Which Employees are the best?

Which Employees are the best?

“An analysis of employee characteristics”

Steven van Barneveld

S 1737805

Faculty of Economics and Business, University of Groningen, Master Thesis BA Marketing Management, September 2010. First supervisor: Wander Jager Second supervisor: Hans Berger

“An analysis of employee characteristics” ₂ Management Summary

The Do-it-yourself (D.I.Y.)-market in the Netherlands is infamous for their

inadequate employees. In order to see what factors determine the interaction quality of Dutch customers in a D.I.Y.-store, this research is done.

After a extensive literature review, based on the SERVQUAL tool of Parasuraman et. al. (1988) and other service measurement tools, a conceptual framework is created which is based on four constructs, which determines the interaction quality between customer and employee. These four constructs are: non-verbal communication and verbal communication, knowledge constructs and employee characteristics (as can be seen in the conceptual framework on page 29).

A questionnaire was filled in by 194 respondents and after deleting the incomplete questionnaires, 157 were used for data collecting. The total dataset consisted of customers of D.I.Y.-stores were evenly distributed regarding age and gender. The questionnaire can be found in appendix 1.

The data is tested using regression analyses, to determine if there are relationships between the four constructs and interaction quality. Furthermore, the relationship between interaction quality and customer satisfaction is measured.

“An analysis of employee characteristics” ₃ Preface

The importance of choosing the right master thesis topic is vital for a good progress in your master thesis process. This was something I knew before starting this thesis. Therefore, I started to think about a topic within my own world of interest. Combining marketing management with something happening in my own life.

Quickly I ended up with my work in a do-it-yourself store. I was working, helping a female customer with buying garden furniture. An hour later, she appeared at the central desk with a small gift, as a thanks for my superb assistance. That’s when my idea started to develop

Customer satisfaction has been a problem in the do-it-yourself market for years. The general image of an employee is that of a non-technical 17-year old student, which couldn’t be bothered with the wishes of the customer. The organization were I work even has the lowest score regarding employee satisfaction in the entire Dutch do-it-yourself market. Still, they are market leader with a massive gap between them and the second largest company. The questions arise; Are good employees even remotely important? Which characteristics are most important? And does the perfect employee even exist? These questions were the basis for this master thesis.

“An analysis of employee characteristics” ₄ Table of Content

“An analysis of employee characteristics” ₅ 4.3 Internal consistency ____________________________________________ 34 4.3.1 Factor analysis ______________________________________________ 34 4.3.2 Four constructs ______________________________________________ 37 4.3.3 Choosing groups _____________________________________________ 38 4.3.4 Sub constructs ______________________________________________ 40 4.3.5 Correlation__________________________________________________ 41 4.4 Hypotheses ___________________________________________________ 42 4.5 Customer Satisfaction ___________________________________________ 44 4.5 Customer Satisfaction ___________________________________________ 44 4.6 Other findings _________________________________________________ 44 Chapter 5: Conclusions and Managerial Implications ________________________ 47 5.1 Conclusions ___________________________________________________ 47 5.2 Scientific contributions _____________________________________________ 49 5.3 Managerial implications _________________________________________ 49 Chapter 6: Research Limitations and Suggestions for Further Research _________ 51 6.1 Research limitations ____________________________________________ 51 6.2 Further research _______________________________________________ 52

Appendix ___________________________________________________________ 54

“An analysis of employee characteristics” ₆ Chapter 1: Introduction and Problem Statement

In this chapter, the introduction of this master thesis will be given. First, the background problem will be described followed by an introduction of the topic and literature. Subsequently the relevance of this thesis will be explained, followed by the central problem statement. Finally, the organization of the rest of this thesis will be described.

1.1 Background problem

In the past few years, Gamma, which is the market leader in the Do-It Yourself market (D.I.Y.-market) in the Netherlands, received bad publicity regarding the service quality of their employees (Consumentenbond, 2009). While Gamma’s success lies in their low prices, good locations and extended marketing campaigns, their employees lacked in service quality and their technical knowledge was inadequate (Gamma, 2009). The lack of qualitative personnel was so commonly known that even companies in other sectors used the bad image of D.I.Y.-stores in their commercials1

1

The author of this thesis, working in a Gamma store as an additional job, wondered how important his influence and that of his co-workers was on the satisfaction of the consumers in that store. Furthermore, he wanted to know what characteristics are most important to achieve consumer satisfaction. Therefore, this thesis will examine the overall interaction quality of personnel in the D.I.Y.-sector. The interaction quality of personnel is divided into four categories; non-verbal communication, verbal communication, knowledge of the personnel (expertise) and the employees’ characteristics. Interaction quality influences satisfaction as will be described in the next chapter.

1.2 Introduction of topic

Consumer satisfaction is a very important competitive advantage for retail stores. “When consumers are unsatisfied 6% contacted the retailer, but 31% told someone else what happened” (Levy & Weitz, 2009).

The three consecutive marketing definitions service quality, consumer satisfaction and consumer loyalty, therefore, are important for successful retail businesses. The

“An analysis of employee characteristics” ₇ three definitions have been linked in several studies (Davis-Sramek et. al., 2009; Edwards, 2010; and Gustafsson et. al, 2005) were high service quality results in higher consumer satisfaction, which leads to more consumer loyalty. Although there are different factors influencing consumer satisfaction such as; promotional activity, environmental cues and product quality, in this master thesis only the effects of employees on consumer satisfaction will be researched. Research done in this field so far examines all factors influencing consumer satisfaction on a general level or, are incomplete regarding the effects of employees. Furthermore, the author of this thesis wants to study his effect on consumer satisfaction based on his own experiences in a D.I.Y.-store.

1.3 Literature introduction

In this paragraph, a short introduction regarding the literature used will be described. Quality

To examine the quality of the employees of a D.I.Y.-Store, a definition of “quality” is necessary. According to Lovelock & Wirtz (2007), quality can be divided into five different levels. The transcendent view, product-based approach, user-based definitions, manufacturing based approach and value-based definitions. The most interesting level is the level of user-based definitions, which starts with the premise that; “quality lies in the eyes of the beholder …. This subjective, demand-oriented perspective recognizes that different consumers have different wants and needs” (Lovelock & Wirtz, 2007). Since this thesis considers only the quality of the service given by employees in a D.I.Y.-store, quality in a sense of product-quality (focused on the product), manufacturing quality (focused on the quality of the production process), transcendent view (often regarding visual arts) or value based definitions (getting the lowest prices) are not relevant. A user-based definition level looks at the perceived quality of the consumers. The highest quality is quality that satisfies consumers in the best way possible, without compromises.

