To what extent does the type of customer behavior has an effect on the type of customer co-creation and how does the customer's past experience with Coolblue moderate this effect?

(1)

Master Thesis

“To what extent does the type of customer behavior has an effect on the type

of customer co-creation and how does the customer’s past experience with

Coolblue moderate this effect?”.

Author: Supervisor:

Karina Zantinge Prof. Dr. Ed Peelen

10871780 Universiteit van Amsterdam

in the

Master Business Administration Marketing Track

Universiteit van Amsterdam

Submission date: 24 March 2017 Final version

(2)

Statement of Originality

This document is written by Karina Zantinge who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Table of Contents

1. Abstract 5

2. Introduction 6

2.1. Setting 6

2.2. Research question 7

2.3. Current state of literature and theoretical gap 7

2.4. Managerial gap 8

2.5. Scope 8

2.6. Structure 8

3. Literature review 10

3.1. Customer Value management 10

3.2. Customer Engagement 10

3.3. Customer Engagement Value 12

3.4. Co-creation 13

3.5. Research gap 14

3.6. Research question, conceptual model and hypotheses 15

4. Data and Method 19

4.1. Research Design 19

4.2. Data collection and sampling technique 19

4.3. Measures and Instruments 20

4.4. Data analysis 20

4.4.1. Step1 – missing values and recoding 20

4.4.2. Step 2 – rotated component matrix 21

(4)

4.4.4. Step 4 – computing variables 21

4.4.5. Step 5 – residual diagnostics 22

4.4.6. Step 6 – moderator analysis 22

5. Results 23

5.1. Descriptive statistics 23

5.2. Multiple regression analyses 24

5.3. Correlation analysis 27

5.4. Backward regression analysis 28

6. Discussion, limitations and further research 29

7. Conclusion 32

8. References 33

(5)

1. Abstract

Customer engagement has been widely researched, with co-creation being one of the constructs to measure customer engagement value. However, since e-commerce has grown tremendously in the last years, it provides grounds for new research to test the existing literature. This study is based on co-creation theory by Yi and Gong (2013) and buying

behaviors study by Ivanova (2016) and examines to what extent the type of customer behavior affects the type of customer co-creation and how the customer’s past experience moderates this effect. The case study is Coolblue, an online Dutch electronics retailer. This is an exploratory study based on an online survey distributed in the Netherlands. The interaction effect is studied with four multiple regression analyses. The results show no significant effect between the type of customer buying behavior and the type of co-creation. The results

indicate that behavioral attitudes, as in consumer shopping characteristics, are not the right predictors for the type of co-creation practiced by the consumer. In addition, the study points out that co-creation types by Yi and Gong (2013) are strongly correlated and further research is needed into measurability of co-creation by itself. In addition, buying behaviors by

Ivanova (2016) might differ in an online setting compared to a physical store and should therefore be reexamined separately.

(6)

2. Introduction

2.1 Setting

Recent literature shows that there is a growing interest in customer engagement value in the field of business and economics (Woodruff, 1997; Verhoef & Lemon, 2012; Hollebeek, 2011). Companies realize that aside from the transactional value, an engaged customer can also add value by leveraging knowledge and feedback. A loyal customer will not only refer the brand to others, but might be willing to help or give advice to other customers. All in all, co-creating with the brand and aiding in its future growth and development. Still, it is unclear when and how the customer is prepared to co-create. So far, there has been little research done on behavioral attitudes and their effects on co-creation.

When looking at behavioral attitudes the first thing that comes to mind is the way customers shop. Even though, buying behaviors have often been researched, there is no literature that links the way people shop to their customer engagement.

In addition, most research on buying behaviors has considered shopping in a physical setting. However, recent reports by CBS (a Dutch governmental institution that gathers statistical information about the Netherlands) show that e-commerce has been on the rise. Nearly 73% of all Dutch internet users made at least one online purchase in 2016, which is around 11 million people (CBS, 2016). Such major trend asks for a more detailed

understanding.

It is interesting to see whether it is possible to combine these two major trends in theory and economics. Can customer’s buying behavior lead to a certain type of co-creation? What happens when we consider an online setting?

(7)

online retailers in the Netherlands (Coolbue, 2017). They offer a great platform to study co-creation as they invest heavily in customer engagement and have a large online presence.

2.2 Research question

The research question of this study is: “To what extent does the type of customer behavior has

an effect on the type of customer co-creation and how does the customer’s past experience with Coolblue moderate this effect?”.

2.3 Current state of literature and Theoretical gap

Current research suggests that customer engagement is a multidimensional construct

comprising both transactional and behavioral aspects (Hollebeek, 2011; Brodie, 2011). While the transactional value of the customer is relatively easy to measure, there is still a lot of different views on the value of non-transactional behaviors.

Most commonly researched motivational behaviors are word-of-mouth activity, recommendations, customer-to-customer interactions, blogging, reviews etc. (MSI, 2010). Kumar et al. (2010) mention customer knowledge value, or providing knowledge and feedback to the firm, as an additional way to enhance value. Also, Prahalad and Ramasway (2004) state that to enhance value the firm needs to do more than just to engage consumers as co-sales agents, the value creation is about developing methods to attain an understanding of co-creation experiences. Yi and Gong (2013) studied this co-creation experience and have developed a scale to measure its value. They distinguish between two dimensions. First dimension is the customer participation, the required (in-role) behavior. The second one is the customer citizenship behavior, the voluntary (extra-role) behavior, which is more interesting to study due to theoretical gap.

(8)

limited to referring or influencing, but also to sharing knowledge, helping other customers, or being a tolerant customer (Yi & Gong, 2013). The question rises when do customers display these behaviors and what are the motivations for these behaviors. This study will tap into this theoretical gap and aid in better understanding of co-creation processes.

2.3 Managerial gap

There were approximately 70 thousand web shops in the Netherlands in 2016 with a turnover of 23 billion euros (Emerce, 2017). This shows that e-commerce is a big part of Dutch economy. Still, there is not enough literature to understand how characteristics or buying behaviors of online customers could affect their total value for the company. This study will tap into this managerial gap and aid in better understanding of types of customers and their non-transactional value for the company.

2.4 Scope

The scope of this study is limited to the Netherlands only with a sample size of 188 respondents. It is a survey research with variables based on co-creation theory by Yi and Gong (2013) and buying behaviors study by Ivanova (2016). The interaction effect is studied with multiple regression analyses. In addition, past experience is tested for its moderating affect. Correlation and backward regression analyses are included for additional outcomes.

2.5 Structure

Chapter 3 includes literature review. First the theory behind customer value and the

development of value metrics are discussed. Than, the broad concept of customer engagement is introduced. Followed by different concepts that try to measure the value of customer

(9)

engagement. Research gap shows which theories are chosen as the starting point and what the focus of this study will be.

Chapter 4 includes research design and elaborates on sampling technique and how the data was collected. A step-by-step data analysis explains how data was validated, whether it was reliable and how the variables were computed.

Chapter 5 includes the results of the data analysis. Descriptive statistics show the demographical characteristics of the respondents and all means are compared. The multiple correlation analyses show whether the hypotheses are accepted or rejected. Additional statistical methods are included.

