How can starting entrepreneurs make the most out
of co-creation with their customers?
Initial findings on the effects of commitment, curiosity, citizenship,
extraversion and generosity on co-creative behaviour.
Thesis Daniël Michiel Fransman FINAL VERSION 1.00
Author: Daniël Michiel Fransman
Amsterdam Business School, University of Amsterdam
Executive Programme in Management Studies – Strategy Track Student number: 10475400
Date: 27-01-2015
Statement of Originality
This document is written by student Daniël Michiel Fransman 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.
Signature:
Index
Abstract ... 7 1. Introduction ... 8 2. Literature Review ... 11 2.1 Co-Creation ... 11 2.1.1 Definition of co-creation ... 11 2.1.2 Outcomes of co-creation ... 122.1.3 When and how to co-create ... 13
2.1.4 Who to co-create with ... 14
2.1.5 Challenges and risks of co-creation ... 16
2.2 Commitment ... 17 2.2.1 Commitment ... 17 2.2.2 Hypothesis development ... 19 2.3 Personality ... 20 2.3.1 Personality ... 21 2.3.2 Curiosity ... 22 2.3.3 Citizenship ... 22 2.3.4 Extraversion ... 23 2.3.5 Generosity ... 24 2.4 Conceptual Model ... 25 3 Method ... 26 3.1 Procedure ... 26 3.2 Data collection ... 26
3.2.1 Population and Sample ... 26
3.3 Measures ... 27
3.3.1 Co-creative behaviour ... 28
3.3.2 Commitment ... 30
4. Results ... 34 4.1 Correlation ... 34 4.2 Factor Analysis ... 37 4.2.1 Case 1 ... 38 4.2.2 Cases 2, 3 and 4 ... 39 4.2.3 Combined results ... 39 4.3 Paired-Samples T-test ... 40 4.4 Regression ... 43 4.4.1 Feedback ... 44 4.4.2 Information Sharing ... 46 4.4.3 Effort ... 49 4.5 ANOVA ... 52 4.5.1 Age Group ... 52 4.5.2 Profession ... 53 5. Discussion ... 55 5.1 Conclusion ... 55 5.3 Implications ... 55 5.2 limitations ... 56 5.3 Future research ... 57 Reference list ... 59
Appendix I: General strengths and limitations of the questionnaire technique ... 65
Appendix II: Overview of question translations ... 67
Co-Creation ... 67
Commitment ... 68
Personality ... 68
Control Variables ... 69
Appendix III: Questionnaire ... 70
Appendix IV: Non significant ANOVA results for gender and education ... 88
Education ... 88
Appendix V: Results Principal Component Analysis ... 89
List of tables and figures
Table 1: The commitment types that were described per case in the questionnaire. ... 30Table 2: Means, Standard Deviations and Correlations. ... 35
Table 3: Average eigenvalues and actual eigenvalues per case. ... 38
Fig. 1: Mean of feedback, information sharing and effort per case. ... 43
Table 4: Hierarchical Regression Model of Feedback with high normative and affective commitment (Case 1). ... 44
Table 5: Hierarchical Regression Model of Feedback with high normative and low affective commitment (Case 2). ... 45
Table 6: Hierarchical Regression Model of Feedback with low normative and high affective commitment (Case 3). ... 45
Table 7: Hierarchical Regression Model of Feedback with low normative and affective commitment (Case 4). ... 46
Table 8: Hierarchical Regression Model of Information Sharing with high normative and affective commitment (Case 1). ... 46
Table 9: Hierarchical Regression Model of Information Sharing with high normative and low affective commitment (Case 2). ... 47
Table 10: Hierarchical Regression Model of Information Sharing with low normative and high affective commitment (Case 3). ... 48
Table 11: Hierarchical Regression Model of Information Sharing with low normative and affective commitment (Case 4). ... 48
Table 12: Hierarchical Regression Model of Effort with high normative and affective commitment (Case 1). ... 49
Table 13: Hierarchical Regression Model of Effort with high normative and low affective commitment (Case 2). ... 50
Table 14: Hierarchical Regression Model of Effort with low normative and high affective commitment (Case 3). ... 50
Table 15: Hierarchical Regression Model of Effort with low normative and affective commitment (Case 4). ... 51
Table 17: Mean and standard deviation for age group in case 1 and 2. ... 53 Table 18: ANOVA results for the effect of profession on co-creative behaviour per case. .... 54 Table 19: Mean and standard deviation for profession in case 4. ... 54 !
Acknowledgements
The researcher would like to thank all the participants of this research. Furthermore, a special thank you goes out to Eva van Ingen, for helping with the translation of the questions and to Judith van Ingen, for giving guidance and feedback.