“An analysis of employee characteristics” ₈ Service quality

Service quality is a widely researched marketing concept which has many perspectives. The paper of Ueltschy et. al. (2007) for example, describes service quality as “the outcome of an evaluation process were the consumer compares his expectations with the service he perceived he has received”. Lovelock & Wirtz (2007) sees service quality as a long term measurement: “service quality is a consumers’ long term, cognitive evaluation of a firms’ service delivery”. While Ueltschy looks at the gap between perceived and expected quality, Lovelock and Wirtz see service quality as a evaluation of the performance.

Service quality is measured in different ways. Parasuraman, et. al. (1988) sees service quality as an overall measure of quality in the form of an average score across five dimensions; tangibles, reliability, responsiveness, assurance and empathy. While providing a general idea of the level of service quality the effect of each of the dimensions is not provided. Furthermore, the research of Parasuraman et. al (1988) measured both expected and perceived quality, while other research shows service quality as a single item measure (Boulding et al., 1993; Cronin and Taylor, 1992;). They expected that consumers automatically compared their expectations with their perceptions and thought that measuring both would be redundant. Although this result explains the quality of the service better, measuring expectations and perceived quality results in a better diagnostic tool.

In this article, service quality is measured as a multidimensional variable, divided into different variables that influence consumer satisfaction. In the next chapter, service quality will be described more extensively.

Consumer satisfaction

For consumer satisfaction, also multiple definitions are available. Cronin & Taylor (1992), describe consumer satisfaction as a transaction specific measurement, stating that; “satisfaction is the level of what a consumer would expect”. Other researchers see consumer satisfaction as the overall evaluation of the purchase and experience of a consumer (Anderson et. al., 2004).

“An analysis of employee characteristics” ₉ met”. Levy and Weitz (2009) describe consumer satisfaction as “a post consumption evaluation of how well a store or product meets or exceeds consumer expectations”. In conclusion, consumer satisfaction measures how well different retail attributes like; store lay-out, products, employees and shopping experience meet the needs and expectations of the consumer.

Consumer loyalty

Consumer loyalty is described by Levy & Weitz (2009) as: ”Consumers that are committed to purchasing merchandise and services from the retailer and will resist the activities of competitors attempting to attract their patronage”.

The impact of consumer loyalty on company success is widely accepted in (marketing) literature (Heskett et al., 1997; Homburg et al., 2009; Paul et al., 2009). Especially the quality of the relationship between consumers and companies is an important variable for consumer loyalty (DeWulf et al., 2001).

Relationship service quality, consumer satisfaction and consumer loyalty

Davis-Sramek (2009) studied retail loyalty. They further confirmed the relationship between service quality, consumer satisfaction, and consumer loyalty, which will be described below. TSQ=Technical Service Quality LB= Loyalty Behavior CC= Calculative Commitment AC= Affective Commitment Consumer Satisfaction RSQ=Relational Service Quality

Figure 1: Link between service quality, consumer satisfaction and loyalty

“An analysis of employee characteristics” ₁₀ The research of Davis-Sramek (2009) explains relational service quality as the quality between consumer and employee. This is a main focus of this article. In this article relational service quality will be divided into verbal and non-verbal communication, and will be further described in the next chapter.

Consumer satisfaction is a mediator in the research of Davis-Sramek (2009), but will be the dependable variable in this thesis. In the research of Davis-Sramek (2009), both relational- as well as technical service quality had a positive effect on consumer satisfaction, were relational service quality had a stronger correlation with consumer satisfaction.

Davis-Sramek (2009) also researched the effect of consumer satisfaction on commitment and loyalty behaviour. They concluded that consumer satisfaction shows no significant effect on loyalty behaviour. To test if this is true in the D.I.Y-market, this research will also look at the effects of employees on customer loyalty. 1.4 Relevance of this thesis

Although the effects of service quality on consumer satisfaction is already researched (Brown & Lam, 2008; Cronin et. al, 2000; Morhart et. al., 2009;), the results are showing different outcomes. While Brown & Lam (2008) concluded that the relationship is significant between service quality and customer satisfaction, Morhart et al (2009), found that it was only significant when the personnel had undertaken specific training. Cronin et al (2000) showed that the relationship between service quality and customer satisfaction was non significant because customers had different interpretation of service quality. Furthermore, evidence is found that the effects are different across market segments. Therefore, this research will add more information regarding the D.I.Y.-sector.

“An analysis of employee characteristics” ₁₁ Taylor (1992) shows, that consumer satisfaction has a strong effect on purchase intention.

Nevertheless, some research show a neutral (e.g. Brown and Mitchell 1993) or even negative relationship (e.g. Silvestro and Cross 2000) between satisfied employees and consumer satisfaction. The main conclusion of Silvestro and Cross (2000) was, that “shaping” employees to be the most profitable does not work, because employees become dissatisfied from being shaped and that this dissatisfaction will negatively influence consumer satisfaction.

Since the author of this thesis works in a D.I.Y.-store, he wondered what the effects of employees were in this market. The fact that there are different outcomes in the marketing literature regarding the effects of employees on consumer satisfaction and that effects differ across markets provides the relevance for this thesis.

With regard to the influence of employees’ technical knowledge less information is known. Although Turley and Milliman (2000) examined 57 different variables divided into five different groups, including the effect of human variables such as employee characteristics and employee uniforms. (Technical) service quality was not explored. Therefore, this research will also focus on the influence of service quality of employees, technical knowledge of those employees (expertise) as well as the employee characteristics, examining the effect of employees in a broad perspective. The construct of technical knowledge is especially interesting, since less information is known about the effects of technical knowledge and the opinion of the author of this thesis is that technical knowledge can greatly determine customer satisfaction in the D.I.Y.-market.

“An analysis of employee characteristics” ₁₂ 1.5 Problem statement

Because the effects of employees in a D.I.Y.-retail store are not specifically studied, the question remains:

“To what extend do employees in the D.I.Y-market influence consumer satisfaction? And which employee characteristics have the highest relevance on interaction quality?”

Sub-questions

Subsequently to the problem statement, sub-questions can be made. These will be formulated in this paragraph.

 Which factors determine the interaction quality between employees and customers of a D.I.Y-store in the Netherlands?

 To what extend does non-verbal communication influence the interaction quality between employees and customers of a D.I.Y-store in the Netherlands?

 Which verbal communication constructs influence interaction quality between employees and customers of a D.I.Y-store in the Netherlands the most?

