How Utilitarian And Hedonic Shopping Motivations Affect The Usage Of Self- Scanning Technologies

(1)

How Utilitarian And Hedonic Shopping

Motivations Affect The Usage Of

Self-Scanning Technologies

And customers’ satisfaction with the shopping experience

(2)

1

HOW UTILITARIAN AND HEDONIC SHOPPING MOTIVATIONS

AFFECT THE USAGE OF SELF-SCANNING TECHNOLOGIES

And customers’ satisfaction with the shopping experience

University of Groningen

Faculty of Economics and Business MSc Marketing Management Master Thesis

22 June 2015

(3)

2

Management summary

Self-service technologies (SSTs) are a new way of service offering for many retail and service environments. In recent years, the implementation of self-scanning checkouts, which belongs to the broad term of SSTs, has grown in the Netherlands and became more and more part of customers daily life. But how do customers perceive the self-scanning checkouts? Does the opportunity of checking out with self-scanning increase customer satisfaction with the shopping experience? And does this differ between customer shopping motivations? In order to take full advantage of the investments in self-scanning checkouts, it is important for companies to find an answer to those questions and to understand which factors influence customers’ self-scanning usage.

In order to provide companies with answers to those questions, this research focuses on the relation between perceived self-scanning characteristics and the usage of self-scanning checkouts. Furthermore, the relationship between the usage of self-scanning checkouts and customers’ satisfaction with the shopping experience is studied. And, most important, the influences of customers’ shopping motivations on those relationships is studied. The goal of this research is to explore the relation between shopping motivations and self-scanning checkouts. Therefore, the following research question is presented:

“To what extent do hedonic and utilitarian shopping motivations influence the relation between SST characteristics and self-scanning usage, and the relation between self-scanning usage and satisfaction with the shopping experience?”

A quantitative research, with survey data from 257 respondents, is conducted to provide an answer to the research question. Four perceptions of self-scanning characteristics are measured. Hedonic and utilitarian shopping motivations are included in this research, and usage of self-scanning and customers’ satisfaction with the shopping experience are measured. Additionally, the results were compared to a model that includes people’s usage intentions, instead of actual usage.

(4)

3

are the less they use self-scanning. This research found indications towards the moderating effects of the shopping motivation on the relation of perceptions of enjoyment and the usage of self-scanning. Likewise, the hedonic motivation seem to negatively influences the relationship between perceived usefulness and the intentions to use self-scanning. Furthermore, hedonic motivated shoppers have lower intentions to use self-scanning, although no negative significant relation is found between hedonic motivation and actual usage. No moderating effects were tested significant on the relationship of self-scanning usage and satisfaction with the shopping experience.

(5)

4

Preface

The master thesis that lies in front of you is the final product of the years I spent as a student at the University of Groningen. It is the result of a last semester of hard work towards a master in Marketing Management. After my bachelor Business Administration I could not decide in what direction I wanted to specialize. A gap year, consisting of working at a worldwide consultancy- and engineering company and traveling the world with my best friend, made me realize that my interests were in business, consumer behavior, strategy and creativity. Therefore, the master Marketing Management was a perfect fit.

The subject of this research is about self-scanning technologies and customer shopping motivations, which I chose because of my enthusiasm for smoothly operating modern technologies and consumer behavior. By studying the perceived characteristics of self-scanning, the usage of self-scanning and the influence of customer shopping motivations, my interests could be perfectly combined within my thesis. Now that I finished the last part of the graduation, it is time to start the working-life. And although it is a pity that this means the end of my time as a student, it is also the start of something new. I am looking forward to apply the knowledge and experiences that I have learned during my years as a student into my future career.

I would like to take the opportunity to thank those who helped me during my thesis process. First of all I would like to thank my supervisors prof. dr. Verhoef and MSc. De Haan, whom provided me with insightful feedback and were always willing to give advice and guidance when needed. I would like to thank my thesis group for the useful discussions and the share of thoughts on the topics. Likewise, I wish to thank my closest friends for always being around and for providing a mood booster when needed. A special thanks to my parents, who always supported me during my study in Groningen. Last but not least, I am grateful for the 257 respondents who completely filled in the questionnaire. Without them, I could not have finished my thesis.

Groningen, 22 June 2015

(6)

5

1. Introduction

In the age of upcoming focus on the customer (the decade of customer experiences and customer-centricity) and developments in Self-Service Technologies (SSTs), it is not surprising that SSTs are being used as a new way of service offering in many retail and service environments. SST is a broad term that can be defined as “technological interfaces that enable customers to produce a service independent of direct service employee involvement” (Meuter et al. 2000, p. 50). It offers benefits to companies as well as to customers. Companies can provide from significant cost savings, increased efficiency and enhanced consistency of their service delivery (Collier and Sherrell, 2010; Weijters et al. 2007). Customers on the other side, are offered the possibility to create a service outcome without direct service employee involvement and can experience increased feelings of control (Collier and Sherrell, 2010; Meuter et al. 2000). Although these developments seem promising, it also raises questions like: do customers become more satisfied by using SSTs? And, with keeping ‘customer-centricity’ in mind, do SSTs please all types of customers or only a specific shopping type?

SSTs appear in many types and forms. Examples are self-service check-ins at airports, online shopping check-outs and mobile payment applications. In recent years, the implementation of another type of SST has grown: the in-store self-scanning technologies. These technologies create the possibility for customers to check out without employee interaction. Although this self-scanning technology exists for almost three decades, it failed in the beginning. However, customers are more technological comfortable nowadays and more ready to use SSTs (Dabholkar, et al. 2003). In the Netherlands, self-scanning technologies become more and more part of customers’ daily life. Supermarkets like Albert Heijn, Plus and Jumbo introduced self-scanning devices that customers can take into the store during their grocery shopping. Convenience store Albert Heijn To Go, furniture store Ikea, and several service environments like Pathé Cinema adopted self-scanning technologies at their payment desks, to (partially) replace the regular employee involved check outs. In March 2015, Albert Heijn even announced that they will start with the introduction of a mobile self-scanning check out application in 200 of their supermarkets (www.ah.nl).

(9)

8

primarily driven by two categories of antecedents; SST characteristics and individual differences (Dabholkar 1996; Meuter et al., 2005). According to literature, the main characteristics of SSTs that drive technology adoption are the perceived usefulness, ease of use, enjoyment, and controllability (Dabolkar and Bagozzi, 2002; Walker et al. 2002, Wang et al., 2013). Additionally there are indications from research that customer shopping motivations might influence SST usage (Marzocchi and Zammit, 2007), but no prior studies have dived into the differences between the hedonic or utilitarian shopping motivations and their influence on self-scanning usage.