Abstract
This investigation reports initial findings on the effects of commitment, curiosity, citizenship, extraversion and generosity on co-creation behavior when customers are asked to co-create with a starting entrepreneur. Co-creation was measured using three different factors: the willingness to give feedback, to share information and to give effort. Results indicate that both normative as well as affective commitment have a positive effect on co-creative behaviour, with normative commitment scoring higher than affective commitment. This research also shows that curiosity has a positive effect on the feedback that customers give and the information they share to a starting entrepreneur. However, no effects were found between citizenship, extraversion and generosity, indicating that they do not influence a customer’s co-creative behaviour. Furthermore, curiosity, citizenship, extraversion and generosity were found to not effect the effort a customer is willing to give when co-creating with a starting entrepreneur. However, a person’s age-group and profession were found to have an effect on effort.
1. Introduction
Imagine yourself waking up with a brilliant idea. You feel exited and ready to conquer the world. But then the questions come to mind. How are you going to develop it, and with who? How do you know for who the product will be? Where and when will you launch the product? These are some of the questions that starting entrepreneurs deal with.
During the early stages of developing a new product or venture, entrepreneurs usually look for resources that can help them with their idea. They do this by consulting their networks and by assessing what resources (e.g. knowledge, skills) they have access to (Sullivan & Ford, 2014). Integrating customers into the value creating process is an important way to reduce uncertainty for firms (Ernst, 2002). Because of this uncertainty, a starting entrepreneur is most likely to develop a new product or idea in an area that he or she is familiar with or at least is closely linked to and the personal characteristics of the entrepreneur draw him or her towards some industries and away from others (Lofstrom et al., 2014). Furthermore, it is found that successful entrepreneurs use customers and the competences of their customers (Prahalad & Ramaswamy, 2000). For starting entrepreneurs, the use of (potential) customers in early stages of the development of their product, can be very useful in reducing the uncertainty they face. But, as will be discuss in the next section, finding the right customers to do this with, can be somewhat of a challenge for them.
The identification of an appropriate set of customers to co-create with is a key challenge for starting entrepreneurs who want to co-create and which type of customer is most likely to reveal their ideas and how they will co-create, depends on certain characteristics of the underlying industry or firm (Nambisan, 2002; Poetz & Schreier, 2012). Access to resources is an important challenge for starting entrepreneurs (Sullivan & Ford, 2014). To deal with this challenge, starting entrepreneurs engage the resources and network available to them (Baker & Nelson, 2005; Sullivan & Ford, 2014, Mollick, 2014). They aim to organize their resources
to meet demands that emerge as they launch new ventures (Sullivan & Ford, 2014). The literature on co-creation tells us that lead users, hobbyists and customers that are connected to a firm in some way are the best individuals to co-create with (Jeppesen & Frederiksen, 2006; Nishikawa, Schreier & Ogawa, 2013). But, this knowledge comes from studies that have been done in large and existing firms, who usually used this value creating process as a marketing tool (Poetz & Schreier, 2012; Nishikawa, Schreier & Ogawa, 2013). Research has also been done on the benefits, risks and organization of the co-creation process. (Lilien et al., 2002; Thomke & von Hippel, 2002; Nambisan, 2002; Nishikawa, Schreier & Ogawa, 2013). These researches typically have been done with larger and well-know firms and with a focus on the co-creative experience of a customer. Starting entrepreneurs typically do not have the luxury of a big pool of customers from were they can select the best customers to co-create with, they usually only have they own small network of resources (Baker & Nelson, 2005; Sullivan & Ford, 2014). For starting entrepreneurs, finding the right customers to co-create with will therefore be a challenge.
This brings up the research question of this thesis: How can starting entrepreneurs make the most out of co-creation with their customers?
As stated earlier, a starting a starting entrepreneur’s personal network is an important factor in the success of co-creating activities, indicating that customers close to the entrepreneur can provide good ideas (Baker & Nelson, 2005; Sullivan & Ford, 2014, Mollick, 2014). This research uses these results to test if the personal ties of a customer to a starting entrepreneur or the idea of the entrepreneur can predict the co-creative behaviour of a customer.
Furthermore Ajzen (2005) found that a person’s personality influences their behaviour, indicating that customer’s personal characteristics can predict their co-creative behaviour. This research tests if certain personal characteristics indeed influence the co-creative
behaviour of customers, if they are asked to help a starting entrepreneur. The personal characteristics that were tested in this research are curiosity, citizenship, extraversion and generosity.
This research investigates if commitment and personal characteristics play a role in the interaction between a starting entrepreneur and their co-creating customers. Since starting entrepreneurs usually cannot just contact their customers, because they do not have customers to start with, it is extremely difficult for them to start co-creating from the beginning of their venture. If they would use a scattergun approach to find customers to co-create with, the chance of finding the right customers would be slim. The aim of this research is to identify specific conditions and characteristics that starting entrepreneurs can use to their benefit to select the customers that will most likely help them the most.