Chapter 6 draws conclusions and answers the research question. In discussion, the results are reflected upon theory. In addition, limitations and advice for future research is given.

(10)

3. Literature review

3.1 Customer value management

In the late nineties, with more demanding customers, global competition and slow-growth economies and industries, customer value was the new source for seeking competitive advantage (Woodruff, 1997). A decade later customer value management (CVM) was well studied and incorporated across different economies and industries. However, new

technologies and changes in customer expectations, experiences and behavior required a fresh look. Verhoef and Lemon (2012) foresaw three emerging perspectives on CVM: (1)

managing customer engagement, (2) managing customer networks, and (3) managing the customer experience. An emerging trend and importance of managing customer engagement (CE) have led to this perspective being the starting point of this study.

3.2 Customer engagement

The process of customer engagement was first conceptualized by Bowden (2009) as an end state of customer loyalty. The author proposed that customer engagement as process included: the formation of a state of calculative commitment for new customers which is considered to be a largely cognitive basis for purchase; increased levels of involvement concomitantly supported by increased levels of trust for repeat purchase customers; and the development of affective commitment toward the service brand which is considered to be a more emotive basis for purchase and which may ultimately eventuate in a state of enduring brand loyalty (Bowden, 2009). Even though previous research had provided insights on customer engagement, Hollebeek (2009) was the first to study whether customer engagement did in fact generate customer value. His findings showed curvilinear relationship between CE/CV and thus customer engagement did serve as a driver of customer value.

(11)

While it can be suggested that customer engagement may require consideration be given to both the psychological aspects of engagement as well as behavioral participation, it appears that there remains a diversity of views in respect to the conceptualization of the concept. Some researchers consider customer engagement to be purely a behavioral construct resulting from a range of motivational drivers (MSI, 2010; Van Doorn et al., 2010; Bijmolt et al., 2010). The Marketing Science Institute (MSI, 2010) defined customer engagement as “customer’s behavioral manifestation toward a brand or firm beyond purchase, which results from motivational drivers including: word-of-mouth activity, recommendations, customer-to-customer interactions, blogging, writing reviews, and other similar activities”. In addition to this view, some researchers have defined customer engagement to include only

non-transactional customer behaviors. Van Doorn, Lemom, Mittal, Nass, Pirner and Verhoef (2010) pointed out that engagement is behavioral in nature and propose that it goes beyond transactions, and is specifically defined as a customer’s behavioral manifestation toward a brand of firm, beyond purchase, resulting from motivational drivers. Bijmolt, Leeflang, Block, Eisenbeiss, Hardie, Lemmens and Saffert (2010) discussed analytical models for customer engagement that are too beyond customer transactions. Their models pertained to the subsequent stages of the customer life cycle: customer acquisition, customer development, and customer retention.

However, other researchers propose customer engagement to be a multidimensional construct comprising both psychological and behavioral aspects (Hollebeek, 2011; Brodie, 2011). Hollebeek (2011) defined customer engagement as a psychological state that occurs by virtue of interactive, co-creative customer experiences with a focal agent/object (brand) in focal service relationships. It occurs under a specific set of context dependent conditions generating differing CE levels; and exists as a dynamic, iterative process within service relationships that co-create value. Brodie, Hollebeek, Juric and Ilic (2011) have researched

(12)

further into the customer engagement concept and have also pointed out that customer engagement, unlike other relational concepts, was based on the existence of a customer’s interactive, co-creative experiences with a specific engagement object (brand). In 2014 Fung So, King and Sparks (2014) developed a framework that was useful to collect insights into customer psychological and behavioral connections with their brands beyond the service consumption experience. The framework demonstrated strong psychometric properties across multiple samples and showed CE to exert a positive significant influence on behavioral intention of loyalty for both hotel and airline customers.

3.3 Customer Engagement Value

To represent customer engagement as a multidimensional construct Kumar, Aksoy, Donkers, Venkatesan, Wiesel and Tillmanns (2010) introduced the customer engagement value (CEV) as an overarching new customer value metric that includes both value from transactions (CLV) and value from non-transactional behavior. Following the conceptualization of van Doorn et al. (2010), CLV would remain the overarching customer value metric to which the value resulting from customer engagement (i.e., CRV, CIV, and CKV) should be added. The components that make up CEV can be determined by aggregating the value of a customer’s own transactions and corresponding CLV, CRV generated by bringing in new customers via referrals thereby aiding in the acquisition process, CIV generated by primarily influencing and encouraging existing customers to continue and/or expand usage post acquisition as well as encouraging prospects (individuals the firm is trying to acquire) to buy, and CKV created by providing knowledge and feedback to the firm to aid in the innovation process (Kumar et al., 2010). In order to develop and implement effective marketing strategies and to ensure the efficient allocation of resources, it is essential for companies to understand the exact nature of each of these various elements.

(13)

3.4 Co-creation

Similarly, other literature shows that there are other value metrics to pinpoint and measure customer engagement. Gronroos and Voima (2011) state that value creation refers to customers’ creation of value-in-use, in other words, co-creation as a function of interaction. Witell, Gustafsson and Lofgren (2011) mention that an organization must develop its

collaborative competence in order to move away from perceiving the customer as a source of information. Customer should be treated as an active contributor with knowledge and skills for overall value co-creation.

The concept of co-creation was first operationalized by Prahalad & Ramaswamy (2004) who stated that the future belonged to those that can successfully co-create unique experiences with customers. According to authors, co-creation is about developing methods to attain a visceral understanding of co-creation experiences so that companies can co-shape consumer expectations and experiences along with their customers (Prahalad & Ramasamy, 2004). Eventhough the authors mention the importance of co-creation they do not propose any measurement scales. Years later, Payne, Storbacka and Frow (2007) state that still relatively little is known about how customers engage in co-creation, as it can be viewed from many different perspectives.

In 2012, Yi & Gong wrote a study on the development and validation of a customer value co-creation behavior scale. They have come up with a scale that is multidimensional and hierarchical, and exhibits internal consistency reliability, construct validity, and

nomological validity. Therefore, their study serves a good starting point to further examine and measure co-creation. In addition, one of the conclusions the authors make is that customer’s behaviors exhibit different patterns of antecedents. For better understanding of antecedents more research is required.

(14)

3.5 Research Gap

Given that customer engagement has the potential to increase a company’s bottom line, it becomes crucial to explore ways in which it can be maximized. Out of all different concepts of customer engagement, co-creation has been researched least and therefore offers more fertile ground for theory building. Seeing that current literature on measuring co-creation specifically is still limited there is an obvious theoretical gap. The scale by Yi and Gong (2012) offers a good starting point, but there is still a lot to research when it comes to cause and effects.

Since most literature on co-creation emphasizes the importance of customer, it is interesting to study the phenomenon from the customer’s perspective as well. In this case, we need to find a way to distinguish customers to understand their involvement and readiness to co-create. The attitudinal drivers, customers’ demographical characteristics and their buying behavior, are the first thing to consider. At the end, personality traits are often seen as enduring factors that can influence behavior.