 To what extend does technical knowledge of the employee influence the interaction quality between employees and customers of a D.I.Y-store in the Netherlands?

 Which employee characteristics influence the interaction quality between employees and customers of a D.I.Y-store in the Netherlands?

 To what extend does the characteristics of the customer, influence the interaction quality between employees and customers of a D.I.Y-store in the Netherlands?

 What are the main determinants for customer satisfaction, regarding the Dutch D.I.Y.-market?

1.6 Organization of the Thesis

“An analysis of employee characteristics” ₁₃ Chapter 2: Literature Review

In this chapter, a literature study will be written, discussing the subjects relevant for this study. First, service quality and the characteristics of the employees will be discussed. Subsequently, the moderators will be described, in the form of consumer characteristics. Finally, hypotheses will be formulated and a conceptual framework will be developed and further explained.

2.1. Service Quality

Service quality first appeared in 1985, in an article of Parasuraman et al,. Although service quality is mentioned in other research, the article is viewed by many researchers as the fundament for service quality (measurement).

Service quality remained an elusive construct. Although there were numerous definitions and philosophies, no clear construct was available.

Parasuraman et. al., (1988) developed a multi-item scale called SERVQUAL which has been a standard in service marketing and a basis for other researchers (Ueltschy 2007, Jain & Gupta 2004). Other variations came, like the SERVPERF scale, which was not focused on the gap between expected and perceived quality but more on performance of the service (Cronin & Taylor, 1992). Different authors compared the two scales in different business sectors and with different variables. Still, there is an ongoing discussion about these two scales, which shows the complexity of these items.

SERVQUAL

“An analysis of employee characteristics” ₁₄ SERVPERF

The SERVPERF scale, developed by Cronin & Taylor (1992) is a direct response to the SERVQUAL scale. Their scale is also based on 22 items, but in comparison to the SERVQUAL scale, SERVPERF only measures performance while SERVQUAL measures the performance and the expectations of the consumers.

The results of Cronin & Taylor show that using a performance-only scale creates a more efficient measurement and has an empirically superior basis. It also explains more of the variation of service quality and is based on the idea that service quality is an attitude and not based on satisfaction. Another important result from the research is that, although most research (at that time) said that satisfaction was an antecedent of service quality, they analyzed that (perceived) service quality leads to consumer satisfaction. This had also been analyzed by Parasuraman et. al. (1985, 1988).

SERVQUAL vs. SERVPERF

Many other researchers supported the superiority of the SERVPERF scale (Boulding et al., 1993; Grewal and Brown, 1994 Hartline and Ferrell, 1996), still in practice, the SERVQUAL is more popular. Furthermore, recent meta-analysis research of Carrillat (2007) shows that both scales are equally valid, but that SERVQUAL is more useful for practitioners, because of its diagnostic value. While SERVPERF better explains the variance quality, SERVQUAL is a better diagnostic tool.

Modifications of SERVQUAL

“An analysis of employee characteristics” ₁₅ Figure 2: Six dimensions of Ma et, al. (2007).

The new structured SERVQUAL is based on three primary dimensions. The first dimension is tangibles, which covers the physical facilities and appearance of personnel. The second dimension is process, which covers how the service process is executed and the last dimension is outcome, which covers what a consumer receives during the service process. The results of the study show that this newly structured SERVQUAL was superior to the original model of Parasuraman et. al. (1988).

Other modifications of SERVQUAL

Besides the study of Ma et. al. (2007), several other authors modified the SERVQUAL tool. The study of Finn and Kayande (2004), examined these modifications and gathered them to examine the consequences. In total, they examined 34 studies who adapted the SERVQUAL by deleting or adding items, often for a specific market(s). They distinguished between scale adoption and scale refinement. Scale adoption refers to adding or deleting certain items based on their suitability for a certain research. Scale refinement refers to a change due to psychometric performance development. They concluded that:

“An analysis of employee characteristics” ₁₆ When scaling retailers, refinement is not advised, because of the lack of scale performance improvement. When SERVQUAL is adopted in this research, it is wise only to use scale adaptation. Furthermore, a close examination of the specific characteristics that needs to be measured should be done.

In conclusion, SERVQUAL has been an important measurement tool for service quality for many years. Although several researchers have criticized this tool, it is still one of the most used tools in the service world. Therefore, in this article the constructs of SERVQUAL, described in the previous paragraphs will be adopted in the conceptual framework. Nevertheless, other measurement tools can also be of interest for this thesis. Therefore, in the next paragraph, these tools will be discussed.

2.2 Other consumer satisfaction measurements

The effect employees have on the consumer satisfaction is diverse. As mentioned in the previous paragraphs, SERVQUAL is a largely accepted measurement tool. Nevertheless, other authors have made different measuring tools and other service categorization methods. In this paragraph, these measurement tools and methods are described, as can be seen in table 1.

Table 1: different service measurements.

Authors Research Independent Dependant Sample Most

Type* Variables Variables significant

Berry,

Wall and T functional clues service experience - - Carbone (2006) mechanic clues humanic clues Brown et. Al. (2002)

E "The big five" (personality psychology)

consumer

orientation 280 Agreeability, Openness

Saucier

(1994) Conscientious-ness Ekcini &

Davis E Based on interaction quality 320 for I.Q. all positive (2009) Brown et.al. consumer

satisfaction Pollack E attitude Service

Quality 550 Attitude, behavior

(2009) behavior consumer

satisfaction

expertise

“An analysis of employee characteristics” ₁₇ “Service clues”

The article of Berry, Wall and Carbone (2006), researched the effect of service clues: “consumers generally expect service companies to know their business and to keep their core service promises. However, small cues can signal the service is exceptional and have a disproportionately larger effect on how a consumer assesses their entire service experience”. The article subdivides these clues into functional, mechanic and humanic clues.

Functional clues can be described as the “what” of service experience. It covers everything that indicates or suggests the technical quality of a service. Material clues come from actual objects, for example, smells and sounds. Humanic clues emerge from the behaviour of the employees of the service provider. The humanic and mechanic clues are the “how” of the experience. Often these clues combine themselves in certain situations. When a salesperson in a D.I.Y.-store explains to a consumer that that specific product is out-of-stock, he or she addresses both functional as well as humanic clues.