The respondents for this research are gathered via a convenient sample and this research aims to draw conclusions based on the results of an online questionnaire, which are analysed using SPSS software (Field, 2009).
This thesis is structured as follows: First, an overview of the literature on co-creation, commitment and the four personality dimensions are given. Then the hypotheses, the conceptual model and the research method are discussed, followed by a quantitative analysis of the data and the findings of the analysis. Finally, the findings are discussed and conclusions, limitations and future research possibilities are outlined.
2. Literature Review
In this chapter, the literature on co-creation, commitment and personality are discussed. Relevant articles are analysed and their implications on this research are discussed. Firstly the literature on co-creation is discussed in order to indicate what we already know about it. Secondly, commitment followed by the four personality characteristics curiosity, citizenship, extraversion and generosity are discussed, together with their hypothesised link to co-creative behaviour.
2.1 Co-Creation
2.1.1 Definition of co-creation
The interaction between the consumer and a firm in a shared value creation process is often referred by as: ‘Co-creation’ (Prahalad & Ramaswamy, 2004; Ramaswamy & Gouillart, 2010; Gebauer, Füller & Pezzei, 2013; Handrich & Heidenreich, 2013; Yi & Gong, 2013). Co-creation is about putting the customer’s experience and idea’s at the centre of the design of a product or service and this shared creation serves the interests of both the customer and the firm (Ramaswamy & Gouillart, 2010). Compared to internal sourcing or external contracting, co-creation seems to be a good mechanism for innovation and problem solving, where the source of a product lies further away from the firm (Afuah & Tucci, 2012).
A user-centred innovation process like co-creation, where customers can contribute to the creation of products and they can benefit from innovations developed and shared by others, offer great advantages over the manufacturer-centric innovation process, where customers rely on firms to develop products (von Hippel, 2005). For instance, customers are found to complement the work of a firm and its professionals in idea generation (Poetz & Schreier, 2012). Furthermore it is found that keeping co-creation sessions with your close network is a way to actively improve a product or service (Hoyer et al., 2010).
2.1.2 Outcomes of co-creation
The recent years, research has been done to analyse the outcomes of co-creation. In an extensive study by Nishikawa, Schreier and Ogawa (2013) of customer-generated products compared to firm-generated products at Muji, a Japanese retail company which sells household and consumer goods, the researchers found that the customer-generated products outperform the firm-generated products in regard to sales (approximately twice as much), sales revenue (more than three times higher) and gross margins (four times higher) (Nishikawa, Schreier & Ogawa, 2013). Furthermore, they found that co-creation has the potential to improve the effectiveness of problem solving activities by a firm (Nishikawa, Schreier & Ogawa, 2013). This suggests that is could be worthwhile for firms to, parallel to their traditional innovation, open up for innovation by customers and indicates that solutions or products that are developed by or with a firm’s customers have the potential of being more cost effective than traditional firm solutions.
Studies also found that customer-generated ideas score on average higher on novelty compared to firm-generated products, indicating that customers are an important source in the process of creating new products or services (Poetz & Schreier, 2012; Nishikawa, Schreier & Ogawa, 2013). Furthermore, research has found that customer-generated products score on average higher on customer benefit, but somewhat lower on feasibility compared to firm-generated products and that firm-firm-generated products are usually easier developed into a marketable product than customer-generated ideas, indicating that customers not always come up with ideas and solutions that are in the best interest of a firm (Poetz & Schreier, 2012; Nishikawa, Schreier & Ogawa, 2013). But in general, external participants are expected to develop better ideas in comparison to professionals from within a firm (Poetz & Schreier, 2012). The full capacity of the crowd is therefore considered to be too big to ignore in the innovation process nowadays (Boudreau & Lakhani, 2013). And in the process of integrating
customers in innovation, the actual ability of a firm to successfully integrate customers into tis process is decisive (Iansiti & Clark, 1994).
2.1.3 When and how to co-create
For firms, the dialog with their customers is important in creating shared value through co-creation. (Prahalad & Ramaswamy, 2004) This interaction is characterized by the joint creation of value, joint problem definition and solving. In order to successfully co-create, all stakeholders must be able to interact with each other to share their experiences (Prahalad & Ramaswamy, 2004; Ramaswamy & Gouillart, 2010). Prahalad and Ramaswamy (2004) describe four building blocks for this interaction as dialog, access, transparency and risk-benefits. Dialog is about having interactivity, making sure that all parties are engaged to participate and there is willingness from both sides to act. Access is concerned with making sure the knowledge and information needed is accessible for the customers. Transparency is about making sure the knowledge and information is transparent for the customers and that they have all the information necessary. And the last, Risk-benefits, states that both parties have to determine if it is worthwhile to start the dialog. (Prahalad & Ramaswamy, 2004).