Ivanova (2016) has written her thesis on consumer decision making styles where she validates the construct of eight consumer characteristics by Sproles and Kendall (1986). The consumer characteristics are: perfectionistic, habitual, price-value conscious, hedonistic, fashion conscious, impulsive, confused by overchoice, and brand conscious (Ivanova, 2016). This study aims to examine the interaction between these consumers’ characteristics and the Yi and Gong’s co-creation scale model. Since the case study is Coolblue, an electronics online retailer, not all characteristics are equally applicable. Fashion or brand conscious buying behavior is not likely to come up when purchasing expensive electronic goods, same goes for impulsive or hedonistic type of behavior. Based on the retailer’s website and product variety (Coolblue, 2017) its typical consumer is likely to seek best quality or best price-value ratio. Coolblue also offers ‘Coolblue’s choice’ category and a reviews system to help those

(15)

who are confused by overchoice and to ensure their customers become habitual. There is also no existing theory on whether past experience has a moderating effect on the interaction between attitudinal drivers and types of co-creation practiced. This study aims to answer the question whether it is logical to assume that for example a price-value conscious consumer that finds a great deal on Coolblue’s website is likely to tell others about it and whether this chance increases if he or she has been a happy customer before.

3.6 Research question, conceptual model and hypotheses

The research question of this study is: “To what extent does the type of consumer has an

effect on the type of co-creation practiced and how does the consumer’s past experience with Coolblue moderate this effect?”.

The conceptual model consists of dependent variables, independent variables and the moderator. The independent variables are the four types of consumers: perfectionistic, price-value conscious, confused by overhoice, habitual (Ivanova, 2016). The dependent variables are the four components of voluntary co-creation (Yi & Gong, 2013). The moderator is the past experience that the customer has had with Coolblue. See four figures below.

Accordingly, there are 32 hypotheses.

(16)

Fig.3 Affect on Helping type of co-creation Fig.4 Affect on Tolerance type of co-creation

The hypotheses for the feedback type of co-creation:

(H1) The ‘Perfectionistic’ type of consumer has an affect on the ‘Feedback’ component of co-creation.

(H1a) Past experience moderates the affect of the ‘Perfectionistic’ type of consumer on the ‘Feedback’ component of co-creation.

(H2) The ‘Impulsive’ type of consumer has an affect on the ‘Feedback’ component of co-creation.

(H2a) Past experience moderates the affect of the ‘Impulsive type of consumer on the ‘Feedback’ component of co-creation.

(H3) The ‘Price-Value Conscious’ type of consumer has an affect on the ‘Feedback’ component of co-creation.

(H3a) Past experience moderates the affect of the ‘Price-Value Conscious’ type of consumer on the ‘Feedback’ component of co-creation.

(H4) The ‘Confused by overchoice’ type of consumer has an affect on the ‘Feedback’ component of co-creation.

(H4a) Past experience moderates the affect of the ‘Confused by overchoice’ type of consumer on the ‘Feedback’ component of co-creation.

(17)

(H5) The ‘Confused by overchoice’ type of consumer has an affect on the ‘Advocacy’ component of co-creation.

(H5a) Past experience moderates the affect of the ‘Confused by overchoice’ type of consumer on the ‘Advocacy’ component of co-creation.

The hypotheses for the helping type of co-creation:

(H9) The ‘Confused by overchoice’ type of consumer has an affect on the ‘Helping’ component of co-creation.

(H9a) Past experience moderates the affect of the ‘Confused by overchoice’ type of consumer on the ‘Helping’ component of co-creation.

(18)

The hypotheses for the tolerance type of co-creation:

(H13) The ‘Confused by overchoice’ type of consumer has an affect on the ‘Tolerance’ component of co-creation.

(H13a) Past experience moderates the affect of the ‘Confused by overchoice’ type of consumer on the ‘Tolerance’ component of co-creation.

(19)

4. Data and method

4.1 Research Design

A quantitative study was conducted in order to answer the research question and to test the above hypotheses. This study examines the case of Coolblue and its customers, as Coolblue is a large and popular retailer with a large number of customers to approach. In addition,

customer engagement is one of the pillars of Coolblue.

For this correlation study, four multiple regression analyses were carried out to test the effect between the type of buying behavior and the type of co-creation and whether

customer’s past experience with Coolblue has an effect on this interaction.

To ensure a reliable unbiased data a dummy question was asked whether the

respondent has ever reviewed Coolblue on social media before. This way not only people who were already motivated to leave feedback, but also those who were not, were taken into consideration.

4.2 Data collection and sampling technique

The study took place in the Netherlands. For the sampling frame all people in the age range 18-65, both male and female, were asked to fill in the survey. There were no restrictions with regard to the education, income or nationality. Demographical data was still asked to provide for additional conclusions.

The survey was created with Qualtrics and was posted online. The link to the survey was supposed to be distributed through Facebook, as Coolblue’s page has over 8000 reviews by existing customers with their given name. However, due to Facebook’s newest regulations it was no longer possible to approach more than thirty people. Therefore, the link to the

(20)

survey was distributed by email and personal social media, but also by giving out flyers at the University of Amsterdam.

4.3 Measures and Instruments

To measure the independent variables, the scale by Ivanova (2016) was used (see Appendix I). Out of eight types of consumers she mentions, only the most applicable to this study were used, namely: perfectionistic, habitual, price-value conscious, and confused by overchoice. The buying behaviors were set as nominal variables.

To measure the dependent variables, the scale by Yi and Gong (2013) was used (see Appendix II). Only the components of customer citizenship behavior were used, namely: the feedback, advocacy, helping, and tolerance. The types of co-creation were set as scale variables.

To measure the moderating effect of past experience the participants were asked to rate their prior experience with Coolblue as: not satisfied, slightly unsatisfied, neutral, slightly satisfied, very satisfied.

All variables were measured with a 5-item Likert scale.

4.4 Data analysis

All raw data was exported from Qualtrics to a personal computer. The latest version of SPSS (24) was used to analyze the data. The data consisted of 237 responses total. First step was data cleaning and dealing with missing values. All respondents that answered a ‘no’ in the dummy question and have never purchased anything from Coolblue before, as well as all incomplete responses, were excluded (N = 49). The data used for all further analyses

consisted of 188 completed responses. There was one counter-indicative item (Q9) that was recoded.

(21)

Second step was to see whether the questions that composed the variables indeed belonged together and did not relate to other variables. With a factor analysis a rotated component matrix provided for eight components (see Appendix III). Not all components were perfect, as expected these factors were contextually interrelated. Most items above 0,3 were grouped together into components: (1) advocacy, (2) confused by overchoice, (3) perfectionistic, (4) tolerance, (5) price-value conscious, (6) helping, (7) habitual, (8) feedback.

Third step was executing a reliability analysis for both, buying behavior items as well as co-creation items (see Appendix III). Perfectionistic buying behavior was reliable with Cronbach’s Alpha 0,716. Habitual buying behavior was unreliable with Cronbach’s Alpha 0,394. After deleting item Q9 the Cronbach’s Alpha became relatively reliable with 0,509. Price-value conscious buying behavior was relatively reliable with Cronbach’s Alpha 0,616. Confused by overchoice buying behavior was reliable with Cronbach’s Alpha 0,761.