“The big five”

This measurement tool is based on the work of researchers such as Costa and McCrae (1985), Goldberg (1992), and Saucier (1994), who, as psychologists, studied the basic five dimensions of personality. These dimensions are:

- Extroversion (or introversion), the extent to which a person is outgoing or shy. - (in) Stability, describes how emotionally steady a person is,

- Agreeability, does person express warmth and feelings toward others. - Conscientiousness, the degree of organization and precision of the person - Openness to experiences and imagination.

“An analysis of employee characteristics” ₁₈ Although this research is not a full service quality measurement it does show that people when they are working, all have different personality traits. These personality traits can influence the satisfaction of consumers. Instable employees will probably be less liked, while agreeable employees will give the consumer a feeling of warmth and cooperation.

Since personality traits are often difficult to see and measure for a consumer in a D.I.Y.-store, especially regarding the brief conversation between customer and employee, these traits cannot be used for this research. Extroversion however, is easily to define. Greeting customers and acknowledging that the customer enters or leaves the store are clear examples of extrovert behaviour. The SERVQUAL tool does not include these greetings into their measurement tool. Therefore, extroversion will be added.

In conclusion, the study confirmed that “the expertise behaviour and service manner of the frontline employee has a strong influence on consumer satisfaction” (Ekinci & Dawes (2009)). Furthermore, acknowledging the customer is not part of the SERVQUAL tool. Therefore the extroversion part of this construct is used in this research.

HSQM-Model

The research by Pollack (2009) is based on the service quality assessment tool, developed by Brady & Cronin (2001). Although this tool not only looks at employees but takes the physical quality and outcome quality into account as well, they do have an interesting way of measuring the interaction quality of service.

“An analysis of employee characteristics” ₁₉ Figure 3: HSQM model

The results of Pollack (2009) were that in different sectors different qualities are important, and that employees’ attitude towards a consumer seems to be the most important factor overall, while expertise is only important when the market sector asks for it. In a supermarket consumers do not expect employees to know everything about a certain product. In a specialty store, consumers expect that the employee has full knowledge of the products.

“An analysis of employee characteristics” ₂₀ 2.3 Employee characteristics

For this research, the ways of measuring the quality of service are very interesting. Nevertheless, these measurements usually cover other variables (environmental cues, product quality, presentation etc.) or are not fully applicable to the D.I.Y.-market specifically. Therefore, this thesis adopts certain aspects of these measurement tools and combines them with other variables to form its own measurement framework. The author decides which aspect(s) from which tool(s) can be used best for this research. These aspects are combined into four different groups; non-verbal communication, verbal communication, knowledge/expertise and employee characteristics. These groups are described more detailed in the next paragraphs.

2.3.1 Non-verbal communication constructs

The article of Gabbott & Hogg (2001) describes the role of non-verbal communication in service encounters. Although no explanation is needed about the existence of non-verbal communication, the effects of this type of communication on the satisfaction of consumers in a (D.I.Y.) store has to the knowledge of the author not been researched. The research of Gabbott & Hogg (2001) uses four types of non verbal communication: proxemics (using distance and personal space), kinesics (body posture), oculesics (eye behaviour) and vocalics (vocal tone and intonation). Gabbott & Hogg (2001) use these types to measure the interaction quality between consumer and employee, while interaction quality influences consumer satisfaction.

Although non-verbal communication is evidentially influencing the interaction between consumers and employees, it is not clear to what extend it influences consumer satisfaction in a D.I.Y.-store, therefore:

H1: Non-verbal communication has a positive effect on interaction quality. 2.3.2 Verbal communication constructs

“An analysis of employee characteristics” ₂₁ Extroversion is an important variable according to the author of this thesis. Extroversion creates an open conversation between employees and customers and employees who have this personality trait will be more likely to approach the consumer. Brown et.al. (2002), confirms this idea. Thus:

H2a: Extroversion has a positive effect on interaction quality.

In line with extroversion is the willingness to help. Is the employee always willing to help the customer? Even if he has other important work to do? And does this influence the interaction quality? These questions are covered by the variable responsiveness

H2b: Responsiveness positively affect interaction quality.

Subsequently, the empathy of the employee is important. Does the employee not only listens to the consumer, but also reserve time for him/her? The research of Parasuraman (1988) shows that empathy, when correctly applied, can highly contribute to consumer satisfaction. Therefore:

H2c: Empathized employees result in higher interaction quality

Assurance, is the fourth variable within the verbal communication dimension. It is based on the empathy variable, but goes more into the relationship with the consumer. “Does the employee tries to help the consumer actively by inspiring them and by giving them extra attention?” is a central question for this variable. Therefore: H2d: When employees give the consumer assurance, it will increase interaction quality

The next SERVQUAL variable “reliability” is important. Does the employee keep their promises? And to what degree does the employee have the ability to convince the consumer that he is right? Note: not to be confused with the actual technical knowledge of the employee.

H2e: The reliability of the employee increases interaction quality.

Finally, a hypothesis is formulated regarding the influence of verbal communication as a whole on interaction quality.

“An analysis of employee characteristics” ₂₂ 2.3.3 Knowledge constructs

In the SERVQUAL dimensions, Assurance covers this part of the service quality. “The knowledge and courtesy of employees and their ability to inspire trust and confidence”, is the definition Parasuraman (1988) gave to assurance. In this thesis however, different definitions are used to describe the “knowledge” of the employee. This is because in a D.I.Y.-store the knowledge of the employee is more deep, than defined by Parasuraman. Especially regarding the below described variable “cooperation”, assurance is considered incomplete.

The degree of knowledge of an employee is a very interesting factor regarding the D.I.Y.-market. The central question is off course; “to what extend is technical (product) knowledge important in a D.I.Y.-retail store?” Is it an essential factor or is it more important to be helped in a good way, if necessary with the help of a more technical developed co-worker.

A study by Murray (1991) concludes that sales people’s knowledge of the product will enhance their capability and motivation to communicate that to their clients. Furthermore, a study by Agnihotri et. al. (2009) shows, that when employees have appropriate technical knowledge about their products and their information communication is perceived as high, it will positively effect consumer satisfaction. According to Agnihotri et. al (2009), this is because experienced sales personnel with high technical knowledge can react better on questions of consumers than employees that are not technical knowledgeable. They can give consumers direct feedback, which consumers perceive as pleasurable. Therefore:

H3a: A higher level of technical knowledge of employees will have a positive influence on interaction quality

Secondly, consumers who have technical questions will appreciate employees, who not only explain the products sold but also help them by offering better solutions to certain problems. In this thesis, this idea is described as cooperation. Therefore: H3b: Employees who help customers by sharing their technical knowledge to create better solutions for the technical problems of consumers, will result in higher interaction quality

“An analysis of employee characteristics” ₂₃ H3c: When employees have high levels of knowledge and share this knowledge with the customer, it will have a positive influence on interaction quality

2.3.4 Employee characteristics

In this paragraph, the importance of the characteristics of employees is described. Here, age and gender and their effects on consumer satisfaction are the main issues. The characteristics of employees can be a very important variable in the interaction between consumers and employees. An older male can be seen as honest, experienced and capable, while a young female, which has equal amount of knowledge as the older male, can be viewed as incapable. Research by Foster (2004) shows that the D.I.Y.-market has a masculine image and consumers would be more likely to go to male staff for any (technical) questions than they would go to female staff. While the research of Foster (2004) is conducted in the D.I.Y.-sector, the number of interviewed consumers is low. Nevertheless the results confirm the presumption of the author that the gender of the employee affects the interaction quality.