Boudreau and Lakhani (2013) indicate that firms who co-create face some important challenges in this dialog. The first challenge is how a firm protects the knowledge and information it needs to share with its customers to co-create. Secondly, a firm must find out if the proposed solution by a customer is actually appropriate and applicable for the firm. And thirdly, a firm must find a way to successfully integrate the solution within the firm (Boudreau & Lakhani, 2013).
Afuah & Tucci (2012) describe important aspects to check when a firm chooses to start co-creation rather than solving the problem internally or to outsource it. The first aspect involves how the problem is defined and the accessibility of potential solvers to it. If the characteristics
of the problem are difficult to access for customers, co-creation is not the best option for the firm. The second aspect to check if the knowledge required for the solution is already available within the firm. If the knowledge is already somewhere available within the firm, co-creation is not the best option for the firm (Coviello & Joseph, 2012). The third aspect to check is the characteristics of the customers: are they able and capable to actually solve the problem. If they are and the other two aspects are in order as well, co-creation might be a good option for a firm (Afuah & Tucci, 2012). Afuah and Tucci (2010) further state that for co-creation, the pool of participants needs to be large (Afuah & Tucci, 2012). This is especially hard for starting entrepreneurs who are developing a new product or service. They usually do not have a large crowd at their disposal. Finding the right participants to co-create from the start with will most likely be somewhat of a challenge for them.
2.1.4 Who to co-create with
Firms compete for the contributions of costumers in co-creation (Chesbrough & Appleyard, 2007). Therefore, a challenge for firms who engage in co-creation is to attract the appropriate customers to co-create with and to keep their input and feedback over time. Starting entrepreneurs face this challenge as well. For them it is even harder to find the appropriate customers to co-create with, because they usually do not have any customers to start with. Customers are willing to interact with firms and be a part of the value creating process (Prahalad & Ramaswamy, 2004; Ramaswamy & Gouillart, 2010). They tend to freely reveal information with a firm when it creates value for them too (Prahalad & Ramaswamy, 2004; Ramaswamy & Gouillart, 2010). Users who freely reveal what they have done, often find that others then improve or suggest improvements to the innovation, to mutual benefit (von Hippel, 2005).
Coviello and Joseph (2012) describe five stages in new product development, each with two distinct roles for customers who participate in this development. The first stage is the ‘opportunity recognition’ stage, where a customer can participate by identifying needs or by requesting for features that are currently not available. The second stage is the ‘Customer-Based Funding’ stage, where a firm sells a concept as a development deal to customers for R&D funding or where a customer approaches a firm for an early sale, providing the firm a revenue base for the R&D activities. The third phase is the ‘Development and Testing’ phase, where a firm asks customers for technical input and technical feedback or where the customer co-develops the product. In the forth stage, called ‘(Wider) Commercialization’, a customer participates by helping the firm getting their product approved or by promoting the product to other potential customers. In the fifth stage, called ‘Feedback’, takes place throughout the whole new product development process. Here, a customer functions by being a sounding board for the firm or by proactively giving feedback on the product (Coviello & Joseph, 2012). This research focuses on the broad ‘Feedback’ stage, where a consumer is asked to help an entrepreneur in all areas by sharing his or her ideas about the product or service.
Customers that want to participate in innovation that leads to value creation, are more likely to be hobbyists than professionals in the field where they innovate, because professionals are less likely to reveal and share sensitive information (Jeppesen & Frederiksen, 2006). This view is supported by Afuah and Tucci (2012) who found that customers who are actively participating in the innovation process are more likely to solve a problem when they are not as close to the problem than participant who are close to it (Afuah & Tucci, 2012). Leading edge users, a specific type of customer that can be best described as early adaptors, are regarded as the type of customer that will most likely make the most important contribution in innovation (Jeppesen & Frederiksen, 2006). Furthermore, technical expertise, additional substantive competences and experience are all factors that positively contribute to the expected quality
of the solution by a customer (Poetz & Schreier, 2012; Nishikawa, Schreier & Ogawa, 2013). Furthermore, customers who have a ‘need’ are better able to give accurate information regarding it than those without it, indicating the involvement and commitment needed from customers in the co-creation process (Urban & von Hippel, 1988). Gruner and Homburg (2000) found that the selection of the customers to co-create with has to be made consciously and carefully by a firm. They found that lead users, financially attractive customers and close customers are attractive groups to co-create with. Lead users are customers that actually encounter a problem, financially attractive customers are customers who represent market segments and can influence the reputation of the firm in the market and close customers are customers who are actually close to the firm. On the other hand, technically attractive customers, who are innovativeness and who have specific know-how, are not (Gruner & Homburg, 2000). This research will build on these studies and will try to identify if user who actually encounter a problem as well as close customers are also important customers to co-create with for starting entrepreneurs.