Feedback type of co-creation was reliable with Cronbach’s Alpha 0,641. Advocacy type of co-creation was reliable with Cronbach’s Alpha 0,921. Helping type of co-creation was reliable with Cronbach’s Alpha 0,806. Tolerance type of co-creation was reliable with Cronbach’s Alpha 0,665.

Fourth step was making constructs by computing variables. All buying behaviors items were computed into: (1) perfectionistic, (2) habitual, (3) price value, (4) overchoice. All co-creation items were computed into: (1) feedback, (2) advocacy, (3) helping, (4) tolerance. When comparing means it appeared that habitual buying behavior scored highest (M= 4,062; SD=,538), followed by perfectionistic buying behavior (M=3,9; SD= ,559). Price value conscious buying behavior scored lower (M=3,672; SD= ,629) with confused by overchoice buying behavior scoring lowest (M=2,901; SD= ,865). As for the types of co-creation, advocacy scored highest (M=3,835; SD=,904), followed by feedback (M=3,225; SD= ,763).

(22)

tolerance type of co-creation scored lower (M=2,923; SD= ,722) with helping scoring lowest (M=2,777; SD= ,769).

Fifth step, was carrying out residual diagnostics. To test for serial correlations, a Durbin-Watson test was conducted with type of co-creation as dependent variable. For the dependent variable feedback and buying behaviors as predictors the value was 1,924. For the dependent variable advocacy the value was 1,927. For the dependent variable helping the value was 1,922. For the dependent variable tolerance the value was 2,097. Since all values are close to 2, the assumptions have been met. The assumption of homoscedasticity and normality of residuals is checked with the Q-Q-Plot of standardized predicted scores versus standardized residuals. Plots for all types of co-creation indicate that in multiple linear regression analysis there is no tendency in the error terms (see Appendix III). The amount of outliers fell in the range of the expected normal distribution and were therefore not filtered out.

The sixth step was the moderator analysis for which new independent variables were computed (i.e. perfectionism score was weighted by past experience by multiplying both outcomes and dividing it by five, so the newly computed variable MODperfectionistic would be a multiplication of perfectionistic and past experience). A linear regression analyses were carried out with dependent variable being the type of co-creation (see example for advocacy type of co-creation in appendix III). The independent variables were filled into block 1 (buying behaviors and past experience) and block 2 (buying behaviors multiplied by past experience and past experience). The results showed no direct interaction effect between the moderator and all types of co-creation.

(23)

5. Results

5.1 Descriptive statistics

Out of all respondents (N=188) the majority were women (N=136) when compared to men (N=52). The statistics show that most respondents bought a relatively expensive product from Coolblue not too long ago. Their latest purchase was between 50 and 250 euros (M=3,61) and they bought it in the last year (M=1,70). On average they were customers of Coolblue for two or three years (M=2,56). The average age of the respondents was 30 years old (M=30,77) and their household income was between 18.000 and 34.000 euros a year (M=2,88).

The statistics show that most types of buying behaviors were equally spread across the respondents. All behaviors consisted of three 5-likert scale items. Most respondents scored relatively high on perfectionistic type of behavior (M=11,38; SD=1,477), on habitual type of behavior (M=11,01; SD=1,887), and on price-value conscious type of behavior (M=11,01; SD=1,887). Confused by overchoice type of behavior scored lower (M=8,70; SD=2,604).

For the type of co-creation, most respondents scored higher on advocacy (M=11,48; SD=2,73) and helping (M=11,12; SD=3,091) types of co-creation. For feedback (M=9,70; SD=,253) and tolerance (M=8,77; SD=2,172) respondents scored lower. Overall, people were less inclined to provide feedback to Coolblue or be more tolerant, but were more open to referring Coolblue to others or helping other customers. However, it is also noticeable that advocacy and helping were also less equally distributed.

An independent t-test was conducted to compare means for male and female

respondents. The results show a significant effect between gender and confused by overchoice type of behavior (Sig. ,000). Men scored relatively lower (M=2,5; SD=,783) than women (M=3,039; SD=,858) showing that women were more confused by overchoice when buying than men.

(24)

5.2 Multiple Regression Analyses

To test the hypotheses four multiple linear regression analyses per type of co-creation were conducted (see Appendix III). For the variable ‘feedback’ the independent variables

perfectionistic, habitual, confused by overchoice and price-value conscious have no

significant result without the moderator F(4,180) is 1,645 with Sig. ,165. With a moderator there is a trend F(5,179) is 2,363 with Sig. ,042. However, since the moderator is the only significant independent variable, its affect on the model is no longer relevant for this research. Therefore,

(H1) The ‘Perfectionistic’ type of consumer has an affect on the ‘Feedback’ component of co-creation is rejected (P>0.05).

(H1a) Past experience moderates the affect of the ‘Perfectionistic’ type of consumer on the ‘Feedback’ component of co-creation is no longer applicable.

(H2) The ‘Impulsive’ type of consumer has an affect on the ‘Feedback’ component of co-creation is rejected (P>0.05).

(H2a) Past experience moderates the affect of the ‘Impulsive type of consumer on the ‘Feedback’ component of co-creation is no longer applicable.

(H3) The ‘Price-Value Conscious’ type of consumer has an affect on the ‘Feedback’ component of co-creation is rejected (P>0.05).

(H3a) Past experience moderates the affect of the ‘Price-Value Conscious’ type of consumer on the ‘Feedback’ component of co-creation is no longer applicable. (H4) The ‘Confused by overchoice’ type of consumer has an affect on the ‘Feedback’ component of co-creation is rejected (P>0.05).

(H4a) Past experience moderates the affect of the ‘Confused by overchoice’ type of consumer on the ‘Feedback’ component of co-creation is no longer applicable.

(25)

For the variable ‘advocacy’ the independent variables perfectionistic, habitual, confused by overchoice and price-value conscious without the moderator have no significant result

F(4,179) is 2,717 with Sig. ,031. With the moderator there is a significant effect with F(5,178) is 26,608 with Sig. ,000. B(perfectionistic) is ,009, B(habitual) is ,026, B(price value) is -,025 and B(overchoice) is -,052 which all have a slightly negative effect. Again since there is no main effect between the behavioral attitude and the type of co-creation, the effect of the moderator is no longer relevant for this research. Therefore,

(H5) The ‘Confused by overchoice’ type of consumer has an affect on the ‘Advocacy’ component of co-creation is rejected (P>0.05).

(H5a) Past experience moderates the affect of the ‘Confused by overchoice’ type of consumer on the ‘Advocacy’ component of co-creation is no longer applicable. (H6) The ‘Confused by overchoice’ type of consumer has an affect on the ‘Advocacy’ component of co-creation is rejected (P>0.05).

(H8a) Past experience moderates the affect of the ‘Confused by overchoice’ type of consumer on the ‘Advocacy’ component of co-creation is no longer applicable.