The study of Fischer et. al. (1997) also studied the effects of female service providers. They used the SERVQUAL dimensions to see what part of the service is most affected. They concluded that gender stereotypes do affect the service quality, but that the market is a critical factor. As Foster (2004) also acknowledges, the market and the “image” of the market, determines if males (or females) are more desirable. It might be that female consumers prefer female employees to help them. Therefore:

H4a: Male employees have a higher positive overall influence on interaction quality than female employees.

H4b: The gender of the consumer influences the effect male or female employees have on interaction quality.

“An analysis of employee characteristics” ₂₄ To test if the employee’s age is relevant in the D.I.Y.-sector, the following hypothesis is formulated:

H4c: young employees have a negative influence on consumer satisfaction

Finally, the fourth hypothesis will describe the employee characteristics as a whole, asking:

H4d: positive employee characteristics have a positive influence on interaction quality.

2.3.5 Interaction quality

The previous parts of this chapter are combined in the interaction quality between employees and consumers. The article of Ekinci and Daves (2009) also included this mediator. The article of Ekinci 92009) combines the different personality traits, like extroversion and stability, that affect a consumer and combines them in an average. This average, grades the overall quality of the interaction. Furthermore, the article of Brady & Cronin (2001) also acknowledges interaction quality as an outcome of the different variables, like performance of employees, clothing and physical appearance that determine the quality of employees. In conclusion:

H5: positive interaction quality perceptions result in more positive consumer satisfaction

Because every consumer is different, the effects of consumer characteristics are also taken into account (as can be seen in the next paragraph).

2.3.6 Consumer characteristics

“An analysis of employee characteristics” ₂₅ results of the questionnaire. Using a cluster analysis, different consumer groups will appear, from which conclusions can be drawn.

2.3.7 Consumer satisfaction

As mentioned in the introduction of this thesis and described extensively in marketing literature, consumer satisfaction is one of the most fundamental concepts. Spreng et. al (1996) describes consumer satisfaction as simple as “satisfying the needs and desires of the consumer" while Kotler and Keller (2006) define satisfaction as “a person’s feelings of pleasure or disappointment resulting from comparing a product’s perceived performance (or outcome) in relation to his or her expectations”. Kotler and Keller (2006) see satisfaction in relation to expectations, while Spreng (1996) sees satisfaction as for filling needs and desires. Anderson et al. (2004), Zeithalm and Bitner (2000), Levy and Weitz (2009) Cronin and Taylor (1992) see consumer satisfaction as a evaluation between purchase and expectations. In this thesis therefore, consumer satisfaction will be based on the conceptualization of these researchers stating that; “consumer satisfaction is the relation between expectations and results”. This will create a good match with it’s primary determinant “interaction quality” were the quality of the employee is measured using the expectations and results from consumers.

As described in different articles, interaction quality is an input for consumer satisfaction. The research of Brady & Cronin (2001) uses this order, as well as Ekcini and Daves (2009). In this thesis, the overall quality of the interaction between employee and consumer is also a logical step to take. A consumer is likely to consider the entire interaction with the employee, instead of just one aspect.

“An analysis of employee characteristics” ₂₆ 2.4. Conceptual Framework

On the bases of the hypotheses in the previous paragraphs, a conceptual framework can be designed (see figure 4).

Non-verbal communication elements: - Proxemics - Kinesics - Oculesics - Vocalics Knowledge elements: - Expertise - Coöperation Verbal communication Elements: - Extroversion - Reliability - Responsiveness - Assurance - Empathy Consumer satisfaction Consumer Characteristics: (Technical level) - Never - Novice - Average - Expert Interaction Quality Employee characteristics: - Gender - Age

Figure 4: Conceptual framework

The first field (non-verbal communication), covers the H1 hypothesis. The second field (Verbal communication), comprises the H2 hypotheses. The third field covers the knowledge elements with the corresponding hypotheses H3. The fourth field (employee characteristics) can be related to H4. The four groups of hypotheses all converge into the field interaction quality. In order to test if the interaction quality, influences customer satisfaction, Hypothesis H5 relates to the link between the field Interaction quality and consumer satisfaction. The last field (consumer

“An analysis of employee characteristics” ₂₇ Chapter 3: Research Design

In the previous chapter, an extensive literature review has been described, from which a conceptual framework could be formulated. The conceptual framework serves as a starting point for this chapter. In this chapter the research design will be described, where the measurements of this thesis, as well as the data collection method and research method will be described. The measurement paragraph will describe what type of measurement will be used for this research. The data collection method will explain how the data needed for this research will be gathered. In the final paragraph, the way the research will convert this data into usable results will be described.

3.1 Measurement

Although some of the constructs that are examined came from the literature, others were adapted to suit the needs of this thesis. For example, the knowledge constructs are described by the author of this thesis, while the verbal communication constructs were described according to the SERVQUAL tool.

The research will be based on a questionnaire, which can be found in appendix 1. Most questions were assessed on a 7-point Likert scale (1= Strongly disagree and 7= strongly agree). This 7 point scale was chosen to create more variance in the answers of the respondents.

Non-verbal communication will be measured in four types, based on the research of Gabbott & Hogg (2001). Proxemics will be used to see what the effects of distancing are on consumer satisfaction. Does standing close to a customer create a negative influence of the interaction quality? Kinesics is the definition for the use of body posture. If an employee is standing up right, does that affects the interaction quality?? Oculesics, will look at the movement of the eye of the employee. Finally, vocalics will determine if the intonation and tone of the voice of the employee will affect the perceived interaction quality of the consumer.