2.1.5 Challenges and risks of co-creation
Satisfaction or dissatisfaction with the outcome, the perceived fairness of the relationship with the firm and sense of community within the co-creation process mediate the relationship between the customer and their co-creation experience (Gebauer, Füller & Pezzei, 2013). This indicates that the co-creation experience of the customer is a determinant for positive or negative reactions of customers who participate in innovation. If a firm thus frequently assesses and rewards these activities accordingly, customers will be more willing to engage in value co-creation behavior (Yi & Gong, 2013).
The literature describes several problems for firms that can arise when dealing with customers in co-creation. Lilien et al. (2002) state that co-creation tend to be time-consuming for firms and that co-creation requires increased efforts from firms and their employees. Also, they
discuss the danger that the end product has a low organizational fit (Lilien et al., 2002). Furthermore, the clear and complete transfer of complex, subtle and fast changing information between a customer and a firm can be a costly and time-consuming process for firms (Thomke & von Hippel, 2002). Co-creation, if not properly managed, can even further increase the uncertainty for firms. (Nambisan, 2002).
There are also risks in collaborating intensively with customers. Djelassi and Decoopman (2013) state that the most important risk for firms is that a customer could feel exploited in the innovation process. Firms must be aware of this risk and try to avoid the customer from feeling exploited or cheated. Otherwise, there is a big chance that the customers will perceive the efforts of the firm as not credible or unfair (Djelassi & Decoopman, 2013).
Customers, on the other hand, need to know that collaborating with firms comes with responsibilities and risks for them as well. Collaborating is a two-way street and customers must take some responsibility for the risks they accept, when collaborating with firms. These kinds of risks arise when a consumer neglects the professional expertise of a firm and when the customer does not focus on a partnership with the firm that is based on mutual benefit (Prahalad & Ramaswamy, 2004).
2.2 Commitment
Now that co-creation and the preconditions, implications, benefits and risks for involving customers in co-creation have been discussed, the next section will focus on elaborating on specific conditions and customer characteristics that are hypothesised to be useful for starting entrepreneurs.
2.2.1 Commitment
First, this research looks at commitment and it proposed influence on co-creative behaviour. Commitment is a psychological state that characterizes a person’s relationship with a certain
cause like a person, a firm, a product or a service and the level this commitment has implications for the person’s decision to continue or discontinue with this relationship (Mowday, Steers & Porter, 1979; Meyer, Allen & Smith, 1993; Meyer & Herscovitch, 2001). Commitment is a force that binds a person to a course of action of relevance to a target, indicating that commitment has a positive effect on a person’s willingness to help (Meyer & Herscovitch, 2001). This makes commitment an interesting factor to investigate when looking at a person’s expected co-creative behaviour.
There are three distinct types of commitment. The first, ‘Affective Commitment’, is characterized by an affective attachment between a person and a cause. A person with a high level of affective commitment is committed to an organization because he or she wants to due to certain values or moral beliefs, indicating a free choice associated with this type of commitment. Typical bases for this type of commitment are the relevance to a person’s identity, values and interest (Meyer, Allen & Smith, 1993; Meyer & Herscovitch, 2001). Customers are attached trough affective commitment with a starting entrepreneur by their shared interest in the product or idea. Therefore, this research will use affective commitment by describing a link with a product or an idea and will test if this influences the co-creative behaviour of the customer.
The second, ‘Continuance Commitment’ is characterized by a perceived cost by a person associated with leaving the cause. A person with a high level of continuance commitment is committed to an organization because he or she needs to. The bases for this type of commitment are investments made and the lack of alternatives (Meyer, Allen & Smith, 1993; Meyer & Herscovitch, 2001). Continuance commitment implies that for it to be in place, there has been a transaction between the customer and the starting entrepreneur. A customer typically does not make an investment in a starting entrepreneur and therefore it is believed that continuance commitment will not link customers to starting entrepreneurs. Continuance
commitment can become an issue after a customer has been involved with co-creating with a starting entrepreneur, but this research focuses on the first step a customer and starting entrepreneur make in the co-creating process, therefore this type of commitment will not be used in this research.
The third, ‘Normative Commitment’ is characterized by the feeling of obligation of a person to remain. A person with a high level of normative commitment is committed to an organization because he or she feels obliged to. The bases for normative commitment are socialization and psychological contract (Meyer, Allen & Smith, 1993; Meyer & Herscovitch, 2001). Customers can be connected with starting entrepreneurs via a social or psychological contract. Usually this will mean they are friends of the starting entrepreneur or somehow related to them. Close customers are attractive groups to co-create with, indicating that this group is interesting as well for a starting entrepreneur (Gruner & Homburg, 2000). Furthermore, the entrepreneur’s personal network is an important factor in the success of co-creating activities, indicating that customers close to the entrepreneur can provide good ideas (Mollick, 2014). This research will use focus on this link with the starting entrepreneur and will test by using normative commitment by describing a link with the starting entrepreneur and will test if this influences the co-creative behaviour of the customer.