Furthermore, although prior research provides insights into customers’ satisfaction with the SST, it remains quite unexplored how SST usage affect customers’ satisfaction with the shopping experience as a whole. Prior research indicates that customer shopping motivations might influence the usage and satisfaction with the shopping experience (Meuter et al., 2005; Marzocchi and Zammit, 2007). Therefore, this research will try to address this knowledge gap, by focusing on two customer shopping motivations as potential moderator for the relation between the perceived characteristics of self-scanning and the usage of self-scanning. Additionally, the influence of the shopping motivations on the relationship between self-scanning usage and satisfaction with the shopping experience will be studied. The customer shopping motivations will be differentiated by hedonic motivations and utilitarian motivations. The utilitarian motivation concentrates on the most generic goal of shopping, while the hedonic motivation represents shopping for entertainment and the emotional worth (Babin et al. 1994).

Furthermore, this research will focus specifically on self-scanning technologies at the checkout desk and is conducted to find an answer to the question:

“To what extent do hedonic and utilitarian shopping motivations influence the relation between SST characteristics and self-scanning usage, and the relation between self-scanning usage and satisfaction with the shopping experience?”

(10)

9

The structure of this research will be as follows. First, the literature review in Chapter 2 will provide theoretical background on the constructs. The hypotheses are formulated and the conceptual model will be presented at the end of the chapter. In Chapter 3, methodology, the research and questionnaire design, measurements and plan of analysis will be shown. Chapter 4 will present and discuss the results of this research. In Chapter 5, the results will be analyzed and a further discussion of the results will be presented. The paper will conclude with managerial implications, limitations and suggestions for further research.

2. Theoretical Framework

This chapter provides an overview of relevant literature in the field of SSTs and self-scanning, and presents the formulated hypotheses. First, the chapter gives an overview on the types of SSTs and the developments related to SSTs and self-scanning. Second, the hedonic and utilitarian shopping motivations will be discussed. The chapter will continue with an elaboration of scientific literature about characteristics of SSTs, that influences the usage. Next, the construct of satisfaction with the shopping experience will be explained. Finally, the conceptual model will be presented.

2.1 SSTs and self-scanning technologies

(11)

10

willing to act as co-creators of the service. They need to have positive perceptions of the technology, as a means for attaining their goals more easily rather than as an obstacle. Since there is relatively little literature that focuses specifically on self-scanning technologies, literature about SSTs in general will be used in this chapter additionally.

2.2 Hedonic and Utilitarian Shopping Motivations

Understanding customer shopping motivations is key to gain insights in how those motivations might influence self-scanning usage and satisfaction with the shopping experience. Hirschman and Holbrook (1982) describe the utilitarian motivated customer as the “problem solvers”. They act to “get something” (Triandis, 1977) and purchase products in a “deliberant and efficient manner” (Babin et al., 1994). Those customers might prefer SST characteristics such as ease of use, usefulness and control, because of their goal-oriented way of shopping. The second customer type that Hirschman and Holbrook describe is the hedonic customer. Those customers are seeking for fun, fantasy, arousal, sensory stimulation and enjoyment, and they shop because “they love it” (Triandis, 1977). The hedonic motivation represents the emotional worth of the shopping experience (Jones et al. 2006). Mano and Oliver (1993) found a positive causal link between hedonism and satisfaction and more recent research demonstrated that enjoyment of self-scanning usage could even positively influence the perceived quality (Anselmsson, 2001).

2.3 Characteristics of Self-Service Technologies

(12)

11

TRA as a theoretical basis to explain the causal linkage between the two determinants for acceptance (perceived usefulness and ease of use), and users’ attitudes, intentions and actual adoption behavior (Davis et al., 1989). This research will include the two variables of the TAM and will evaluate them separately, since prior factor analyses suggested that perceived ease of use and perceived usefulness are statistically distinct dimensions (Hauser and Shugan 1980; Larcker and Lessig 1980; Swanson 1987).

Furthermore, several studies found additional drivers for the usage of SSTs. Next to ‘perceived ease of use’ and ‘perceived usefulness’ (the two major variables derived from the TAM), ‘perceived enjoyment’ (or fun) and ‘perceived control’ appear in the literature as drivers that have a positive impact on the usage of SSTs and the intentions to use (Dabholkar, 1996; Dabholkar and Bagozzi, 2002; Curran and Meuter, 2005; Weijters et al., 2007; Zhao et al., 2008).

Perceived ease of use is one of the two variables that the TAM assumes as determining for customers’ acceptance of technology systems. It is defined as the extent to which a customer believes whether using the SST will be free of effort (Davis et al., 1989). It is tied to an individual’s assessment of effort involved in using the system (Venkatesh, 2000) and focuses on the process that leads to the final outcome (Weijters et al., 2007). Utilitarian motivated shoppers prefer the shopping process to be efficient and functional, and they focus on how useful and beneficial the process is (Batra and Athola, 1991). Therefore, it is expected that positive perceptions of ease of use of utilitarian motivated shoppers has a higher positive influence on their usage behavior, than the perceptions of ease of use of hedonic shoppers. This leads to the following hypotheses:

H1a: Perceived ease of use will have a positive relationship with the usage of self-scanning. H2a: The utilitarian shopping motivation does strengthen the relation between perceived ease of use and the usage of self-scanning, more than the hedonic shopping motivation does.

(13)

12

performances will be enhanced by the system and one’s perceptions of usefulness will increase. However, since main goal of this research is to discover the topic of self-scanning and the influences of customers shopping motivations, the relationship between ease of use and usefulness will not be included in the model. These two perceptions will be treated as distinct dimensions. Perceived usefulness is a goal-oriented variable, and is expected to be most important to utilitarian motivated customers that are concerned with buying products in a timely and efficient manner (Childers et al. 2002; Sherry et al., 1993). Therefore:

H1b: Perceived usefulness will have a positive relationship with the usage of self-scanning. H2b: The utilitarian shopping motivation does strengthen the relation between perceived usefulness and usage of self-scanning, more than the hedonic shopping motivation does.

Perceived enjoyment is an addition to the TAM model and refers to the extent to which a customer perceives that the activity of using technology is enjoyable in its own right (Davis, et al. 1989). Enjoyment as a characteristic of SSTs that influences the usage behavior, is consistent with prior research into retail shopping behavior. Perceived enjoyment is an intrinsic motivator to use technology (van der Heijden, 2004) and it contains the fun and entertaining aspects of the shopping experience that hedonic motivated shoppers appreciate (Babin et al. 1994, Childers et al., 2002). Considering the TAM framework, where ease of use and usefulness reflect the more instrumental aspects of shopping, enjoyment relates to the hedonic aspects of shopping. Therefore:

H1c: Perceived enjoyment will have a positive relationship with the usage of self-scanning. H2c: The hedonic shopping motivation does strengthen the relation between perceived enjoyment and usage of self-scanning, more than the utilitarian shopping motivation does.