“An analysis of employee characteristics” ₂₈ Verbal communication will be measured according to 4 of the 5 SERVQUAL dimensions (tangibles are not used in this thesis, because it is about the physical facilities, equipment and appearance of the personnel). The other four each will describe a part of the interaction between customers and employees (as described in paragraph 2.4). Furthermore, Extroversion is added.

Extroversion is a concept that looks at overall enthusiasm of the employee. The research of Brown et. al. (2002) also used extroversion in their research and revealed the significance for extrovert employees on customer significance. Furthermore, from his own experiences, the author believes that extroversion is an important variable in the interaction between employees and consumers in the D.I.Y.-market.

Knowledge constructs are divided into expertise and cooperation. Expertise will describe the overall technical knowledge, while cooperation will describe the pro-activity in which an employee uses his knowledge, i.e. does he/she help the consumer when he/she thinks there are better solutions for a problem.

Employee’s characteristics are the last independent variable for interaction quality. It looks at the characteristics of the employee. What are the differences between young and old, or male or female employees?

Interaction quality is the entire interaction between an employee and a consumer. It is measured based on the previous four groups. According to the book of Aaker, Kumar and Day (2004) it is important to get an overall rating at the beginning of the questionnaire. This question can be used as a dependable variable. Therefore, the average score of the four groups, at the beginning of the questionnaire the respondents are asked to give a grade to the overall satisfaction with D.I.Y.-store employees.

“An analysis of employee characteristics” ₂₉ The overall goal of this thesis is to see what the satisfaction of consumers is regarding D.I.Y.-store employees. Therefore, this will be the dependant variable. 3.2 Data collection method

To assess to what extend employees influence consumer satisfaction, an empirical study is conducted in the Dutch D.I.Y.-market. This data will be collected via questionnaires distributed among D.I.Y.-customers. Respondents are consumers in Gamma stores in the Netherlands and consumers approached via an internet survey. We limit ourselves to respondents above the age of 18 years, since consumers below 18 year old will mostly not have developed a consumer satisfaction level regarding D.I.Y.-stores. Since the questionnaire is held in a D.I.Y.-store the respondents of the first sample already are the target group for this research.

The population for this research are all customers of D.I.Y.-stores from 18 years and older. Since this is a large group, the limit for this research is at least 150 respondents.

The questionnaire is translated into Dutch, and will be tested among a group of 30 respondents. After this test, the questionnaire will be modified before the final questionnaire is held.

As mentioned in the first chapter of this thesis, the questions asked in the questionnaire do not measure the score of employees in the D.I.Y. Market, but test the importance of certain characteristics. The respondent is asked (as can be seen in the questionnaire in appendix 1), to value characteristics as described in the conceptual framework. Respondents give their opinion about these characteristics, without testing the score for a certain (group of) employees. To test the employees score, more different factors are needed. First, the actions of the employee need to be measured unobtrusively, secondly the respondent needs to view the performance of the employee in the right way without being biased in any way.

“An analysis of employee characteristics” ₃₀ 3.3 Research method

The research design is conclusive and descriptive. The collection of data will be done once, making this research cross sectional instead of longitudinal.

Since most questions are based on a 7-point Likert scale, the data is based on an interval scale. Two questions, where respondents have to rank some statements from best fitting to worst fitting, are based on a ordinal scale and questions about the respondent themselves (age, gender) are nominal. The questionnaire consists primarily of non-comparative scaling techniques.

In order to test the internal validity of this research a Cronbach’s Alpha test will determine if the questions asked for every group determined in the conceptual framework, i.e. the non-verbal communication construct but also the proxemics part specifically correlates significantly. When the correlation is positive, the individual questions can be used into one construct. From there on, calculations can be made to see what constructs have most influence on the interaction quality. When correlation is low, the questions will need to be used individually.

“An analysis of employee characteristics” ₃₁ Chapter 4: Results

In this chapter, the results which came from the respondents who filled in the questionnaire will be presented and discussed.

4.1 Introduction

The results will be drawn from the data received via the questionnaire and put into statistically underlined discussion points

In paragraph 4.2 some general calculations will be made, explaining the number of visits a year and the D.I.Y.-store that is visited most often. In paragraph 4.3 the internal consistency of the measurements will be discussed, using Cronbach’s Alpha. In paragraph 4.4 the four different constructs presented in chapter 3 will be calculated, determining the interaction quality and therby accepting or rejecting the hypotheses. In the last paragraph, the different consumer characteristics will be combined into consumer groups, followed by the consumer satisfaction of these consumer groups.

4.2 General remarks

“An analysis of employee characteristics” ₃₂ 4.2.1 Gender

The dataset of 154 respondents (three respondents did not fill in their gender) has a somewhat equal number of male (81) and female (73) respondents. Regarding the number of visits a year, using a 95% significance level (independent t-test test), there is a difference between male and female with significance of 0.001. this indicates that male respondents visit the store more often than female

respondents, as can be seen in figure 5, Male respondents visit D.I.Y.-stores more often than females. Were most females visit the store 10 times a year, most men visit a store 25 times a year.

4.2.2 Age

The dataset of 154 respondents (three respondents did not fill in their age) is somewhat evenly divided. In graph 6 the age of the respondents is shown in relationship to their store visits. Using an ANOVA test, differences between the groups presented in figure 6 are tested. The number of visits to the store shows a non significant trend. (Sig.= 0.072). this indicates that there is no

“An analysis of employee characteristics” ₃₃ are found (Sig.=0.085). In conclusion, there are no initial differences found within the respondents group were age was a determinant. In paragraph 4.5, the age of the respondents will be discussed in more detail.

4.2.3 Technical level

Another important variable used to cluster respondents is their technical level. Four levels of technical expertise were

selected, from never doing any technical work, to being a novice regarding technical knowledge, having average technical knowledge and being an experts.

The results are presented in table 2. Most respondents see themselves as people with novice or average technical knowledge. Only 12.7% of the respondents say that they never do any technical work.

4.2.4 General remarks

The average number of visits a respondent makes a year is 16.8 times a year. The median is 15. In total, the range of visits was between 2 and 50 times a year. Gamma is the store most visited by the respondents (63.7%) followed by Praxis (17.8%). Other stores had small percentages. This is because the questionnaires were handed out in Gamma stores.

The most important reason for the consumer to go to a certain D.I.Y.-store is when the store is nearby. Helpful employees is the second most important reason, but the difference is immense, respectively 76.8% to 8.3%. Respondents gave an average grade of 6.62 point on a 10 point scale to the question what grade they want to give to their most visited D.I.Y.-store. The median and mode are 7. Overall, respondents were modestly satisfied with employees quality in D.I.Y.-stores in general, giving them a 5.13 on a 7 point Likert scale.