2.2.2 Hypothesis development
This research will test if commitment influences a person’s co-creative behaviour and tests if close customers as described by Gruner and Homburg (2000), are also important for starting entrepreneurs. In this research, two types of commitment, affective and normative commitment, are tested. As stated earlier, continuance commitment will not be used, because it implies that a customer has made an investment that commits them to the starting entrepreneur. However, it seems unlikely that starting entrepreneurs will have many
customers that are committed to the starting entrepreneur via an investment. Therefore, this type of commitment is not tested.
Affective commitment was tested via personal involvement, indicating a person’s link to the idea. Normative commitment was tested via a psychological contract, indicating a person’s link to the entrepreneur. Because commitment has a positive effect on a person’s willingness to help (Meyer & Herscovitch, 2001), it is expected that both affective as well as normative commitment will have a positive effect on a person’s co-creative behaviour. Meyer et al. (2002) found that out of the three types of commitment, affective commitment has the strongest and most favourable correlations with organization-relevant outcomes and that normative commitment is also associated with these outcomes, but not as strongly as affective commitment (Meyer et al., 2002). Therefore, any affective commitment is expected to have the higher positive effect on co-creative behaviour, compared to normative commitment. Furthermore, it is expected that normative and affective commitment combined will have a higher positive effect on co-creative behaviour, compared to affective and normative commitment alone. This leads to the following hypotheses:
1a) The higher the level of normative commitment and affective commitment, the higher the chance of co-creative behaviour.
1b) Affective commitment has a stronger positive effect on co-creative behaviour than normative commitment.
2.3 Personality
Next, this research looks at personal characteristics and their proposed influence on co-creative behaviour.
2.3.1 Personality
A person’s personality influences its behaviour and personality is assumed to predispose overt behavior relevant to the trait or attitude under consideration (Ajzen, 2005). For example, an extrovert person is considered to be rather talkative than silent and a conscientious person is considered to be more tidy than careless. The wish to be recognized and acknowledgement by the firm are both related to a customers motivation to participate in innovation (Jeppesen & Frederiksen, 2006). Customers also find the hedonic and symbolic benefits provided to them by participating in innovation very important, in addition to utilitarian benefits. Besides monetary motives, they can be motivated to participate by needs or desires such as altruism, reputation building, skill demonstration or the desire to belong to a specific group (Djelassi & Decoopman, 2013; Afuah & Tucci, 2012). Non-monetary incentives are found to be more important than monetary incentives in this process and most firms use non-monetary incentives to motivate customers to participate in new product development, although some firms combine both monetary and non-monetary incentives to motivate customers to participate (Rohrbeck, Steinhoff & Perder, 2010). There are thus several indications in the literature that a persons characteristics influence their co-creative behaviour. However, few researches focus on the actual personal characteristics and investigate their influence on co-creative behaviour.
This research focuses on this gap by examining four specific personality dimensions and investigates if these personal characteristics have an effect on co-creative behaviour. The four dimensions, curiosity, citizenship, extraversion and generosity, are investigated in this research because they are each expected to influence a persons’ co-creative behaviour. Out of these dimensions, citizenship and extraversion give an indication of a customer’s willingness to help others. (Sherrod, Flanagan & Youniss, 2002). Curiosity is about the customers desire to search for new information (Litman, 2005; Loewenstein, 1994; Kashdan, Rose & Fincham
2004). The latter, generosity is about a customer’s willingness to share this information (Goldberg, 1990). Combined it is expected that these four personality dimensions give an overall indication of the co-creative behaviour of a customer. Each dimension and their proposed link to co-creative behaviour is separately discussed in the following sections.