For the variable ‘helping’, the independent variables perfectionistic, habitual, confused by overchoice and price-value conscious have no significant result without the moderator

(26)

F(4,179) is ,701 with Sig. ,593. With a moderator there is a significant effect F(5,178) is 3,551 with Sig. ,004. B(perfectionistic) is -,021 which is a slightly negative effect. B(habitual) is ,065 which is a slightly positive effect. B(price value) is -,045 which is slightly negative effect. B(overchoice) is ,056 which is a slightly positive effect. Again, since the moderator is the only significant independent variable, its affect on the model is no longer relevant for this research. Therefore,

(H9) The ‘Confused by overchoice’ type of consumer has an affect on the ‘Helping’ component of co-creation is rejected (P>0.05).

(H9a) Past experience moderates the affect of the ‘Confused by overchoice’ type of consumer on the ‘Helping’ component of co-creation is no longer applicable.

For the variable ‘tolerance’, the independent variables perfectionistic, habitual, confused by overchoice and price-value conscious have no significant result with or without moderator

(27)

past experience with F(4,179) is ,741 met Sig. ,565 without moderator. With moderator F(5,178) is 1,802 with Sig. ,115. Therefore,

(H13) The ‘Confused by overchoice’ type of consumer has an affect on the ‘Tolerance’ component of co-creation is rejected (P>0.05).

(H13a) Past experience moderates the affect of the ‘Confused by overchoice’ type of consumer on the ‘Tolerance’ component of co-creation is rejected (P>0.05).

5.3 Correlation analysis

Since the original models did not yield any significant or relevant results, other statistical methods have been applied to find out whether some hidden relations could be found. First, a bivariate correlation analysis between all variables was conducted. All items, such as all behavioral attitudes, type of co-creation and past experience, were included (see Appendix III).

(28)

The analysis shows that co-creation constructs strongly correlate with each other. Also, past experience correlates with all types of co-creation, while only with perfectionistic and habitual type of behavior. These results will further be interpreted in the discussion section.

5.4 Backward regression analysis

A backward regression analysis which included interaction variables was also added to this research. The dependent variable was the past experience, while the independent variables were all behavioral attitudes and all types of co-creation.

Backward regression analysis shows in model 6 an effect between past experience and perfectionistic behavior (Sig. ,030), past experience and habitual behavior (Sig. ,018), and past experience and advocacy type of co-creation (Sig. ,000) (see Appendix III). These results will further be interpreted in the discussion section.

(29)

6. Discussion, limitations and further research

As we can see from the data analysis there are no significant results on the effect between the type of behavior and the type of co-creation. So to answer the research question, the types of consumer behavior have no significant effect on the types of co-creation. Past experience as moderator does have an effect in most cases, but since the moderator is the only significant independent variable, its affect on the model is no longer relevant for this research.

In addition, there is no effect whatsoever (with and without the moderator) between buying behaviors and tolerance type of co-creation. This suggests that even though a

consumer has had a positive experience with Coolblue before, it will in no possible way affect consumer’s tolerance towards Coolblue.

Lately there have been many discussions about soft, difficult to measure, factors that have an effect on customer’s engagement. In the age of consumerism, the most important question is how can customer add more value to the company and what the company should do in order to increase customer’s readiness with regard to the brand. I have tried to

operationalize types of co-creation by Yi and Gong (2012) by seeking interaction with attitudinal behaviors. However, as there are no significant results, there are few things that can be questioned. First, Yi and Gong mention four types of voluntary co-creation, but as we see in the correlation matrix those components are strongly correlated with each other. Therefore, the question rises whether co-creation according to Yi and Gong (2012) does indeed consist of those four types and whether or not it should be regrouped differently. Secondly, for the attitudinal behaviors I have chosen characteristics mentioned by Ivanova (2016). However, her case study was based on GlassShopWall, a demo store in a shopping mall in the Hague. It is plausible to assume that there is a difference between shopping behaviors in a physical store and online. Customers characteristics may come through more pronounced when the customer is alone behind a computer in his own home

(30)

rather than when he is in a physical store. For example, a customer might spend hours searching for best price-quality ratio and the best deal at home, but will not want to appear cheap when confronted by a shop assistant.

In addition, it can be questioned whether this manner of measurement was the right choice to do. A survey might be too impersonal and also might have been too long, as we see from the total of incomplete responses. In an experiment setting, monitoring people the way they shop and browse Coolblue might have yielded different results.

Another limitation of this study is that it is only limited to the Netherlands. It is possible that customers in different countries display different patterns of buying behavior or co-creation when it comes to online retailers. Some developing countries for example do not even have the same variety of online retailers (or products) to be confused in the first place. In some cultures, it is less common to advocate a brand (or product) and people are more

reserved. Also, variables were translated into Dutch and might have been interpreted

differently. Even though the translated versions were checked by three Dutch natives before conducting the study, there is still room for possible misinterpretations.

Also, it is fair to question whether behavioral attitudes were the right predictors from the beginning. However, with limited literature in mind, it was difficult to predict beforehand whether there could be a relation between buying behaviors and readiness to co-create until further researched.

Backward regression analysis shows that past experience has an effect with

perfectionistic and habitual types of behaviors and also advocacy type of co-creation. In other words, somebody who found the best quality product or is just a very loyal customer, will indeed be a happy customer and vice versa. Consequently, that customer would be happy to tell others about that product or brand.

(31)

So, the results show that past experience has a moderating effect, however, is not just a simple view that positive past experience can be a predictor for most variables? And, that past experience is in turn a result of good products and services and good branding strategy? So in short, is co-creation, as a construct, not just a hype? At least it it does not hold up when researched in the given setting.

Future research might consider different predictors to operationalize types of creation by Yi and Gong (2012). Future research might also question these types of co-creation and see if these can be regrouped. It is also advisable to conduct a study outside the Netherlands or with international customers. Future research can also examine whether there is indeed a difference between buying behaviors by Ivanova (2016) in a physical store versus an online setting.

(32)

7. Conclusion

Multiple regression analyses have showed that the types of consumer behavior have no significant effect on the types of co-creation. Past experience turned out to have a moderating role, but due to lack of main effect is no longer relevant. The results might not be significant, but have showed no interaction effect between two existing theories. Therefore, future

research into customer engagement might explore co-creation with different predictors, or in a different country or setting. The results also show that the types of co-creation by Yi and Gong (2012) are too strongly correlated and additional research might be needed.

(33)

8. References

Bijmolt, T. H. A., Leeflang, P. S. H., Block, F., Eisenbeiss, M., Hardie, B. G. S., Lemmens, and Saffert, P. (2010). Analytics for customer engagement. Journal of Service Research, 13, 341-356.

Bowden, J. L. H. (2009). The process of customer engagement: A conceptual framework. Journal of Marketing Theory and Practice, 17, 63-74.

Brodie, R. J., Hollebeek, L. D., Juric, B. and Ilic, A. (2011). Customer engagement:

Conceptual domain, fundamental propositions, and implications for research. Journal of Service Research, 14, 252-271.