Regarding the general loyalty questions, the highest score on a 7 point Likert scale, is that respondents will visit a D.I.Y.-store more often when satisfied about the employees of that certain store (5.77). Nevertheless, respondents are less willing to make a detour to that certain store when there is another store nearby (5.23). The

Table 2: Technical level

Technical level Frequenties Percentage

Never 20 12,7

Novice 60 38,2

Average 43 27,4

“An analysis of employee characteristics” ₃₄ location of other stores seems to have influence on the loyalty behaviour of the respondents.

4.3 Internal consistency

In this paragraph, the consistency of the different questions asked in the questionnaire will be measured using a factor analysis and the Chronbach’s Alpha. 4.3.1 Factor analysis

In order to test if the questions (see appendix 1) in the survey can be grouped into a lower number of variables, we need to conduct a factor analysis. Although the questions are already grouped into four constructs, the factor analysis needs to verify that there are indeed four constructs.

We first need to check if the factor analysis is reliable and significant, using a Kaiser-Meyer-Olkin(KMO) and Bartlett test. In the table below, these tests are presented.

Table 3: KMO and Bartlett's Test Kaiser-Meyer-Olkin Measure of

Sampling Adequacy. 0.865

Bartlett's Test of Sphericity Approx. Chi-Square 2967.298

df 496

Sig. 0.00

“An analysis of employee characteristics” ₃₅ The factor analysis showed six factors in the component matrix, as can be seen in the table below.

Table 4: Factor analysis

Component Rotation Sums of Squared Loadings

Total % of Variance Cumulative %

1 5,494 17,167 17,167 2 4,043 12,635 29,803 3 3,830 11,969 41,772 4 2,869 8,964 50,736 5 2,602 8,131 58,867 6 1,522 4,756 63,623

The factor analysis shows six factors. Regarding the rotation sums (performed using VARIMAX) an Eigenvalue of 1.0< is needed in order to be a reliable factor. Another indicator is the minimal % of variance of 5. (Which only the 6th _{factor does not have).}

“An analysis of employee characteristics” ₃₆ To see which questions have a high loading on the factors, a factor loading matrix is needed. It was rotated using VARIMAX to receive a better view of the correlation of the questions. In table 5, this matrix is presented.

Table 5: Rotated Component Matrix Component 1 2 3 4 5 6 NVC1 Distance ,117 ,226 ,686 -,103 ,017 -,015 NVC2 Straight ,157 ,004 ,789 ,129 ,131 ,021 NVC3 Still ,091 ,091 ,766 ,166 ,143 -,045 NVC4 Look -,035 ,132 ,778 ,087 ,292 ,099 NVC5 Be Clear ,046 ,198 ,677 ,075 ,253 ,088 NVC6 Loud ,090 ,190 ,684 ,339 -,034 ,200 VBE1 Greet ,195 ,022 ,169 ,641 ,278 -,158 VBE2 Direct ,413 -,052 ,008 ,691 ,073 ,178 VBE3 Succes ,358 ,150 ,059 ,713 ,018 ,116 VBR1 Promiss -,148 ,259 ,228 ,028 ,643 ,192 VBR2 sympathize ,112 ,267 ,286 ,465 ,252 ,092 VBRE1 Always assist -,056 ,253 ,106 ,146 ,731 -,023 VBRE 2 Not Always react ,428 ,042 ,073 ,139 ,196 ,482

VBA1 Trust -,028 ,330 ,232 ,063 ,609 ,259

VBA2 Polite ,105 ,256 ,314 ,335 ,579 ,008

VBE1 Personal ,086 ,166 ,179 ,314 ,418 -,242

VBE2 Wishes ,193 ,040 ,159 ,054 ,102 ,690

“An analysis of employee characteristics” ₃₇ scores a 0.700 in factor four and five. This indicates that these factors have a relatively low loading.

The factor analysis shows six groups, two more than the initial four groups presented in the conceptual framework. Using a Chronbach’s Alpha, we can test these factor groups for internal consistency, as can be seen in table 6.

Table 6: internal consistency of factors

Factor Number of items Cronbach's Alpha

1 8 0.923 2 6 0.873 3 6 0.863 4 5 0.736 5 5 0.774 6 2 0.576

Except for factor 6, the internal consistency of these groups is good. This could indicate that there are not four, but 6 groups of constructs. In the next paragraph, other tests will be done to determine the groups that will be used. In paragraph 4.3.3 the factor analysis groups and the groups which will be discussed in the next paragraph, will be compared.

4.3.2 Four constructs

In this paragraph, we will test the four initial constructs to see if their reliability is high enough for grouping variables. Using the Cronbach’s Alpha, the reliability of the scales was tested. According to Nunally & Bernstein (1994) the reliability is accepted when Cronbach’s Alpha is 0.70 or higher. Between 0.50 and 0.70, the reliability is accepted, but needs additional measurements to check if the questions are reliable enough to combine into one construct.

* Variable of which the separate items become one variable

In table 7 the different reliability scores are presented, showing the different types of constructs used in the conceptual framework. The questions used for each construct can be seen in appendix 1.

Table 7: Reliability scores of constructs

Construct Nr. of

items Cronbach’s Alpha Items if deleted

Non-verbal Communication 6 0.863* ,510

Verbal Communication 12 0.795* ,387

Knowledge constructs 6 0.873* ,606

“An analysis of employee characteristics” ₃₈ As can be seen in the table all four variables are above the 0.70 acceptance score. Furthermore, none of the constructs will create a higher Cronbach’s Alpha score if an item is deleted. It can be concluded that the constructs are internally consistent and further calculations can be done with these variables.

4.3.3 Choosing groups

When comparing the factor analysis and the four groups from the conceptual framework, there are some similarities and some differences. Firstly, the constructs Non-verbal communication, knowledge constructs and employee characteristics are identical to respectively factor three, two and one. Secondly, the Cronbach’s Alpha’s are almost all relatively high indicating some consistency. Nevertheless, regarding factor four, five and six, which have the same questions as the construct verbal communication, it can be seen that the factors score lower than the verbal communication construct. Table 7 shows that verbal communication has a Chronbach’s Alpha of 0,795, while the three factors in table 6 (factor 4, 5 and 6) have lower Chronbach’s Alpha scores. Therefore the four original constructs will be used.

Compare constructs

Since the four constructs proved to be internally consistent, calculations can be made to see the differences between the constructs. In the table below, the means are presented to see which construct has the highest score and is most important.

