 10 20 30 40 50 >3,25 >3,5 >3,75 >4,0 >4,25 >4,5 >4,75 >5,0 >5,25 >5,5 >5,75 >6,0 N um ber o f r et ail el em en ts Descriptiveness mean

Study 1a: Number of retail elements with descriptiveness cut off rate

0 10 20 30 40 50 >3,25 >3,5 >3,75 >4,0 >4,25 >4,5 >4,75 >5,0 >5,25 >5,5 >5,75 >6,0 N um ber o f r et ail el em en ts Descriptiveness mean

(24)

Table 1: Most Descriptive Retail Elements

Study 1a – cut-off rate 4.75 Study 1b – cut-off rate 4.5

Retail element N Mean Retail element N Mean

Products sold in the store 213 5.58 Products sold in the store 195 5.55

Cleanliness 213 5.42 Placement of products 195 5.15

Exterior signs 213 5.37 Cleanliness 195 5.13

Employee characteristics 213 5.16 Grouping of products 195 5.12

Lighting 213 5.15 Product displays 195 5.12

Placement of products 213 5.13 Exterior signs 195 5.07

Entrance 213 5.05 Racks and cases 195 4.86

Racks and cases 213 5.02 Price displays 195 4.72

Product displays 213 5.01 Lighting 195 4.71

Address and location* 213 5.00 Space design in store 195 4.68

Exterior display window* 213 4.98 Colour schemes 195 4.54

Grouping of products 213 4.95 Entrance 195 4.49

Music* 213 4.92 Traffic flow of customers* 195 4.49

Space design in store 213 4.86 Signs and cards* 195 4.48

Colour schemes 213 4.84 Size of building* 195 4.46

Price displays 213 4.79 Employee characteristics 195 4.37

It can be concluded from the table above that the two studies conducted are fairly similar in their outcome. Both studies show that products sold in the store is the highest rated retail element with a rounded descriptiveness mean of 5.6. However, there are few differences that should be mentioned. To begin with study 1b levels off more quickly than study 1a. This can easily be seen in figure 1 and 2 presented on the previous page. To demonstrate this, study 1b only has 11 retail elements rated at a descriptiveness level of 4.5 or higher, whereas study 1a has 32 retail elements at that level or higher. Moreover, minor differences can be seen when looking at the 16 most descriptive retail elements for both studies (including the grey text in table 1). Address and location, exterior display window and music are not among the top 16 most descriptive retail elements for atmosphere in study 1b, while they are in 1a. When looking to study 1b, traffic flow of customers, signs and cards and size of building are rated midst the most descriptive retail elements for atmosphere in study 1b, however they do not make the top 16 in study 1a. These retail elements include an asterisk. An explanation for these differences could be that study 1a asks respondents to rate how descriptive the retail elements are for the atmospheric variable it belongs to. Study 1b asks the respondent how descriptive the retail elements are for the overall atmosphere. This could result in the mentioned differences because, consumers might think different retail variables describe the

(25)

overall atmosphere in a store, not even thinking about a few retail elements that might come to mind when they are asked to rate the descriptiveness of retail elements for e.g. the exterior of the store. This could be an explanation why address and location and exterior display window are included amongst the most descriptive retail elements in study 1a and not in 1b. These elements are only found to be amongst the top descriptive retail elements when asked for the exterior atmospheric variable and not for the overall atmosphere.

In order to move forward to study 2 it is important to decide upon which retail elements to carry forward. Due to the extensive body of knowledge in literature, which divides the retail atmosphere in five atmospheric variables it is decided to carry forward the 16 retail elements of study 1a. This distinction also made sure that the distribution between the five atmosphere variables was better. (e.g. when taking the cut-off rate of 4.5 of study 1b no retail element belonging to the human variable will be included). The 16 elements of study 1a make up a comprehensive scale that is able to measure the customer’s perception towards the most influential retail elements. Moreover, the scale is reasonably short to be usable in the main study and beyond this research in practical retail settings.

(26)

Conceptualization

In study 1 I uncovered the 16 most descriptive variables for retail atmosphere. This study took the division of the retail atmosphere in five atmospheric variables into account. These 16 retail elements will be carried forward in study 2. The aim of study 2 is twofold (i) it measures whether the perceptions of the 16 retail elements have an effect on brand image directly and (ii) indirectly through the overall perceived atmosphere and consumer’s emotion.