CBS. (2016). Bijna een op vijf koopt levensmiddelen online. Retrieved from

https://www.cbs.nl/nl-nl/nieuws/2016/47/bijna-een-op-vijf-koopt-levensmiddelen-online

Coolblue. (2016). Home pagina. Retrieved from https://www.coolblue.nl Emerce. (2016). CBS: 70 duizend webwinkels in Nederland. Retrieved from

https://www.emerce.nl/nieuws/cbs-70-duizend-webwinkels-nederland Fung So, K.K., King, C. and Sparks, B. (2014). Customer engagement with tourism

brands: scale development and validation. Journal of Hospitality & Tourism Research, Vol.38, No.3, 304-329.

Hollebeek, L. D. (2009). Demystifying customer engagement:

Toward the development of a conceptual model. Paper presented at the ANZMAC 2009 conference, Monash University, Melbourne, Australia.

Hollebeek, L. D. (2011). Demystifying customer brand engagement: Exploring the loyalty nexus. Journal of Marketing Management, 27(7/8), 1-23.

Grönroos, C., & Voima, P. (2013). Critical service logic: making sense of value creation and co-creation. Journal of the academy of marketing science, 41(2), 133-150.

(34)

Ivanova, D. (2016). Digital shopping displays: and the future of retail. (Unpublished master’s thesis). University of Amsterdam, Amsterdam, the Netherlands.

Kumar, V. (2008a). Managing Customers for Profit. Upper Saddle River, NJ: Wharton School Publishing.

Kumar, V. (2008b). Customer Lifetime Value: The Path to Profitability. The Netherlands: Now Publishers.

Kumar, V., Petersen, J. A. and Leone, R. P. (2007). ‘‘How Valuable Is Word of Mouth?’’ Harvard Business Review, 85 (October), 139-146.

Kumar, V., Petersen, J. A. and Leone, R. P. (2010). ‘‘Driving Profitability by Encouraging Customer Referrals: Who, When, and How,’’ Journal of Marketing: September 2010, Vol. 74, No. 5, pp. 1-17.

Kumar, V., Aksoy.L,, Donkers B., Venkatesan R., Wiesel T. and Tillmanns S. (2010). ‘‘Undervalued or Overvalued Customers: Capturing Total Customer Engagement Value,’’ Journal of Service Research, 13 (3), 297-310.

Lemke, F. Clark, M. and Wilson, H. (2010).Customer experience quality: an exploration in business and consumer contexts using repertory grid technique. Journal of the Academy of Marketing Science, 39(6), 846-869.

MSI (2010), 2010-2012 Research Priorities. Boston, MA: Marketing Science Institute. Payne, A., Storbacka, K., and Frow, P. (2008). Managing the co-creation of value. Journal of

the Academy of Marketing Science, 36, 83–96.

Prahalad, C. K., & Ramaswamy, V. (2004). Co-creation experiences: The next practice in value creation. Journal of interactive marketing, 18(3), 5-14.

van Doorn, J., Lemom, K. N., Mittal, V., Nass, S., D., P., Pirner, P. and Verhoef, P. C. (2010). Customer engagement behaviour: Theoretical foundations and reserach directions.

(35)

Journal of Service Research, 13, 253-266.

Verhoef, P.C. and Lemon, K.N. (2012). Successful customer value management: Key lessons and emerging trends. European Management Journal, 31, 1-15.

Verhoef, P. C., Reinartz, W. and Krafft, M. (2010). Customer engagement as a new perspective in customer management. Journal of Service Research, 13, 247-252. Witell, L., Kristensson, P., Gustafsson, A., & Löfgren, M. (2011). Idea generation: customer

co-creation versus traditional market research techniques. Journal of Service

Management, 22(2), 140-159.

Woodruff, R. B. (1997). Customer value: The next source for competitive advantage. Journal of the Academy of Marketing Science, 25(2), 139–152.

Yi, Y. and Gong, T. (2013). Customer value co-creation behavior: scale development and validation. Journal of Business Research, 66, 1279-1284.

(36)

9. Appendices

(37)

(38)

Appendix III: Data Analysis Rotated Component Matrix

Component

1 2 3 4 5 6 7 8

Ik koop liever bij winkels die het beste kwaliteit bieden

,105 -,073 ,058 ,742 ,047 -,053 -,007 ,037

Winkels van hoge kwaliteit zijn belangrijk voor mij

,046 ,017 -,019 ,835 ,000 ,013 ,135 ,065

Ik doe mijn best om kwalitatief de beste winkels te kiezen

-,007 ,129 -,048 ,789 ,039 ,062 ,028 ,024

Zodra ik een winkel leuk vind, koop ik er regelmatig

-,072 ,041 ,236 -,014 ,115 -,022 ,101 ,693

Ik heb favoriete winkels waar ik vaak koop

,140 ,068 -,153 ,154 -,025 ,058 ,072 ,835

Ik wissel regelmatig van winkels waar ik vaak koop

-,156 -,049 -,164 ,176 -,097 -,127 ,593 ,116

Ik koop zoveel mogelijk met korting

-,069 ,030 -,073 -,013 -,037 ,864 -,048 -,016 Ik probeer altijd de

beste prijs-kwaliteit verhouding te vinden

,123 -,065 ,096 ,138 ,066 ,795 -,130 ,166

Ik kies meestal voor de lager geprijsde

producten

-,001 ,082 ,348 -,183 ,133 ,559 ,156 -,210

De hoeveelheid keuze die er is, verwart mij wel eens

,071 -,106 ,814 -,098 -,052 ,077 ,052 ,162

Ik vind het soms moeilijk om te kiezen waar ik moet winkelen

-,053 ,015 ,849 ,000 -,040 ,005 -,212 ,012

Hoe meer ik over producten of winkels lees, hoe moeilijker ik het vind om de juiste keuze te maken

-,205 ,080 ,746 ,108 ,063 ,059 ,029 -,073

Als ik een goed idee heb voor Coolblue, laat ik dat weten

,271 ,272 ,101 -,005 ,137 -,036 ,676 ,035

Als de service van Coolblue goed is, laat ik dat weten

,377 ,145 -,044 ,019 ,143 ,064 ,704 ,063

Als ik een probleem heb met Coolblue, zou ik dat melden

,634 ,007 -,058 -,062 -,092 ,004 ,018 ,121

Ik heb positieve dingen gezegd over Coolblue tegen anderen

(39)

Ik heb Coolblue aanbevolen aan anderen

,855 ,276 -,071 ,104 ,111 ,035 ,099 -,058

Ik raad mijn vrienden en familie aan om bij Coolblue te winkelen ,762 ,342 ,001 ,142 ,054 ,066 ,159 ,016 Ik zou andere consumenten van Coolblue helpen ,272 ,767 ,021 ,069 ,044 ,067 ,181 ,099 Als andere consumenten een probleem ervaren met Coolblue, ben ik bereid om te helpen

,010 ,843 ,038 ,012 ,092 ,066 ,004 -,021

Ik leg anderen uit hoe Coolblue werkt ,435 ,695 -,025 -,018 ,047 -,181 -,053 ,012 Ik geef andere consumenten van Coolblue advies ,118 ,722 -,037 ,009 ,112 -,003 ,134 ,042

Als Coolblue een keer niet naar de

verwachting is, vind ik dat niet zo erg

-,122 ,106 ,116 ,058 ,784 ,023 ,098 -,049

Als Coolblue een keer een fout maakt, ben ik geduldig

,191 ,097 -,164 -,011 ,750 ,045 ,063 -,023 Ik vind het geen ramp

om iets langer te wachten op een bestelling van Coolblue dan ik had verwacht

,112 ,060 ,025 ,040 ,729 ,022 -,038 ,152

Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. a. Rotation converged in 6 iterations.