2.3.2 Curiosity
Curiosity is defined as a “positive emotional-motivational system associated with the recognition, pursuit, and self-regulation of novel and challenging opportunities.” (Kashdan, Rose & Fincham 2004). It induces both inquisitive and exploratory behaviours, is described as the search for new information and energizes a persons desire to know, see and experience (Litman, 2005; Loewenstein, 1994; Kashdan, Rose & Fincham 2004). This indicates that people that are more curious than others will feel a stronger urge to help a starting entrepreneur than people that are less curious. Therefore, a higher level of curiosity is expected to have a positive effect on co-creative behaviour. This leads to the following hypothesis:
2) The higher the level of curiosity, the higher the level of co-creative behaviour. 2.3.3 Citizenship
Citizenship is about helping someone else. It consists of characteristics related to moving beyond self-interest and it involves the sense of being connected to a group other than one’s own group such as family, community, race, or religion (Sherrod, Flanagan & Youniss, 2002). Furthermore, citizenship is about having concern for others who are outside their own group (Sherrod, Flanagan & Youniss, 2002). Gebauer, Füller and Pezzei (2013) found that a persons’ sense of community within the co-creation process mediates the relationship between the customer and their co-creation experience (Gebauer, Füller & Pezzei, 2013). Furthermore, Yi and Gong (2013) describe a specific type of customer behaviour in co-creation, customer citizenship behavior, which is described as voluntary behavior that is not necessarily required
for value co-creation (Yi & Gong, 2013). Citizenship can therefore be an important dimension for starting entrepreneurs to look for in their customers when they want to co-create. Customers that tend to move more beyond their self-interest will more likely help a starting entrepreneur, than customers who are less likely to move beyond their self-interest, because they are more concerned with the entrepreneur (Sherrod, Flanagan & Youniss, 2002). A higher level of citizenship is therefore expected to have a positive effect on co-creative behaviour.
3) The higher the level of citizenship, the higher the level of co-creative behaviour. 2.3.4 Extraversion
Extraversion is “the act, state, or habit of being predominantly concerned with obtaining gratification from what is outside the self” (Costa & McCrae, 1992). Extraverts like to interact and are characterized by being enthusiastic, optimistic, talkative, assertive, and sociable. Extravert persons like to be being around other people and take pleasure in activities that involve social gatherings. They enjoy the companionship of others and are energized when they are around other people, and they are likely to enjoy to time spent with people rather than spending time alone (Costa & McCrae, 1992; Watson & Clark, 1997). Opposed to Extraverts are Introverts, who are characterized as being more reserved, quiet and shy (Costa & McCrae, 1992; Watson & Clark, 1997). People that are more introvert tent to ‘avoid’ social contact and will therefore less likely help an entrepreneur compared to extravert people. Therefore, it is expected that an extravert customer will probably help a starting entrepreneur more in co-creating than an introvert customer. This leads to the following hypothesis:
2.3.5 Generosity
Generosity reflects a person’s passion to help others and their readiness to share with others. Goldberg (1990) characterizes generosity by giving something without expecting anything in return. This can involve spending time, money or labour. It includes a person’s intentions of looking out for a common good, giving from the heart and it is not based a person’s economic status. People who are Generous are typically described as: charitable, indulgent and lenient (Goldberg, 1990). Due to these characteristics, generous customers are expected to be a useful group to co-create with for a starting entrepreneur, because customers who are more generous than others, will more likely help an entrepreneur with his or her idea, compared to customers who are less generous. This leads to the following hypothesis:
2.4 Conceptual Model + + + + + Curiosity (H2) Citizenship (H3) Extraversion (H4) + Normative Commitment Affective Commitment Co-creative Behaviour Generosity (H5) +
3 Method
3.1 ProcedureThe data for this research was gathered via an online survey. This method was chosen over other research methods, because with the aim of getting as much respondents possible and with limited resources and time, an online survey is considered to be the best option (Saunders, Lewis & Thornhill, 2012). Respondents were contacted via email and asked to participate. In order to answer the research question and hypothesis, respondents were asked questions on their personal characteristics and co-creative behaviour, after controlling for commitment. In order to control the type of commitment, the respondents were presented with for the four different cases. The respondents were asked to consider these different cases and subsequently asked them to answer questions about their expected co-creative behaviour on the given case. The focus on these questions was to gather information about what the respondent would do in these different cases.
3.2 Data collection
The data for this research was collected online via Qualtrics (http://uvafeb.qualtrics.com).
3.2.1 Population and Sample
For this research, the convenient sampling strategy with a snowball effect was chosen. This method was chosen because this method is fast, easy and cost effective. (Saunders, Lewis & Thornhill, 2012). Respondents were contacted via the researchers private and business network and via fellow students of the Amsterdam Business School. Each participant was asked to forward the questionnaire to other possible participants. The respondents are therefore connected to the researcher in some way or via someone who knows the researcher, making it very hard to have a representative sample for a larger population (Saunders, Lewis & Thornhill, 2012).
The original sample size of this research was N=126. After checking for missing values, the sample was reduced to N=84. Unfortunately, none of the 42 respondents that were dropped out of the sample were partially useful for this research. Out of these 42, 23 had incomplete personality data and the rest did not answer enough questions to record a score per case. From the 84 respondents that were left over, two appeared to be faked by the respondent, because of repetitive equal answers and a very short completion time, compared to the other respondents. These two respondents were also deleted from the sample, resulting in a final sample of N=82 (65%). This dropout rate is possibly explained by the fairy large amount of questions in the questionnaire. The average completion time was around 15 minutes, which is quite long for an online questionnaire of this type (Saunders, Lewis & Thornhill, 2012).