(14)

13

Sherrell, 2010). Hui and Bateson (1991) argue that perceived control is the key to unlock customers’ emotional and behavioral responses to service encounters. They empirically showed that greater perceived control is associated with greater customer satisfaction, deeper customer loyalty, and better service quality. Later studies also found that perceived control influences self-service quality evaluations (Dabholkar, 1996; Dabholkar et al. 2003). Therefore:

H1d: Perceived control will have a positive relationship with the usage of self-scanning. H2d: The utilitarian shopping motivation does strengthen the relation between perceived control and usage of self-scanning, more than the hedonic shopping motivation does.

2.4 SSTs and the satisfaction with the shopping experience

Although drivers of SST usage have been studied extensively, less is known about what happens next. When customers have been convinced to use self-scanning; can the usage of these self-scanning checkout desks influence their satisfaction with the shopping experience? Oliver (1996, p. 13) defines customer satisfaction as “the consumer’s fulfillment response.” “It’s a judgment, providing a pleasurable level of consumption-related fulfillment”. Satisfaction is an important factor, because prior research has shown that satisfaction is a primary driver of repurchase intention (Szymanski and Henard, 2001). Bitner et al. (2002) stress that when the SST is viewed as a better solution in some way than the interpersonal alternative, the customer satisfaction can actually increase. Meuter et al. (2000) gained more insights in the relation between SST and satisfaction and studied reasons for satisfaction and dissatisfaction. They found three ways in which SSTs can that increase customer satisfaction. First, satisfaction arises when the SST solves an intensified or immediate need. Those needs are defined as “situations in which external environmental factors add a sense of urgency to the transaction” (p. 55). Translated to the topic of this research, self-scanning checkouts can help customers who are in a hurry, have utilitarian shopping goals, and don’t feel like interacting with a service employee. Therefore, self-scanning might increase satisfaction with the shopping experience for utilitarian people, since self-scanning can solve an immediate need and contributes in the achievement of their shopping goals.

(15)

14

Some customers prefer using SSTs over traditional service because they find it easy to use, or it helps them avoid interaction with employees (Dabholkar 1996; Meuter et al., 2000). However customers sometimes have to do some extra work with SSTs in order to benefit, this is offset by the fact that customers can save time and are able to control the service production process (Marzocchi and Zammit, 2007). Additionally, in the research of Meuter et al. (2000) customers mentioned that using the SST saved them time, and was experienced as more quickly or efficiently than an employee involved checkout. These arguments are very similar to the utilitarian motivation for shopping, where customers shop to fulfill a specific goal or need and prefer to do this in an efficient manner.

The third category for satisfaction of Meuter et al. (2000) is called “it did its job”, and emphasizes that many customers are still fascinated by technologies and are surprised that it really works. Bateson (1985) stresses that customers participate in SSTs because they take a pleasure in doing so. Arguments like saving time or saving money, are irrelevant for those customers. These are arguments that point to amazement, fun and enjoyment as drivers to become satisfied, and fit with the characteristics of a hedonic shopping motivation.

Prior research shows that usage of SSTs can lead to more satisfaction. Literature suggests that customers’ shopping motivations can influence the usage and satisfaction (Collier and Sherrell, 2010). Therefore, this research also focuses on influences of the shopping motivations on customer satisfaction with the shopping experience. It seems that both shopping motivations can become more satisfied with the shopping experience by using self-scanning, although customers with a utilitarian motivation might be slightly more satisfied than customers with a hedonic shopping motivation, since it can improve the accomplishment of their shopping goals. Therefore:

(16)

15

The formulated hypotheses lead to the following conceptual model:

Shopping Motivations S S T c h a ra ct e ri st ic s Shopping Motivations Usage of Self-Scanning Technology Satisfaction with shopping experience Hedonic Shopping Motivation Utilitarian Shopping Motivation Perceived Fun Perceived Ease of Use Perceived Usefulness Perceived Control Hedonic Shopping Motivation Utilitarian Shopping Motivation H1a t/m H1d H1a t/m H1d H2a t/m H2d H2a t/m H2d H3H3 Conceptual Model

3. Methodology

The previous chapter summarized prior literature, and showed the hypotheses and the conceptual model of this research. To test the conceptual model, an empirical research is conducted. This chapter will discuss how the research will be performed and how the data will be analyzed. First, the research design and data collection method will be explained, followed by the questionnaire design and the measurement of variables. Finally, the plan of analysis will be presented.

3.1 Research Design

(17)

16

will be spread via multiple channels, like e-mail, Facebook and LinkedIn. Only people who have ever visited a store of IKEA can participate in this research.

3.2 Questionnaire design

The questionnaire will start with a short introduction to thank respondents for their participation, to explain what is expected from respondents and to provide information about the time it will take to fill in the questionnaire. Then the questionnaire starts, and will consist of five sections. The first section will contain control questions regarding respondents’ gender, age, education, and their prior visits to IKEA. In case respondents have never visited IKEA, the questionnaire will close and respondents will be thanked for their participation. Respondents who did visit the IKEA before, will all be exposed to the same questionnaire from that moment on. The second section will measure satisfaction with the latest shopping experience at IKEA, which is related to the final part of the conceptual model. Four statements will be presented and respondents are asked to judge these statements on a range from totally disagree to totally agree. The third section will consist of statements about people’s perception of four SST characteristics, which relate to the independent variables of the conceptual model. Two pictures of a self-scanning checkout desk at IKEA will be shown, to make sure respondents have the right technology in mind when judging the statements about the SST characteristics. Next, in the fourth section, participants have to judge statements about the moderators of the model: their hedonic and utilitarian shopping motivations. Finally, the questionnaire will conclude with questions about their previous usage of self-scanning checkout desks at IKEA and their intentions to use self-scanning during their next visit. After closing the questionnaire, respondents will be thanked for their participation. The questionnaire is presented in Appendix 1.

3.3 Measurements

(18)

17 3.3.1 Measuring Satisfaction

Much research has been done into the measurement of satisfaction. Maxham and Netemeyer (2002) developed a three-item measurement scale for measuring overall store satisfaction. However, those statements are very similar to each other. To prevent respondents from becoming inattentive, caused by the presentation of comparable questions, the questionnaire will be adapted with items from Collier and Sherrell (2010) and Liu (2012). Respondents are asked to respond to the items on a 5-point Likert scale, ranging from totally disagree to totally agree.

3.3.2 Measuring perceived SST characteristics

To measure the perceived SST characteristics, the scales of Childers et al. (2002), Weijters et al. (2007) and Collier and Sherrell (2010) will be used. For those scales is chosen, because they are developed for measuring self-service in retail settings (Weijters et al., 2007) or focus specifically on self-scanning technologies (Collier and Sherrell, 2010). The items will be adapted to the self-scanning situation at IKEA, therefore, some irrelevant questions will be excluded from the questionnaire.