Residual Diagnostics

Residual Statistics: Feedback

Model Summaryb

Model R R Square

Adjusted R Square

Std. Error of

the Estimate Durbin-Watson

1 ,183a ,034 ,010 2,20637 1,924

a. Predictors: (Constant), Overchoice, Perfectionistic, Pricevalue, Habitual b. Dependent Variable: Feedback

Residual statistics

Minimum Maximum Mean Std. Deviation N

Predicted Value 8,5095 10,7880 9,7384 ,40654 172

Residual -6,41757 5,15445 ,00000 2,18041 172

Std. Predicted Value -3,023 2,582 ,000 1,000 172

(40)

(41)

Residual statistics Advocacy Model Summaryb Model R R Square Adjusted R Square Std. Error of

1 ,235a ,055 ,033 2,68703 1,927

a. Predictors: (Constant), Overchoice, Perfectionistic, Pricevalue, Habitual b. Dependent Variable: Advocacy

Residuals Statisticsa

Predicted Value 9,6663 13,6141 11,5146 ,64246 171

Residual -9,19194 4,55395 ,00000 2,65523 171

Std. Residual -3,421 1,695 ,000 ,988 171

(42)

Residual statistics Helping Model Summaryb Model R R Square Adjusted R Square Std. Error of

1 ,093a ,009 -,015 3,13796 1,922

a. Predictors: (Constant), Overchoice, Perfectionistic, Pricevalue, Habitual b. Dependent Variable: Helping

Predicted Value 10,2762 11,9690 11,1744 ,28968 172

Residual -7,12583 8,48431 ,00000 3,10105 172

Std. Residual -2,271 2,704 ,000 ,988 172

(43)

Residual statistics Tolerance Model Summaryb Model R R Square Adjusted R Square Std. Error of

1 ,123a ,015 -,008 2,17273 2,097

a. Predictors: (Constant), Overchoice, Perfectionistic, Pricevalue, Habitual b. Dependent Variable: Tolerance

Predicted Value 8,0987 9,4993 8,8035 ,26667 173

Residual -5,85722 6,03499 ,00000 2,14732 173

(44)

(45)

Moderator analysis Feedback ANOVA:

Model Sum of Squares df Mean Square F Sig.

1 Regression 6,657 5 1,331 2,363 ,042b Residual 100,876 179 ,564 Total 107,534 184 2 Regression 9,448 9 1,050 1,873 ,059c Residual 98,086 175 ,560 Total 107,534 184

a. Dependent Variable: CC_Feedback

b. Predictors: (Constant), Hoe tevreden bent u met Coolblue?, Beh_PriceValue, Beh_Overchoice, Beh_Perfectionistic, Beh_Habitual

c. Predictors: (Constant), Hoe tevreden bent u met Coolblue?, Beh_PriceValue, Beh_Overchoice,

Beh_Perfectionistic, Beh_Habitual, MODoverchoice, MODperfectionistic, MODpricevalue, MODhabitual

Coefficientsa

Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) 1,421 ,658 2,160 ,032 Beh_Perfectionistic ,119 ,103 ,087 1,160 ,248 Beh_Habitual ,119 ,108 ,083 1,099 ,273 Beh_PriceValue ,001 ,090 ,001 ,013 ,990 Beh_Overchoice ,001 ,066 ,001 ,011 ,991 Hoe tevreden bent u met

Coolblue? ,194 ,086 ,174 2,255 ,025 2 (Constant) -1,145 4,461 -,257 ,798 Beh_Perfectionistic -,801 ,663 -,587 -1,208 ,229 Beh_Habitual ,705 ,754 ,492 ,936 ,351 Beh_PriceValue 1,113 ,628 ,918 1,773 ,078 Beh_Overchoice -,157 ,405 -,175 -,387 ,699 Hoe tevreden bent u met

Coolblue? ,778 ,994 ,699 ,783 ,435 MODperfectionistic ,215 ,149 1,113 1,443 ,151 MODhabitual -,131 ,168 -,697 -,780 ,437 MODpricevalue -,261 ,144 -1,348 -1,819 ,071 MODoverchoice ,035 ,091 ,189 ,389 ,698

Excluded Variablesa

Model Beta In t Sig. Partial Correlation

Collinearity Statistics Tolerance 1 MODperfectionistic ,817b _1,152 _,251 _,086 _,010 MODhabitual -,084b _-,099 _,921 _-,007 _,007 MODpricevalue -1,220b _-1,672 _,096 _-,124 _,010 MODoverchoice ,005b ,010 ,992 ,001 ,023

b. Predictors in the Model: (Constant), Hoe tevreden bent u met Coolblue?, Beh_PriceValue, Beh_Overchoice, Beh_Perfectionistic, Beh_Habitual

(46)

Appendix IV: Results Descriptive Statistics N Minimu m Maximu m Mean Std. Deviation Skewness Kurtosis Statistic Statistic Statistic Statistic Statistic Statistic Std.

Error

Statistic Std. Error Hoe duur was uw

laatste aankoop?

188 1 5 3,61 1,199 -,160 ,177 -1,448 ,353 Hoe lang geleden

hebt u voor het laatst iets gekocht bij Coolblue?

187 1 5 1,70 1,004 1,319 ,178 ,894 ,354

Hoeveel jaar bent u al klant van Coolblue?

182 ,00 10,00 2,5604 1,89625 1,535 ,180 3,216 ,358

Bent u een man of een vrouw? 188 1 2 1,72 ,449 -1,007 ,177 -,997 ,353 Wat is uw leeftijd? - 1 188 20,00 77,00 30,7713 9,97997 1,573 ,177 2,499 ,353 Wat is uw gezinsinkomen op jaarbasis? 187 1 5 2,88 1,551 ,050 ,178 -1,520 ,354 Valid N (listwise) 181

Independent t-test: interaction gender and buying behaviors

Group Statistics

Bent u een man of een vrouw? N Mean Std. Deviation Std. Error Mean

Beh_Perfectionistic Man 52 3,9679 ,56112 ,07781 Vrouw 136 3,8664 ,55808 ,04785 Beh_Habitual Man 52 4,1058 ,58859 ,08162 Vrouw 136 4,0441 ,51450 ,04412 Beh_PriceValue Man 52 3,6122 ,58085 ,08055 Vrouw 135 3,6840 ,66082 ,05687 Beh_Overchoice Man 52 2,5449 ,78390 ,10871 Vrouw 134 3,0398 ,85844 ,07416