Exploratory testing of the conceptualization

This research has an exploratory approach towards the uncovering of the effects the 16 retail elements may have on brand image, both directly and indirectly through the overall perceived

Products sold in the store Cleanliness Exterior signs Employee characteristics Lighting Placement of products Entrance Racks and cases Product displays Address and location Exterior display window

Grouping of products Music Space design in store

Colour schemes

Figure 3:Conceptualization of Retail Atmosphere and the Effect on Brand Image

Overall Perceived

Atmosphere Brand Image

Consumer’s Emotion

(27)

atmosphere and consumer’s emotion. In doing so, several main effects were explored, which are discussed here.

I tapped into the influence the retail elements have on (i) the overall perceived atmosphere, (ii) consumer’s emotion and (iii) brand image directly. I take this first step to test which retail elements actually affect the three constructs. This is beneficial for practitioners because retail atmospherics and their elements are of strategic importance (Turley & Chebat, 2002) and even minor differences in the retail atmosphere can have impact on consumer behaviour (Turley & Milliman, 2000). Therefore, the knowledge of which retail elements significantly influence the overall perceived atmosphere, consumer’s emotion and brand image directly are of strategic value for retailers. Retail stores are a tool for brands to differentiate themselves in the minds of consumers (Baker, Levy, & Grewal, 1992). With the knowledge of which retail elements are most important to focus on, the road to a sustained competitive advantage may become clearer.

Next to that, I looked into the effect the 16 retail elements have on brand image through both the overall perceived atmosphere and consumer’s emotion. Both cognitive evaluation and emotions play an important role in the retail setting (Burns & Neisner, 2006). The effect of the retail atmosphere on customer’s emotion is widely studied (Andreu et al., 2006; Mehrabian & Russell, 1974; Spies et al., 1997). However, the linkage of the effects of the retail atmosphere with corporate and marketing objectives is hardly present (Turley & Chebat, 2002). Therefore, I take a macro approach and explore the effects of the 16 retail elements on brand image, a fundamental part of the marketing plan, possibly through by the overall perceived atmosphere and consumer’s emotion. To the best of my knowledge this has not been studied yet.

(28)

Data and Method Study 2

Research Design and Data Collection Procedure

This quantitative research was conducted using a cross-sectional survey design. A questionnaire was administered in Qualtrics in cooperation with Asics to collect data in four different cities, namely Amsterdam (the Netherlands), Brussels (Belgium), Barcelona (Spain) and London (United Kingdom). The survey was originally administered in English, hereafter the survey was translated to Dutch, French, Spanish and German. Language errors are uncovered using back-translation.

The questionnaire was distributed by four researchers in the flagship Asics store in each of the aforementioned cities in order to collect real-life data. Researchers were not to mistaken for shop personnel and therefore wore Asics clothes, however distinct from what the store personnel were wearing. Moreover, researchers had to wear a key cord with the University of Amsterdam student card visibly around their neck at all times. Customers were approached in store either when they intended to leave the store or, in the case of a purchase, right after they finished at the cashier’s desk. Important to note is that customers were approached right after they experienced the store and the emotions that came with it. Customers were motivated to take part in the 5-6 minute survey by offering them an immediate incentive in the form of a candy bar. Moreover, they were told they had the chance to win a pair of running shoes if they completed the survey and filled in their email address in the end. A typical intro text was the following and always was initiated in the language of the country: “Good morning/day/afternoon, my name is ___. Could I ask you a question? I am from the University of Amsterdam and currently writing my Master’s Thesis. I am conducting research on retail and it would be of great help if you could spare 5-6 minutes to fill in this survey. If you complete the survey you have the chance to win a pair of running shoes and there is also some candy. Would you be willing to help?” When the customer agreed to participate in the survey they filled in the survey on a tablet. The researchers always highlighted that the customer can answer the questions in an honest fashion. That the research is carried out independent and that their

(29)

response is strictly anonymous. One of the researchers was always close to the participant in order to answer any questions the respondent may have. This induced the respondent to complete the survey in a serious manner and the response bias was minimized. The complete approach guideline that was created for this research is presented in Appendix E.

The data collection stretched over a total of two weeks. Data collection commenced Monday 21st of November where two students collected data in Barcelona and two in London. The week after, the second data collection period started where two students collected data in the Amsterdam store and two in the Brussels store.