3.3.3 Measuring hedonic and utilitarian motivations

Hedonic and utilitarian shopping values are studied several times in prior research. The scales of Babin et al. (1994) and Griffin et al. (2000) are frequently used today and found much academic support. The statements are applicable to the shopping situation of IKEA. For this research however, an adaption will be made since the scales measure hedonic and utilitarian shopping values instead of motivations. Shopping values are being measured in the past tense, and ask customers -for example- what they liked about their shopping trip. This research will measure shopping motivations, which focus on why the customers went to IKEA in first instance. For this research, shopping motivations will provide more insights into customers general shopping preferences than shopping values would do. The original items of the measurement scales were adapted to the situation of IKEA and translated into Dutch.

3.3.4 Measuring the usage and usage intentions of self-scanning

(19)

18

likelihood that they will use self-scanning at IKEA in the future. This will be measured on a 5-point Likert scale.

Variable Type

Self-scanning usage Binair (yes= 1, no = 2)

Usage intention 5-point Likert scale

SST characteristics 5-point Likert scale

Satisfaction 5-point Likert scale

Shopping motivations 5-point Likert scale

Table 1: types of variables used in the research

3.3.5 Control variables

Control question will be included in the research. Those questions provide insights in the background of the sample and make it possible to compare the sample of this research to the population of the Netherlands. Additionally, the variables give more information about the effects in the model. The control variables are presented in Appendix 1, the questionnaire.

3.4 Plan of analysis

The plan of analysis consist of three parts. First the demographic analysis will be conducted, which consists of a description of the sample based on the control variables, and the comparison with the Dutch population. The second part contains basic insights into the data, such as checking the reliability of the constructs and looking at the correlations between variables. In the last part, the hypotheses are tested by logistic and linear regression analyses.

3.4.1 Demographic analysis

(20)

19

3.4.2 Basic insights

To gain basic insights, new variables need to be created first. Multiple items were used to measure the perceptions of SST characteristics, the hedonic and utilitarian motivations, and satisfaction. Although the measurement scales in this research are developed and tested by other academics, some questions will be combined or adapted to the situation of IKEA and the questions will be translated into Dutch (which might cause interpretation differences). Therefore, a factor analysis will be performed on the items of the SST characteristics and shopping motivations, to check the underlying constructs and to see whether those items can be combined into new factors or not. Furthermore, the items will be tested on internal consistency, to measure the strength to proceed with these variables instead of the original variables. A reliability analysis is performed and tested with Cronbach’s Alpha. As a rule of thumb, an alpha of 0.6 or higher indicates sufficient internal consistency between items and we can be confident that the items measure the same construct (Malhotra, 1999). Next, new variables will be computed based on the factor scores of SST characteristics and shopping motivations. To compute a new variable from the items of satisfaction, the standardized values will be saved. Likewise, the values of usage intentions will be standardized. Additionally, the averages of the items are saved into a new variable, to use in for the descriptive statistics and the multicollinearity analysis. Dummies will be created for the variable self-scanning usage and for the control variables education and gender. Furthermore, the descriptive statistics of the variables and a correlation table will be presented in order to gain pre-insights into the relations between the variables.

3.4.3 Testing hypotheses

(21)

20

characteristics and usage. The moderators will be computed by multiplying the independent variables with the moderating variables. Third, a linear regression analysis will be executed to test hypothesis H3, which looks at the moderating effect of the shopping motivations on satisfaction. Also the main effect of self-scanning usage on satisfaction will be explored.

The same procedure will be executed for the variable ‘usage intention’, to see whether the results show interesting differences or not. However, since usage intention was measured on an interval scale, a linear regression instead of a logistic regression will be conducted to test H1a to H1d, and H2a to H2d. Finally, both models will be tested for mediation by looking at the significances and coefficients of path A (which will be already known from prior analyses), and of path B and C that will be discovered with additional regression analyzes.

4. Results

This chapter provides findings of the quantitative research that has been performed. Descriptives of the dataset are summarized to see if the sample correspondents to the Dutch population. Factor analyzes are conducted to discover whether items can be combined into new variables. Reliability analyzes are executed to test internal consistency of the constructs of the conceptual model. Finally, the hypotheses are tested by using three regression models.

4.1 Descriptive analysis

In total 335 respondents started the survey, of which 75 did not finish the survey. Since the dropouts were not remarkably high at a specific question, which could be a sign that there is a reason for not answering the question, these respondents were removed from the dataset. Two people have never visited IKEA, thus the questionnaire closed automatically for them. Three respondents did not complete the final three questions about their usage of self-scanning, which are important for the analyzes. Therefore, those respondents were removed. One respondent indicated she was six years old, but this seems to be a typo since her answer distribution was normal. Therefore, this answer was reported as missing value and replaced by the mean.

(22)

21

education level. Those respondents did answer all the other questions of the survey. Therefore, mean substitution is applied for those missing values and those respondents are included in the sample.

(23)

22

Demographic variable CBS, 2014 Sample (N=257)

Gender Male (%) 49.5 42.0 Female (%) 50.5 55.6 Age < 20 (%) 22.9 1.6 20-39 (%) 24.5 80.5 40-64 (%) 35.3 13.2 65 > (%) 17.4 4.7 Education (CBS, 2012) Mavo (%) 22.3 2.7 Havo/Vwo (%) 10.6 10.5 MBO (%) 29.8 11.7

HBO (%) 28.3 (HBO + WO) 37.5

WO (%) 36.2

Other (%) 9.0 0.8

Table 2 : Demographic descriptives

In order to gain insight into respondents’ experience with self-scanning, they were asked if they ever used self-scanning at IKEA and if they used self-scanning the last time they visited IKEA. Figure 1 shows the distribution. 63% of the female respondents had used self-scanning before, against 49% of the male respondents.

Figure 1: prior usage of self-scanning at IKEA of males and females

49% 63%

51% 37%

Male Female

Have you used selfscanning at IKEA before?

(24)

23

4.2 Basic analysis and pre-insights

Before testing the hypotheses, basic analysis of the data is performed and the variables are restructured. Additionally, pre-insights of the data are provided. First, a manipulation check will be presented to evaluate how participants responded on the manipulation of two

questions. Second, a factor analyses is performed to determine if items can be combined into a new variable. Third, the descriptive statistics and the correlations of the variables are shown in order to gain pre-insights into the relations between variables.