Independent Samples Test

Levene's Test for Equality of Variances

t-test for Equality of Means F Sig. t df Sig. (2-tailed) Mean Differen ce Std. Error Differen ce 95% Confidence Interval of the Difference Lower Upper Beh_Perfe ctionistic Equal variances assumed ,025 ,875 1,114 186 ,267 ,10153 ,09113 -,07825 ,28130 Equal variances not assumed 1,111 91,90 7 ,269 ,10153 ,09135 -,07991 ,28296 Beh_Habit ual Equal variances assumed 1,399 ,238 ,706 186 ,481 ,06165 ,08736 -,11070 ,23400

(47)

Equal variances not assumed ,664 82,49 2 ,508 ,06165 ,09278 -,12291 ,24621 Beh_Price Value Equal variances assumed 1,033 ,311 -,687 185 ,493 -,07177 ,10442 -,27778 ,13423 Equal variances not assumed -,728 104,6 30 ,468 -,07177 ,09860 -,26729 ,12375 Beh_Overc hoice Equal variances assumed ,130 ,719 -3,613 184 ,000 -,49493 ,13699 -,76519 -,22466 Equal variances not assumed -3,761 101,1 15 ,000 -,49493 ,13159 -,75597 -,23389

Independent t-test: interaction gender and types of co-creation

Group Statistics

Bent u een man of een vrouw? N Mean Std. Deviation Std. Error Mean

CC_Feedback Man 52 3,1859 ,86171 ,11950 Vrouw 134 3,2400 ,72418 ,06256 CC_Advocacy Man 51 4,0000 ,88192 ,12349 Vrouw 134 3,7724 ,90703 ,07836 CC_Helping Man 51 2,8284 ,82839 ,11600 Vrouw 134 2,7575 ,74714 ,06454 CC_Tolerance Man 51 2,9412 ,79918 ,11191 Vrouw 134 2,9167 ,69301 ,05987

Independent Samples Test

Levene's Test for Equality of Variances

t-test for Equality of Means F Sig. t df Sig. (2-tailed) Mean Differen ce Std. Error Differen ce 95% Confidence Interval of the Difference Lower Upper CC_Fee dback Equal variances assumed 1,349 ,247 -,433 184 ,665 -,05415 ,12495 -,30067 ,19237 Equal variances not assumed -,401 80,46 9 ,689 -,05415 ,13488 -,32255 ,21425 CC_Adv ocacy Equal variances assumed ,519 ,472 1,537 183 ,126 ,22761 ,14812 -,06462 ,51985 Equal variances not assumed 1,556 92,71 3 ,123 ,22761 ,14625 -,06283 ,51805 CC_Hel ping Equal variances assumed 2,746 ,099 ,560 183 ,576 ,07097 ,12672 -,17905 ,32099 Equal variances not assumed ,535 82,77 0 ,594 ,07097 ,13274 -,19307 ,33500

(48)

CC_Tole rance Equal variances assumed 2,154 ,144 ,206 183 ,837 ,02451 ,11905 -,21038 ,25940 Equal variances not assumed ,193 80,24 3 ,847 ,02451 ,12692 -,22805 ,27707

Multiple Regression analyses

Multiple Regression analysis for Feedback

Model Summaryc:

Model R R Square Adjusted R

Square Std. Error of the Estimate Change Statistics R Square Change

F Change df1 df2 Sig. F Change 1 ,188a _,035 _,014 _,75917 _,035 _1,645 ₄ ₁₈₀ _,165

2 ,249b ,062 ,036 ,75070 ,027 5,083 1 179 ,025 a. Predictors: (Constant), Beh_Overchoice, Beh_Habitual, Beh_PriceValue, Beh_Perfectionistic b. Predictors: (Constant), Beh_Overchoice, Beh_Habitual, Beh_PriceValue, Beh_Perfectionistic, Hoe tevreden bent u met Coolblue?

c. Dependent Variable: CC_Feedback

ANOVAa

1 Regression 3,793 4 ,948 1,645 ,165b Residual 103,741 180 ,576 Total 107,534 184 2 Regression 6,657 5 1,331 2,363 ,042c Residual 100,876 179 ,564 Total 107,534 184

b. Predictors: (Constant), Beh_Overchoice, Beh_Habitual, Beh_PriceValue, Beh_Perfectionistic c. Predictors: (Constant), Beh_Overchoice, Beh_Habitual, Beh_PriceValue, Beh_Perfectionistic, Hoe tevreden bent u met Coolblue?

Multiple Regression analysis for Advocacy

Model Summaryc

Model R R Square Adjusted R Square Std. Error of the Estimate Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 ,239a _,057 _,036 _,88743 _,057 _2,717 ₄ ₁₇₉ _,031 2 ,654b _,428 _,412 _,69335 _,370 _{115,238 1} ₁₇₈ _,000

a. Predictors: (Constant), Beh_Overchoice, Beh_Habitual, Beh_PriceValue, Beh_Perfectionistic b. Predictors: (Constant), Beh_Overchoice, Beh_Habitual, Beh_PriceValue, Beh_Perfectionistic, Hoe tevreden bent u met Coolblue?

c. Dependent Variable: CC_Advocacy

ANOVAa

1 Regression 8,558 4 2,140 2,717 ,031b Residual 140,968 179 ,788 Total 149,526 183 2 Regression 63,957 5 12,791 26,608 ,000c Residual 85,570 178 ,481 Total 149,526 183

(49)

a. Dependent Variable: CC_Advocacy

Multiple Regression analysis for Helping

Model Summaryc

Model R R Square Adjusted R Square Std. Error of the Estimate Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 ,124a _,015 _-,007 _,76802 _,015 _,701 ₄ ₁₇₉ _,593 2 ,301b _,091 _,065 _,74014 _,075 _{14,739 1} ₁₇₈ _,000

c. Dependent Variable: CC_Helping

ANOVAa

1 Regression 1,653 4 ,413 ,701 ,593b Residual 105,583 179 ,590 Total 107,236 183 2 Regression 9,727 5 1,945 3,551 ,004c Residual 97,509 178 ,548 Total 107,236 183

a. Dependent Variable: CC_Helping

Multiple Regression analysis for Tolerance

Model Summaryc

Model R R Square Adjusted R Square Std. Error of the Estimate Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 ,128a ,016 -,006 ,72245 ,016 ,741 4 179 ,565 2 ,219b _,048 _,021 _,71265 _,032 _5,962 ₁ ₁₇₈ _,016

c. Dependent Variable: CC_Tolerance

ANOVAa

1 Regression 1,547 4 ,387 ,741 ,565b Residual 93,427 179 ,522 Total 94,974 183 2 Regression 4,575 5 ,915 1,802 ,115c Residual 90,400 178 ,508 Total 94,974 183

a. Dependent Variable: CC_Tolerance

b. Predictors: (Constant), Beh_Overchoice, Beh_Habitual, Beh_PriceValue, Beh_Perfectionistic c. Predictors: (Constant), Beh_Overchoice, Beh_Habitual, Beh_PriceValue, Beh_Perfectionistic, Hoe