Pilot test study 2. The researchers executed a pilot test in the Amsterdam store over the period of two days. The first day two researchers went into the store for around three hours and conducted questionnaires with the customers. Next to the approach text mentioned earlier, the customers were told that they were part of a pilot test. The researchers asked the customers to give their honest feedback. It was immediately noticed that the questionnaire took quite long – around 10-12 minutes. Customers gave the feedback to shorten the questionnaire. Next to this the customers that participated found the questions to be clear. The researchers took the feedback into account and the questionnaire was shortened at the end of the first pilot test day. This was done through creating two versions of the questionnaire which are named sub-questionnaire 1 and sub-questionnaire 2. Instead of including all four research topics in one questionnaire the questionnaire was split into two in order to avoid a high dropout or non-response rate. Sub-questionnaire 1 includes two out of four research subjects namely the research on retail atmosphere and technology in retail. Sub-questionnaire 2 includes the other two research subjects namely, shopping companion and body weight. In both questionnaires some questions are the same, including demographics. Qualtrics randomly assigned either one of the two questionnaires to the next respondent.

A concern from the researchers examining the research topics of sub-questionnaire 2 was that they would not get enough specific customers that they needed for their research. The researcher on shopping companion needed both shoppers who were alone as in a group. Next to that the research on bodyweight needed athletic, normal and over-weighted customers in order to

(30)

complete the research. This was ensured by adding a randomization at the start of Qualtrics, which would only be visible for the researchers. The randomization included the following three variables gender (male, female), shopping formation (alone, group) and body weight (athletic, normal and overweight). The researcher tried to select the “ideal” next respondent based on the randomization (e.g. female, alone, normal). If this was not possible the researcher selected the next customer leaving the store.

The two sub-questionnaires, revised with a randomization of the “ideal” next participant to select, were tested on the second pilot test day. Researchers again approached the customers with the approach text and told them this was part of a test run for their research. The researchers again asked the customers for honest feedback. The second pilot test day confirmed the customers thought the questions were clear. Moreover, the questionnaire now took around 5-6 minutes and this was acceptable for most customers. The researchers therefore used the spilt design of the questionnaire in the data collection period in the four cities.

Sub-questionnaire 1, which includes the questions for this research and for the technological research, can be found in Appendix F.

Sample

The population for this survey consists of all retail consumers in the Netherlands, Belgium, Spain and the United Kingdom. The population is large and the sampling frame is unknown. The research was conducted using both non-probability purposive and convenience sampling. The researchers chose the purposive sampling technique in order to ensure that all necessary different respondents were selected. For example, the researcher from the bodyweight research needed both athletic and overweight participants. The researchers therefore added a randomization, as explained before, in Qualtrics. If it was not possible to choose the “ideal” next respondent, researchers approached the next customer that left the store as a convenience sampling technique.

(31)

Measure Development

With the results of the first study the 16 most descriptive retail elements form a scale which is used in study 2. Respondents were asked to rate the perception of the 16 retail elements on a 7-point Likert scale (E.g. How did you perceive the following retail atmosphere elements of the Asics store? Products sold in the store: 1= Very poor, 7= Exceptional). I developed this scale for the purpose of this research and therefore no report on the Cronbach’s alpha was available at this point.

The overall perceived atmosphere construct was measured with the validated four item 7-point semantic differential scale from Rayburn and Voss (2013). In their article they report on a Cronbach’s alpha of .91 (E.g. Please rate the overall perceived store atmosphere on the following traits: Uncomfortable – Comfortable)

The construct consumer’s emotion was measured with an adapted version of the 17 item scale by Tsai, Chang, Chuang and Wang (2008). They report on a Cronbach’s alpha of .85. The researchers adapted the scale to a 6-item semantic differential scale containing the dimensions of pleasure, arousal and dominance as described in literature (Mehrabian & Russell, 1974). (E.g. Please indicate what emotions you experienced during the store visit: I was... Unhappy – Happy).

The validated four item 7-point semantic differential scale from Spears and Singh (2004) was used to measure brand image and has a Cronbach’s alpha of .95. (E.g. Please describe your overall feelings about Asics as a brand: Unappealing – Appealing).

All respondents were asked several demographic questions including gender, year of birth, education level and yearly income. Moreover, buying behaviour was recorded. If the respondent bought something the number of items and amount spent was recorded.