4.2.1 Manipulation check

Two pairs of questions with contrary meaning were included in the questionnaire, to check whether respondents were paying attention or not. The questions did not contain a literal contradiction, but when a respondent agrees on one of the questions it is unlikely that he/she would also agree with the reversed question. To check whether respondents did agree or totally agree on both questions, two cross tabulations are presented (tables 3 and 4). In table 3, four respondents agreed on both questions. Those respondents were checked in the dataset, but they did not provide other extreme answers. As shown in table 4, a total of 19 respondents agreed/totally agreed on both questions. Also those respondents were checked in the dataset, but no other unlikely answers were found. An explanation could be that respondents did not read the denial in one of the questions, and read it as a confirmatory question. Those respondents are included in the sample, since they showed a normal pattern on the other questions of the survey.

Afrekenen met zelf-scannen lijkt me vermakelijk Afrekenen met zelf-scannen

lijkt me saai

Agree Totally agree

Agree 4 0

Totally agree 0 0

Table 3: cross tabulation of respondents’ confirming contrary questions

Ik zie winkelen bij IKEA niet als prettige vrijetijdbesteding Ik zie winkelen bij IKEA

als een waar plezier

Agree Totally agree

Agree 15 0

Totally agree 3 1

(25)

24 4.2.2 Recoded items

Prior to performing the factor analysis, the reversed questions were recoded. The survey contained three reversed questions, which were measured on a 5-point Likert scale. Those items are “self-scanning would be uncomfortable”, “checking out with self-scanning seems boring to me”, and “I don’t see shopping at IKEA as pleasant leisure”. Those items need to be recoded into the opposite direction, to avoid problems in the factor analysis. The values 5 (totally agree) were recoded into values of 1, the values 4 (agree) was recoded into values of 2, and so forth.

4.2.3 Factor analysis and reliability analysis

A factor analysis is performed on two constructs of the conceptual model, SST characteristics and shopping motivations. The factor analysis is conducted for the 19 items of SST characteristics, of which four underlying variables are expected, and the 10 original items of shopping motivations, of which two underlying variables are expected. Satisfaction was measured with four items and no underlying constructs are expected. Therefore, a reliability analysis with Cronbach’s Alpha will be executed to check whether those items are internally consistent and can be combined into a new variable. Additionally, a reliability analysis will be performed for the results of the factor analysis. Cronbach’s Alpha should be above 0.6 to combine the items into new variables.

Factor- and reliability analysis for self-scanning characteristics

Two factor analyzes were performed on the self-scanning characteristics. First, the 19 items were put into an explanatory factor analysis, to check if the analysis would show a four factor solution. The results are shown in appendix II. The Bartlett’s test of Sphericity should be significant and the KMO should be above 0.5 in order to consider the factor analysis as appropriate. This was both the case (p < 0.01, KMO 0.913). The communalities were above 0.4, which is the rule of thumb. However, when looking at the rotated component matrix, some items loaded on different factors than would make logically sense. Although Cronbach’s Alpha is greater than 0.6, the items do not logically fit within the factors. Therefore, four items were removed and the factor analysis was redone (those items are marked bold and italic in appendix II).

(26)

25

fourth factor became below a value of 1 (Ev= 0.973), it still had 5% of variance for each factor and the cumulative percentage was 72.68% for the four factors. Finally, the Cronbach’s Alpha’s were checked. The first factor ‘perceived ease of use’, has an alpha of 0.886 for the five original items. The second factor, ‘perceived usefulness’, has an alpha of .815 for the three remaining items. The third factor, ‘perceived enjoyment’ has an alpha of 0.846 for the four remaining items. The final factor ‘perceived control’ has an alpha of 0.798, for the three remaining items. After performing this second factor analysis, the items fit statistically and logically into the constructs that are measured.

Factor analysis perceived SST characteristics

Component Cronbach

’s Alpha

1 2 3 4

Perceived Ease Of Use

1. Zelfscannen zal duidelijk en begrijpelijk voor mij zijn .791 .033 .089 .020 .884 2. Met zelfscannen zal ik niet veel hoeven nadenken .788 -.037 .067 .116

3. Zelfscannen zal makkelijk zijn in gebruik .816 .254 .127 .213 4. Zelfscannen zal me weinig moeite kosten .829 .192 .079 .242 5. Zelfscannen zal gebruiksvriendelijk zijn .711 .197 .177 .367

Perceived Usefulness

6. Zelfscannen zal me helpen om sneller te kunnen winkelen.

.311 .302 .207 .679 .815

7. Zelfscannen verlaagd de wachttijd bij de kassa .130 .081 .168 .874 8. Ik zou afrekenen met zelfscannen nuttig vinden .457 .292 .197 .640

Perceived Enjoyment

9. Afrekenen met zelfscannen lijkt me leuk .359 .715 .105 .287 .846 10. Afrekenen met zelfscannen zou me een goed gevoel

geven

.314 .646 .242 .337

11. Afrekenen met zelfscannen lijkt me vermakelijk .000 .850 .202 .004 12. Afrekenen met zelfscannen lijkt me interessant .031 .822 .192 .154

Perceived Control

13. Afrekenen met zelfscannen zou me een gevoel van controle geven

.113 .229 .806 .242 .798

14. Afrekenen met zelfscannen geeft klanten de leiding .114 .186 .799 .018 15. Afrekenen met zelfscannen zou me meer controle

geven dan afrekenen bij een bemande kassa

.128 .148 .787 .222

(27)

26

Factor- and reliability analysis for shopping Motivations

(28)

27

Factor analysis Shopping Motivations

Component Cron- bach’s Alpha

1 2

Hedonic Motivation

1. Ik zie winkelen bij IKEA als een waar plezier .847 -.184 .890 2. Ik ga meestal winkelen bij IKEA omdat ik dat graag wil,

niet omdat het moet

.746 -.205 3. Vergeleken met andere dingen die ik kan doen, vind ik tijd

die ik besteed aan winkelen bij IKEA erg vermakelijk

.840 -.200

4. Ik geniet van het winkelen bij IKEA zelf, en niet alleen voor de producten die ik misschien koop

.839 -.226

5. Tijdens winkelen bij IKEA voel ik me opgewonden over de jacht naar nieuwe producten

.711 -.160 6. Tijdens het winkelen bij IKEA krijg ik een gevoel van

avontuur

.716 -.097

Utilitarian Motivation Pearson

Corre-lation 7. Tijdens het winkelen bij IKEA, koop ik alleen wat ik van

plan was om te kopen

-.148 .860 .664 .498 .000*** 8. Tijdens het winkelen bij IKEA, probeer ik alleen de

producten te vinden waar ik naar op zoek was

-.231 .826

***Significant on a 1% level

Table 6: Results of the final factor analysis of shopping motivation, including Cronbach’s Alpha and Pearson Correlation

Reliability analysis for satisfaction

A reliability analysis with Cronbach’s Alpha was executed for the third construct “satisfaction with the shopping experience”, which was measured on a four-item scale. Cronbach’s Alpha was 0.880 and could not be improved by removing an item (appendix II). Therefore, the four items can be combined into a new variable.