Table 8: Four constructs of interaction quality

Constructs: Mean:

Non-verbal Communication 5.6433 Verbal Communication 5.5663 Knowledge constructs 6.1932 Employee characteristics 4.3376

“An analysis of employee characteristics” ₃₉ In figure 7, four histograms show the distribution of the grades respondents gave (on average) to the four constructs. As can be seen, the knowledge constructs had the highest average

score, and most of the respondents rate these questions on average with 6 out of 7 points. Additionally, verbal- and non-verbal

communication also shows 6 out of 7 points as being the highest average score, but scores of 4 and 5 were more common in comparison to the knowledge constructs. Employee charac-teristics however, show a very widespread

distribution of the scores given. This indicates that respondents have different opinions about this construct. In conclusion, the constructs verbal-, non-verbal communication and knowledge constructs show high average scores, which indicate that respondents find these constructs important, while the widespread distribution of employee characteristics indicates various levels of importance.

“An analysis of employee characteristics” ₄₀ 4.3.4 Sub constructs

As can be seen in the table below, some of the sub-constructs show an internal consistency.

Table 9: Reliability of sub-constructs

Sub-construct: Nr. of items: Cronbach’s Alpha

Non-verbal

communication Proxemics 1 n.a.

Kinesics 2 0.783* Oculisics 1 n.a. Vocalics 2 0.637 Verbal communication Extroversion 3 0.753* Reliability 2 0.458 Responsiveness 2 0.199 Assurance 2 0.673 Empathy 3 0.304 Knowledge constructs Expertise 3 0.710* Cooperation 3 0.901* Employee characteristics Gender 5 0.906* Age 3 0.816*

* Variable of which the separate items become one variable

“An analysis of employee characteristics” ₄₁ 4.3.5 Correlation

The new variables made in the previous paragraph will be used in this paragraph to determine if it is possible to use these variables when determining the most important constructs for interaction quality.

Firstly, the four constructs which are the variables for interaction quality need to be tested, to see if there is any correlation. In the table below, the four constructs are tested using the Pearson correlation.

Table 10: Correlations

Non-verbal

Communication Verbal communication Knowledge Constructs Employee Characteristics Non Verbal Communication (NVC) Pearson Correlation 1.000 0.519 0.390 0.217 Sig. 0.000 0.000 0.006 N 157 157 157 157 Verbal Communication (VC) Pearson Correlation 0.519 1.000 0.450 0.481 Sig. 0.000 0.000 0.000 N 157 157 157 157 Knowledge constructs (KC) Pearson Correlation 0.390 0.450 1.000 0.036 Sig. 0.000 0.000 0.654 N 157 157 157 157 Employee Characteristics (EC) Pearson Correlation 0.217 0.481 0.036 1.000 Sig. 0.006 0.000 0.654 N 157 157 157 157

“An analysis of employee characteristics” ₄₂ 4.4 Hypotheses

In this paragraph, the hypotheses that are described in the theoretical framework will be accepted or rejected.

In the table below, the four constructs used in the theoretical framework to determine interaction quality are described. A regression analysis is used to check if there is a relationship between the four constructs and the interaction quality. If certain constructs change, the influence on the interaction quality can be calculated. First, using the model summary of the regression analysis, the correlation of the dependent variable with the independent variables combined will be tested.

The model shows that there is almost no explained variance between the dependent and independent variables. The adjusted R square, which is used best regarding the number of cases, is even negative. The model summary indicates that there is no representation of a relationship between the variables. The results indicate that there is a non-significant relationship between the dependent and independent variables. When regarding the ANOVA test, we can see that the explained variance (regression) is very low and that the residual is high. This further indicates that the relationship between the dependent and independent variables are insignificant.

Table 12: ANOVA

Model Sum of Squares df Mean Square F Sig.

1 Regression 5,965 4 1,491 ,702 ,592a

Residual 322,863 152 2,124

Total 328,828 156

Table 11: Model Summary

Model R R Square Adjusted R Square Std. Error of the Estimate

“An analysis of employee characteristics” ₄₃ The final regression model shows the relationships between the dependent variable and the independent variable individually.

Table 13: Hypothesis Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta (Constant) 6,045 1,137 5.317 0.000 H1: Non-verbal communication -0,192 0,151 -0,123 -1.278 0.203 H2f: Verbal Communication 0,305 0,22 0,158 1.385 0.168 H3c: Knowledge constructs 0,024 0,188 0,012 0.125 0.900 H4d: Employee Characteristics -0,029 0,098 -0,028 -0.293 0.770 As can be seen in the previous tables, none of the constructs are significant. The model itself is insignificant but also the four main constructs show no significant relationship to the dependent variable. Therefore, these hypotheses need to be rejected. All other results are based on the assumption that the model is significant, to give a overview of the results.

In conclusion, since the overall model shows a clear insignificant result, with almost no explained variance and high residual scores in the ANOVA test, the hypotheses regarding the four constructs (hypothesis 1 till 4) need to be rejected.

This shows that the four constructs in this model, used to determine the interaction quality between employee and customer do not influence the interaction quality significantly. The last hypothesis, regarding the influence of interaction quality on customer satisfaction is discussed in the following paragraph.

“An analysis of employee characteristics” ₄₄ 4.5 Customer Satisfaction

Although the constructs for interaction quality were insignificant, the relationship between interaction quality and customer satisfaction can be examined. In this paragraph, the link between interaction quality and customer satisfaction will be described.

The last hypothesis, H5: “positive interaction quality perceptions result in more positive consumer satisfaction” determinants if interaction quality has a positive effect on customer satisfaction. In table 14, these results are presented.

Table 14: Link between interaction quality and consumer satisfaction

Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) 3,142 ,376 8,354 ,000 Rate quality of D.I.Y.-Store ,300 ,056 ,402 5,413 ,000

As can be seen in the regression analysis above, there is a clear significant result regarding the relationship between interaction quality and customer satisfaction. There is a .300 positive relation between the two variables. This indicates that for every increase of 1 point on a 10 point scale measuring the interaction quality, customer satisfaction (which is based on a 7-point Likert scale), increases with .300 point, with a constant of 3.142.

When the interaction quality is rated with a 10, this gives an average satisfaction of 6.142, which is very high. Although a rise of .300 is not a very high positive increase, it is substantially enough to accept hypothesis H5.

4.6 Other findings

Although the main hypothesis were rejected, other findings can be extracted from the data. In this paragraph, two findings regarding gender and loyalty are described. Gender