Analytical strategy

This exploratory research aimed to uncover the effects the 16 retail elements have on brand image, both directly and indirectly through the overall perceived atmosphere and consumer’s emotion. In doing so I explored these effects with several statistical tests using SPSS Statistics 21.

(32)

First of all, I computed regression analyses to test the direct effect of the 16 retail elements on the overall perceived atmosphere, consumer’s emotion and brand image. I examined these direct effects through a multiple regression with the overall perceived atmosphere, consumer’s emotion and brand image as the dependent variable.

I explored the mediation effects of the overall perceived atmosphere and consumer’s emotion on the effect the 16 retail elements have on brand image. To do this the PROCESS SPSS macro of Hayes (2013) was used. This macro uses bias corrected bootstrap confidence level intervals of 95% with 5000 bootstrap resamples to test the mediation effects. Significant effects are present when zero is excluded from these intervals. Model 6 from this macro is used two times to test for the serial mediation effect. With the first analysis I tested the effect of the 16 retail elements together on brand image through both the overall perceived atmosphere and consumer’s emotion. More interesting perhaps is the second analysis. With this analysis I tapped into the actual effects the 16 retail elements have separately on brand image through both the overall perceived atmosphere and consumer’s emotion. Therefore, in this analysis the retail elements were not added as one but as 16 separate independent variables. In all statistical procedures p-values of <.001, <.01 and <.05 were used to indicate the statistical significant effects.

(33)

Results Study 2

Sample Description

As said before, the total questionnaire consisted of four research subjects. The survey was split in two to make the questionnaire more manageable for participants. Qualtrics randomly assigns either one of the two surveys to the next respondent. First a description of the total data set is given. After, the focus is on the specific survey that contains the questions for this research.

In total the four researchers asked 1061 customers if they wanted to participate in the questionnaire. The response rate was 70.8%, which leaves 751 customers who filled in the questionnaire. Hereof 11 questionnaires were only partly completed and therefore manually deleted from the data set. This leaves 740 usable surveys (N=740). The ages of the respondents ranged between 12 and 90 years old with the average age being approximately 37 years old (N=740, mean=37.3, mode=37). Of the total respondents 65% are locals and 35% are tourists. A few respondents did not want to disclose information on their education level (3.5%). Most respondents obtained 36.5% a Master’s degree, hereafter 36.1% obtained a University bachelor degree. Many respondents did not want to specify any information on their income (29.5%), most respondents earned less than €10.000 a year (9.2%) after which between €20.000 and €29.999 was picked most often (8.6%). Several complete frequency tables can be found in Appendix G.

As said before the total questionnaire (N=740) had two sub-questionnaires in order to research four separate subjects (sub-questionnaire 1 n=372, sub-questionnaire 2 n=368). Sub-questionnaire 1 included two of these subjects, retail atmosphere – this research – and technology in retail. The description of the sample of sub-questionnaire 1 is given here. The ages of the respondents range between 12 and 90 years old (n=372, mean= 36.9, mode=27). 51.9% of the respondents are male and 1.3% did not want to specify their gender. The distribution among the different research locations can be seen in table 2 on the next page.

(34)

Most respondents for sub-questionnaire 1 are obtained in Belgium, Brussels. Most respondents obtained a Master’s degree (37.3%), after which 36.3% obtained a University bachelor. Again, a big part of the respondents

did not want to specify their income level (29%). From the respondents who did state their income level most respondents earned between €30.000 and €39.999 (10.5%), hereafter between €10.000 and €19.000 (9.4%). Several complete frequency tables can be found in Appendix G.

Preparing the Data

I took several steps to prepare the data for the analysis. First, frequencies are checked of the variables that are used for this

research in order to examine errors in the data, however no errors were found. After, the means are computed for the following constructs: 16 retail elements, overall perceived atmosphere, consumer’s emotion and brand image. Next, normality of these constructs is assessed to give insight in the distribution of the outcomes. Normal distributions have values of both skewness and kurtosis of 0. The skewness and kurtosis values of the construct can be viewed 3 above. The skewness of the variables range between -.1.047 and .018. The complete descriptive tables are presented in Appendix H. Brand image is substantially negatively skewed, atmosphere is moderately negatively skewed, and emotion and retail elements are approximately normally distributed. A negative value of skewness indicates a concentration of scores on the right side of the distribution. The kurtosis values range between -.684 and .725. The kurtosis values of atmosphere and brand image indicate a heavy-tailed distribution, whereas the values for emotion and retail elements indicate a flat and light-tailed distribution. Third, a test of normality is done with the Shapiro-Wilk statistic, which