4.2.4 Computing new variables

(29)

28

compute new variables for satisfaction and usage intention, the Z-scores are saved. Additionally, the average of the items are combined into a new variable. The reason for this is that the averages are used to gain insight into the means and to check for multicollinearity, while the factor scores are needed for the regression analysis (especially required for the moderation analysis). In order to use the dichotomous and categorical variables in further analysis, dummy variables were created for the variables gender, education, and usage of self-scanning at a previous visit.

4.2.5 Descriptive statistics

The descriptive statistics of the new variables are shown in table 7 and provide an overview of the means and standard deviations of the new variables. The means of the perceptions of self-scanning characteristics show that ease of use (M= 3.86, sd= 0.74) and usefulness (M= 3.69, sd= 0.85) were rated as highest values within the perceptions of self-scanning, higher than the perceptions of enjoyment (M= 3.14, sd= 0.83) and control (M=3.11, sd= 0.86). Satisfaction has the highest mean, M= 3.93, sd= 0.70. Usage intention consists of one item, which explains the higher standard deviation (M= 3.68, sd= 1.22).

Variable Number of items Mean SD Ease of use 5 3.86 .74 Usefulness 3 3.69 .85 Enjoyment 4 3.14 .83 Control 3 3.11 .86 Hedonic 6 2.93 .88 Utilitarian 2 2.81 .90 Satisfaction 4 3.93 .70 Usage Intention 1 3.68 1.22

Table 7: descriptive statistics (N=257)

4.2.6 Correlations between the variables

(30)

29

hedonic shopping motivation was positively and significant correlated to perceptions of usefulness (r=0.194, p <0 .01), enjoyment (r=0.383, p < 0.01) and control (r=0.221, p < 0.05). Furthermore, the table shows that an increase in one’s hedonic motivation is significantly related to a decrease in one’s utilitarian motivation (r= -0.432, p < 0.01). There was a positive significant relation between the intention to use self-scanning and the perceptions of ease of use, usefulness, enjoyment and control. Additionally, usage intentions had a negative significant correlation with the utilitarian motivation (r= -0.136, p < 0.05). The table shows that an increase of usage intentions is significantly related to an increase of actual usage (r=0.484, p <0.01). The variable satisfaction is positively and significantly related with perceptions of ease of use (r=0.204, p < 0.01), usefulness (r=0.235, p < 0.01) and enjoyment (r= 0.195, p <0.01). Furthermore, there was a negative significant relation between satisfaction and the utilitarian motivation (r= -0.161, p < 0.01) and a positive significant relation between satisfaction and the hedonic motivation (r= 0.363, p < 0.01). Hence, an increase in the utilitarian motivation is related to a decrease in satisfaction, while an increase in the hedonic motivation is related to an increase in satisfaction. There was a positive significant relation between usage intentions and satisfaction (r= 0.181, p < 0.05).

(31)

30

4.3 Testing hypotheses

In this section the hypotheses are tested. Starting with testing for multicollinearity by looking at the VIF statistics and tolerance of the independent variables. Then, the main effects of the perceived SST characteristics on the self-scanning usage (H1a to H1d) will be tested by a logistic regression. Next, the moderators (the shopping motivations) will be included in the logistic regression analysis order to test hypotheses H2a to H2d. Subsequently, the second part of the model will be tested with a linear regression analysis to test H3. Finally, the model will be tested with the variable ‘usage intentions’ and a mediation analysis will be conducted.

4.3.1 Multicollinearity

The predicting variables of the logistic regression are tested on multicollinearity. Multicollinearity is present when a high correlation is found between two or more predicting variables. This can lead to incorrect results of the analysis. For this research, the VIF statistics should not be greater than 5 and the tolerance should not be smaller than 0.20. The multicollinearity analysis is conducted on the averages of the original variables, not on the factor scores. Using the factor scores would lower the VIF statistics, while this is not representative. The results are shown in table 9. All the variables have a VIF score of < 5 and a tolerance > 0.20. Therefore, no multicollinearity issues are found.

Variables VIF < 5 Tolerance > 0.20

Ease of Use _1.68 _.60 Usefulness _2.03 _.50 Fun _1.90 _.53 Control _1.47 _.68 Hedonic _1.52 _.66 Utilitarian _1.31 _.77 Education _1.12 _.90 Gender _1.24 _.81 Age _1.12 _.89

Table 9: multicollinearity statistics of the predicting variables

4.3.2 Main effects of SST characteristics

The first hypotheses that will be tested are the influences of perceived ease of use (H1a),

(32)

31

usage of self-scanning. Two models are consulted to test the main effects. Table 10 shows the

results. The first model consists of a logistic regression including the control variables, the perceived SST characteristics and the shopping motivations, to get insights in the main effects. The model is significant (p < 0.01) and has a Nagelkerk R² of 0.309, which means that the model has an explanatory power of 30.9%. Perceived ease of use has an odds ratio of 2.644 and is significant (p <.01). Perceived enjoyment also showed a positive significant relationship with self-scanning usage, OR= 1.862 and p <0.01. This means that both higher perceived ease of use and higher perceived enjoyment will lead to an increase in the usage of self-scanning. This leads to confirmation of hypotheses H1a and H1c. Perceived usefulness and perceived control also showed a positive relationship with self-scanning usage, but those relations were not significant. Therefore, hypotheses H1b and H1d are rejected. There was a negative significant relation between age and self-scanning usage (OR= -0.28, p < 0.05), which holds that the older people are, the less they make use of self-scanning.

4.3.3 Moderating effect of the hedonic and utilitarian motivation on SST usage

The main hypotheses of this research are about discovering the influence of shopping motivations and the perceptions of self-scanning on the actual usage of self-scanning checkout desks (H2a to H2d). To test those hypotheses, the interaction effects are included in a second logistic regression model. The results are shown in table 10. Also the control variables are added. The model is significant (p <0.01) and has a Nagelkerke R² of 0.358, which means that the model explains 35.8% of the variance. Perceived ease of use and enjoyment are still significant, which means those effects are strong enough to stay significant when the interaction variables are added to the model. The interaction effect of perceived enjoyment and the utilitarian shopping motivation is negatively significant on a 5% level (OR= 0.657, p= 0.011). This means that one who is utilitarian motivated to shop, is less likely to use self-scanning when the perceived enjoyment increases. The interaction effect of hedonic and enjoyment shows a positive relationship on self-scanning usage. This seems to support hypothesis H2c, the hedonic shopping motivation does strengthen the relation

between perceived enjoyment and usage of the self-scanning, more than the utilitarian shopping motivation does. However, the interaction effect of hedonic and enjoyment was not

(33)

32

Similarly, the interaction effect of utilitarian and ease of use showed a positive direction (OR=1.007), while hedonic and ease of use showed a negative direction (OR=0.995), as expected with H2a, the utilitarian shopping motivation does strengthen the relation between

perceived ease of use and usage of self-scanning, more than the hedonic shopping motivation does. Also those relationships were not found significant (p= 0.967, p= 0.976 respectively).