Table 2: Research Location

Research location n Percentage

Amsterdam 94 25.3% Brussels 113 30.4% Barcelona 90 24.2% London 75 20.2% Total 372 100%

Table 3: Skewness and Kurtosis

Skewness Kurtosis n Statistic Std. Error Statistic Std. Error Atmosphere 372 -0.835 0.126 0.467 0.252 Emotion 372 -0.224 0.126 -0.684 0.252 Retail Elements 372 0.018 0.126 -0.554 0.252 Brand Image 372 -1.047 0.126 0.725 0.252 Valid N (listwise) 372

(35)

looks if the distribution of scores deviates from a similar normal distribution. These scores are significant (p <.05). This means that the distributions of the constructs are significantly different from a normal distribution (Field, 2013). Last, a check was done for univariate outliers in the data set. When looking into the minimum and maximum of the Z-scores computed it can be seen that four cases for atmosphere and three cases for brand image below -3 are identified. These are excluded from further analysis, thus n=365.

In sum, the results of the preliminary analysis show that the data is not normally distributed. Field (2013) states that when the sample gets bigger the assumption of normality becomes less important. This is because the sampling distribution is normal regardless of what the population data looks like. Moreover, normality is considered as the least essential of assumptions required for linear regression (Hayes, 2013). Plus research has uncovered that normality violations only impact statistical inferences in small samples (e.g. N < 30) (Hayes, 1996).

Correlation and Reliability Analysis

An overview of the descriptive statistics, correlations and scale reliabilities is presented in table 4 on the next page together with the reliability analysis in table 5 and 5.1.

Correlation Analysis

The correlation matrix includes the independent variables: products sold in the store, cleanliness, exterior signs, employee characteristics, placement of products, lighting, entrance, product displays, racks and cases, address and location, exterior display window, grouping of products, music, space design in store, colour schemes and price displays and the 16 retail elements taken together. The two mediators: overall perceived atmosphere and consumer’s emotion and the dependent variable: brand image. Moreover, two control variables, gender and research location, are presented as well. Notable outcomes are discussed here.

The independent variables, the 16 retail elements separate and the 16 retail elements taken together, all significantly positively correlate with the two mediators, overall perceived atmosphere

(36)

Table 4: Correlation analysis

Variables M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

1. Gender 1.50 0.52 -

2. Research location 2.41 1.06 0.02 - 3. Products sold in the store 5.16 1.11 -0.04 .10 - 4. Cleanliness 5.73 1.04 -0.02 .11* .56** - 5. Exterior sign 5.09 1.27 -.12* .14** .50** .43** - 6. Employee characteristics 5.50 1.18 -0.06 .08 .55** .52** .47** - 7. Placement of products 5.21 1.13 -0.08 .18** .68** .57** .53** .57** - 8. Lighting 5.36 1.18 -0.04 .22** .51** .57** .47** .48** .55** - 9. Entrance 5.14 1.20 -.11* .13* .60** .55** .58** .57** .60** .58** - 10. Product displays 5.26 1.10 -0.08 .12* .62** .52** .54** .55** .66** .57** .61** -

11. Racks and cases 5.08 1.11 -0.05 .22** .62** .54** .48** .48** .73** .56** .58** .66** -

12. Address and location 5.40 1.18 -0.10 .20** .46** .49** .39** .42** .46** .46** .54** .51** .56** -

13. Exterior display window 5.07 1.24 -0.10 .12* .49** .45** .60** .47** .58** .50** .63** .62** .54** .44** -

14. Grouping of products 5.18 1.11 -0.07 .15** .60** .57** .48** .56** .72** .54** .56** .62** .68** .45** .54** - 15. Music 4.80 1.29 -0.03 .13* .43** .40** .34** .42** .44** .45** .39** .42** .39** .39** .41** .41** - 16. Space design 5.20 1.10 -0.02 .14** .67** .59** .50** .56** .70** .61** .61** .67** .70** .54** .54** .69** .44** - 17. Colour schemes 5.22 1.20 -0.03 .23** .58** .57** .50** .53** .64** .58** .64** .65** .66** .48** .55** .64** .39** .59** - 18. Price display 4.74 1.35 -0.07 .12* .50** .48** .43** .45** .53** .49** .50** .54** .49** .40** .40** .52** .39** .53** .49** - 19. 16 Retail elements 5.20 0.88 -0.08 .20** .78** .73** .69** .72** .83** .75** .80** .81** .80** .67** .73** .80** .60** .82** .79** .69** (.95) 20. Overall p. atmosphere 6.08 0.79 0.04 .07 .46** .31** .28** .41** .41** .33** .46** .48** .43** .30** .40** .33** .32** .43** .44** .32** .51** (.87) 21. Consumer’s emotion 5.58 0.93 -0.03 .06 .50** .35** .32** .48** .45** .36** .42** .48** .45** .31** .38** .43** .40** .43** .41** .35** .55** .64** (.84) 22. Brand Image 6.16 0.87 0.02 .14** .39** .33** .31** .41** .41** .38** .39** .40** .40** .31** .36** .40** .30** .38** .47** .34** .50** .53** .51** (.93)

*. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed). n=365. Cronbach’s alphas are presented on the diagonal

Table 5.1: Reliability analysis - continued

Construct n Number of

items

Cronbach’s alfa Overall perceived atmosphere 365 6 .87

Emotion 365 6 .84

Pleasure 365 3 .88

Arousal 365 1 -

Dominance 365 2 .66

Brand Image 365 5 .93

Table 5: Reliability analysis

Construct n Number of items Cronbach’s alfa Retail Elements 365 16 .95 External 365 4 .81 Interior 365 5 .83

Layout and design 365 4 .90

Point of purchase and decoration 365 2 .70

(37)

and consumer’s emotion, and with the dependent variable brand image at the p <.01 level. These findings suggest that when the perception of the retail elements goes up also the overall perceived atmosphere, consumer’s emotion and brand image are more positive evaluated. None of the retail elements have a correlation below .2, with any other variable, which would indicate there was an absence of relation.

It should be noted that the highest correlations are found in row 19, because this variable is the mean of the 16 retail elements taken together, therefore it is logical that the retail elements separately correlate the strongest with this variable. Especially, placement of products (r = .83, p <.01), product displays (r = .81, p <.01), racks and cases on which products are displayed (r = .80, p <.01) and space design and allocation hereof (r = .82, p <.01), as these variables have correlations of .8 or higher.

The 16 retail elements taken together correlate the highest with the overall perceived atmosphere (r = .51, p <.01), consumer’s emotion (r = .55, p <.01) and brand image (r = 50, p <.01). When looking at the retail elements separately it can be seen that products sold in the store correlates highest to both the overall perceived atmosphere (r = .46, p <.01) and consumer’s emotion (r = .50, p <.01), and employee characteristics correlates the highest with brand image (r = .41, p <.01).

Gender has two significant negative correlations, namely with exterior signs (r = -.12, p <.05) and entrance (r = -.11, p <.05). However, these correlations are between -.2 and .2, which suggest an absence of relation. Research location significantly correlates with 14 variables (all but products sold in the store and employee characteristics) and the 16 retail elements taken together. Thereof, lighting (r = .22, p <.01), racks and cases on which the products are displayed (r = .22, p <.01), address and location (r = .20, p <.01) and colour schemes (r = .23, p <.01) have an effect above .2 indicating there is a tendency to positive relation between the variables.

Referenties

GERELATEERDE DOCUMENTEN

This work is dedicated to my mother and father who gave their life to proclaim the good news in Yekepa and the surrounding region, so that the Mano and Gio

At the end of 2015, IUPAC (International Union of Pure and Applied Chemistry) and IUPAP (International Union of Pure and Applied Physics) have officially recognized the discovery of

Note that the optimal size of a shopping area is the larger the higher the demand for retail products (in our case measured by footfall) and the lower the rent of the

The current study aims to gain insight in the effects of store sales in the context of a Dutch retail chain by examining whether higher levels of store sales are indirectly

According to chapter 2, the partner selection criteria which should be applied for co-development projects between manufacturer and supermarket chains are the following:

When performing a regression analysis on the whole model, with all the factors and their components and the moderating effects included, only responsiveness of

market shares Sum of the market shares of Lidl and Aldi stores (computed on the basis of the store’s net sales floor) in the City quarterly Supermarket Gids..

are perhaps the only openly religious group to have reached beyond their Christian flock of fans and had a couple of Top 20 hits.. But their singer, Martin Smith, believes that