Therefore, hypothesis H2a is rejected. No significant results were found for H2b and H2d, which leads to rejection of those hypotheses.

(34)

33

Model 1 Model 2

Variable B Exp (B) Sig. B Exp (B) Sig.

Constant 1.196 3.307 .097 _1.149 _3.155 .132 Control Gender(0=female; 1=male) -.489 .613 .132 _-.489 _.613 .146 Age -.028 .973 .032** _-.026 _.975 .050** Education .059 1.061 .622 _.056 _1.057 .666 Independent Ease of Use _.972 _2.644 .000*** _1.102 _3.009 .000*** Usefulness _.101 _1.106 .503 _.088 _1.091 .589 Enjoyment _.622 _1.862 .000*** _.665 _1.944 .000*** Control _.026 _1.026 .864 .039 _1.040 .808 Moderators Hedonic _-.202 _.817 .230 _-.190 _.827 _.283 Utilitarian .035 _1.036 .823 _-.011 _.989 _.949 Hedonic*ease of use _-.005 _.995 _.976 Hedonic*usefulness _-.237 _.789 _.114 Hedonic*enjoyment _.085 _1.089 _.576 Hedonic*control _.063 _1.065 _.657 Utilitarian*ease of use .007 1.007 .967 Utilitarian*usefulness _-.040 _.961 _.820 Utilitarian*enjoyment _-.421 _.657 _.011** Utilitarian*control _-.234 _.792 _.134 Sig. .000 .000 Cox&Snell R-square .230 .267 Nagelkerk R-square .309 .358 *** Significant at a 1% level ** Significant at a 5% level * Significant at a 10% level

(35)

34

4.3.4 Main effect of self-scanning usage on satisfaction

Hypothesis 3 states that the utilitarian shopping motivation will have a stronger influence on

the relationship between SST usage intention and satisfaction, than the hedonic shopping motivation will have. Before testing this hypothesis, the pure main effect of self-scanning

usage is regressed on satisfaction. In regression model 1, shown in table 11, the control variables, the dummy of self-scanning usage and the shopping motivations are added as independent variables and satisfaction as the dependent variable. The model is significant (F= 7.26, p <0 .01) and explains a variance of 13% (R²= 0.13). The relationship between self-scanning usage and satisfaction is not significant (B= 0.10, p= 0.118), meaning that usage of self-scanning checkouts is not expected to significantly increase satisfaction with the shopping experience. The regression of the hedonic shopping motivation on satisfaction shows a positive significant relationship (B=0.37, p < 0.01). This means that customers who have higher hedonic motivations, are also more satisfied with the shopping experience.

4.3.5 Moderating effect of the hedonic and utilitarian motivation on satisfaction

One of the main hypotheses of this research states that the utilitarian shopping motivation has

a stronger influence on the relationship between self-scanning usage and satisfaction, than the hedonic shopping motivation has (H3). In order to test this final hypothesis a second

regression model is executed, including the interaction effects of the shopping motivations. The results are shown in table 11. The model is significant (F=5.613, p < 0.01) and has a R² of 0.129. The interaction effects of the utilitarian motivation and self-scanning usage (B=0.045, p =0.616) and the hedonic motivation and SST usage (B=0.090, p= 0.276) were not significant. This leads to the rejection of hypothesis H3.

(36)

35

Variables _Beta Sig. Beta Sig.

Constant .097 .113

Control

Age .097 .118 .094 .129

Gender(0=female; 1=male) .008 .905 .000 .998

Education .040 .523 .040 .522 Independent SST usage .095 .118 .093 .128 Hedonic motivation .370 .000*** .332 .000*** Utilitarian motivation -.095 .119 -.157 .060* Moderators Usage*Hedonic .090 .276 Usage*utilitarian .045 .616 Model Sign. .000 .000 F-value 7.259 5.613 R² .151 .157 Adjusted R² .131 .129 *** Significant at a 1% level ** Significant at a 5% level * Significant at a 10% level

Table 11: regression analysis with satisfaction as dependent

4.3.6 Comparing the results to customers usage intentions

(37)

36

In model 2 the interaction effects were included. In this model, the perceptions of SST characteristics are still significant, and the hedonic motivation has stayed negatively significant to usage intentions on a 5% significance level (B= -0.101, p <0.05). The interaction of hedonic motivation and perceived usefulness showed a negative significant relationship on a 10% significance level (B= -0.074, p < 0.10). This means that an increase in one’s hedonic shopping motivation is related to a decrease in one’s perception of usefulness. The more hedonic motivated people are, the less does perceived usefulness influence their usage intention.

Variables Beta Sig. Beta Sig.

Constant _.532 _.525 Control Gender(0=female; 1=male) -.052 .239 -.069 .130 Age .046 .272 .050 .238 Education .012 .768 .016 .701 Independent Ease of Use _.524 _.000*** _.534 _.000*** Usefulness _.373 _.000*** _.391 _.000*** Enjoyment _.396 _.000*** _.243 _.000*** Control _.248 _.000*** _.361 _.000*** Moderators Hedonic _-.095 _.037** _-.101 _.027** Utilitarian _-.024 _.562 _-.019 _.654

Hedonic * ease of use _-.019 _.668

Hedonic * usefulness _-.074 _.087*

Hedonic * enjoyment _.025 _.560

Hedonic * control _-.040 _.337

Utilitarian * ease of use _.040 _.362

Utilitarian * usefulness _.050 _.232 Utilitarian * enjoyment _-.031 _.472 Utilitarian * control _-.043 _.333 Model Model Sign. .000 .000 F-value 43.080 23.251 R² .617 .629 Adjusted R² .602 .602 *** Significant at a 1% level ** Significant at a 5% level * Significant at a 10% level

(38)

37

An additional regression analysis is conducted to gain insight into the relationship of usage intentions and satisfaction, as shown in table 13. In model 1, that includes the control variables and the effects of the usage intentions on satisfaction, a positive significant relationship is found, B= 0.17, p < 0.01. This means that if one’s intentions to use self-scanning increase, one’s satisfaction with the shopping experience also increases. Although a significant relationship is found, the adjusted R² is 2.9%, which indicates the model does not have strong explanatory power. The model is significant p < 0.05. Model 2 includes the variables hedonic and utilitarian shopping motivations, and their interaction effects. The model is significant (p < 0.01) and has an R² of 14.1%. There was a positive significant direct relationship between the hedonic shopping motivation and satisfaction, B= 0.341, p < 0.01. This means that if one has higher hedonic shopping motivations, one’s satisfaction with the shopping experience increases. No significant moderating effects of hedonic or utilitarian motivations are found.

Variables _Beta Sig. Beta Sig.

Constant .754 .160 Control Gender(0=female, 1=male) -.098 _.128 _.012 _.858 Age .060 .355 .090 .141 Education .007 .916 .047 .447 Indepedent Usage intention .170 .008*** .153 .011** Hedonic motivation .365 .000*** Utilitarian motivation -.088 .156 Intention*Hedonic -.044 .466 Intention*utilitarian -.043 .485 Model Model Sign. _.023 _.000 F-value _2.893 _6.111 R² _.045 _.168 Adjusted R-square _.029 _.141 *** Significant at a 1% level ** Significant at a 5% level * Significant at a 10% level

(39)

38

4.3.7 Mediation

Finally, for both models (the main model with actual self-scanning usage and the additional model with usage intentions) mediation will be discussed. In case of full mediation, the effects of the SST characteristics on satisfaction would only appear through the mediator (self-scanning usage). With full mediation, the A-path and B-path should be significant, while the C-path is insignificant. Partial mediation is present when the A-path and the B-path are significant, as well as the C-path. Table 14 shows the results.

When actual usage is included in the model to test the relationships, path A, which tested the logistic regression of SST characteristics on the self-scanning usage, was significant for perceived ease of use and perceived enjoyment. The other perceptions, of usefulness and control, did not show a significant relation. The model had a Nagelkerke R² of 0.31 and was significant (p < 0.01). Path B tested the regression of self-scanning usage on satisfaction, while controlling for the perceptions of SST characteristics. The model had an R² of 0.069 and was significant (p< 0.01). Path B was not significant, B= -0.009, p = 0.890. As such, no partial mediation is found. Last, testing path C while controlling for the mediator self-scanning usage, three of the four variables of SST characteristics showed a significant relationship with satisfaction. Those perceptions are the ease of use (B=0.166, p < 0.05), usefulness (B= 0.157, p < 0.05), and enjoyment (B= 0.126, p <0.05). As such, no full or partial mediation is found in this model.

(40)

39

Actual usage Intentions to use

B Sig. Beta Sig.

Path A Logistic regression

Ease of use B (Exp) = 2.655 .000*** .526 .000***

Usefulness 1.101 .513 .366 .000*** Enjoyment 1.803 .000*** .372 .000*** Control .996 .976 .217 .000*** Path B -.009 .890 -.047 .623 Path C Ease of use .163 .008*** .163 .008*** Usefulness .156 .011** .156 .011** Enjoyment .124 .042** .124 .042** Control .047 .439 .047 .439 Partial/full mediation No No *** Significant at a 1% level ** Significant at a 5% level * Significant at a 10% level

Table 14: regression analysis with mediators actual usage and intentions to use

5. Discussion

In the introduction of this research the growth of self-scanning technologies in the Netherlands was discussed. This technology seems to pop out of the ground in several retail and service environments. Therefore, it is important to understand customers perceptions of self-scanning, which characteristics attract customers to the self-scanning checkouts, and if this attraction differs among customer shopping motivations. The following research question was formulated: “To what extent do hedonic and utilitarian shopping motivations influence

the relation between SST characteristics and self-scanning usage, and the relation between self-scanning usage and satisfaction with the shopping experience?” This research focused

(41)

40

Based on the theoretical framework, nine hypotheses were formulated. The expected relations were visualized in the conceptual model. Additionally, the results were compared to people’s usage intentions instead of their actual usage. The study showed that from the perceptions of the SST characteristics, ease of use (H1a) and enjoyment (H1c) had a positive influence on the actual usage of self-scanning. All the perceptions had a positive effect on the intentions to use. A negative interaction effect was found for the utilitarian motivation and perceived enjoyment, on self-scanning usage. Furthermore, a positive direction was found for the interaction between the hedonic motivation and perceived enjoyment, despite the fact that this relation was not significant. The findings of these interaction effects seem to support hypothesis H2c (the hedonic shopping motivations does strengthen the relation between

perceived enjoyment and self-scanning usage, more than the utilitarian shopping motivation does), although further research should be done to confirm these effects.

No significant interaction effects were found between the utilitarian motivation and the perceived ease of use (H2a), usefulness (H2b), or control (H2d) and self-scanning usage. Neither were significant interaction effects found for the hedonic motivations and ease of use, usefulness and control, and self-scanning usage. When the variable ‘usage intention’ was used as dependent variable, the directions of the interaction effects of H2b were as expected; the interaction of the hedonic motivation and usefulness showed a negative significant relationship with usage intentions, while the interaction effect of the utilitarian motivation and usefulness was positive, but not significant.

(42)

41 Hypothesis Usage/int ention Exp-ectation Significant relation? Result

H1a: Perceived ease of use  usage Usage + Positive

significant

Accepted Intention Positive

significant

Accepted H2a: The utilitarian shopping motivation

perceived ease of use & usage

Usage + Positive, not significant

Rejected Intention Positive, not

significant

Rejected

H1b: Perceived usefulness  usage Usage + Positive, not

significant

Rejected Intention Positive

significant

Accepted H2b: The utilitarian shopping motivation 

perceived usefulness & usage

Usage + No Rejected

Intention Positive, not significant

Rejected

H1c: Perceived enjoyment  usage Usage + Positive

significant

Accepted Intention Positive

significant

Accepted H2c: The hedonic shopping motivation 

perceived enjoyment & usage

Rejected Intention Positive, not

significant

Rejected

H1d: Perceived control  usage Usage + Positive, not

significant

Rejected Intention Positive

significant

Accepted H2d: The utilitarian shopping motivation 

perceived control & usage

Usage + No Rejected Intention No Rejected H3: The utilitarian shopping motivation 

usage & satisfaction

Rejected Intention No Rejected

Additional findings

The utilitarian motivation has a negative influence on perception of enjoyment and the SST usage The hedonic motivation has a negative influence on perceptions of usefulness and intentions to use Age has a negative relationship with the usage of self-scanning

The hedonic shopping motivation has a positive effect on satisfaction with the shopping experience The utilitarian shopping motivation has a negative effect on satisfaction with the shopping experience The hedonic shopping motivation is negatively related to intentions to use self-scanning

How Utilitarian And Hedonic Shopping Motivations Affect The Usage Of Self- Scanning Technologies

How Utilitarian And Hedonic Shopping

Motivations Affect The Usage Of

Self-Scanning Technologies

And customers’ satisfaction with the shopping experience

HOW UTILITARIAN AND HEDONIC SHOPPING MOTIVATIONS

AFFECT THE USAGE OF SELF-SCANNING TECHNOLOGIES

And customers’ satisfaction with the shopping experience

Management summary

Preface

Table of Contents

1. Introduction

2. Theoretical Framework

3. Methodology

4. Results

5. Discussion