Of the 82 respondents in the final sample that was used, 51 are female (62,2%) and 31 are male (37,8%). The average age is 41 with a standard deviation of 14. 71 respondents (86,6%) have first or second-degree personal ties to the researcher and 11 respondents (13,4%) are co-students of the researcher.
As for education, 48 respondents (58,5%) have a masters degree or higher. 27 (32,9%) have a Bachelor’s degree, three (3,7%) a college degree and four (4,9%) a high-school diploma. 54 respondents (65,9%) are employed, 16 are self-employed (19,5%), six (7,3%) are retired, three (3,7%) are currently looking for a job, two (2,4%) are not looking for a job and one (1,2%) did not answer.
3.3 Measures
The research was done in Dutch, therefore all questions were translated from English into Dutch. In order to assure that the content of the items remained the same in Dutch, a person not involved with the research translated these items on her own. After combining and inspecting these translations, no problems were found and therefore no extra help was needed
in order to properly translate the questions. The items used in the questionnaire were derived from different studies. Each item will be discussed in the next sections.
3.3.1 Co-creative behaviour
Co-creative behaviour was measured based on two scales developed by Handrich and Heidenreich (2013) and by Yi and Gong (2013). Yi and Gong (2013) developed a scale to measure customer value co-creation behavior and treat customer value co-creation behavior as a multidimensional concept consisting of two high-order factors called customer participation behavior and customer citizenship behavior. These factors are each made up of four dimensions. The four dimensions of customer participation behavior are: information seeking (to know what to do), information sharing (to actually share the information with the firm), responsible behavior (being cooperative, accepting directions from the firm and observing their rules and policies), and personal interaction (the interpersonal relations between customers and employees). Those of customer citizenship behavior are: feedback (both solicited and unsolicited information that customers provide to the firm), advocacy (recommending the firm to others), helping (the willingness to help other customers), and tolerance (the willingness to be patient about the development of the product or service) (Yi & Gong, 2013). Because the measurement developed by Yi & Gong is applicable to asses an actual co-creative process and therefore some of their dimension are not applicable to the cases described in the questionnaire, a selection of their measurements was made. Out of these dimensions, this research measures information sharing and feedback, because they are best linked to an intention to co-create, compared to the other measurements.
Another method to measure co-creative behaviour, developed by Handrich and Heidenreich (2013) was used. They developed a scale measure the willingness of a customer to engage in co-creation. Their scale distinguishes three dimensions: customization, effort and information Sharing. Again, some of these dimension are not applicable to the cases described in the
questionnaire and therefore a selection of these measurements was made. Out of these dimensions, effort was uses in this research, because it is best linked to an intention to co-create, compared to the other measurements. Each question was slightly adjusted for the subject of this research (Handrich and Heidenreich, 2013).
Combined, these three dimensions, feedback, information sharing and effort, are expected to measure a person’s intended creative behaviour. In order to determine the respondent’s co-creative behaviour, these questions were asked in the survey. Respondents were asked to indicate their agreement or disagreement with each of the items, using a five point Likert-scale ranging from one ‘strongly disagree’ to five ‘strongly agree’, with a midpoint labelled ‘neither agree nor disagree.’ The questions can be found in appendix II.
In the present study, feedback was measured for each of the four cases. The obtained Cronbach’s alpha reliability estimate was in case one .82, in case two .70, in case three .81 and .83 in case four, indicating a strong and reliable scale was used to measure feedback (Field, 2009).
Furthermore, information sharing was measured for each of the four cases. The obtained Cronbach’s alpha reliability estimate was in case one .59, in case two .70, in case three .70 and .78 in case four, indicating an issue with the reliability of the scale (Field, 2009). After assessment of the individual questions, the first question regarding information sharing: “I would clearly explain what I wanted the entrepreneur to do.” was eliminated from the results, significantly changing the reliability for the first two cases and slightly changing it for the last two cases (Field, 2009). This question focuses on explaining by the respondent, whereas the three other questions focus more on actually giving the information and answering questions, which could explain the difference in reliability. This adjustment altered the Cronbach’s alpha
reliability estimate to .74 in case one, .82 in case two, .74 in case three and .76 in case four, indicating a strong and reliable scale to measure information sharing.
Also, effort was measured for each of the four cases. The obtained Cronbach’s alpha reliability estimate was .85 in case one, .86 in case two, .86 in case three and .85 in case four, indicating a very strong and reliable scale to measure effort (Field, 2009).
3.3.2 Commitment
According to Meyer, Allen and Smith (1993), “Commitment can take different forms, and it is therefore imperative that researchers state clearly what form or forms of commitment they are interested in and that they ensure that the measures they use are appropriate for the intended purpose” (Meyer, Allen & Smith, 1993). As stated earlier, this research focuses on affective as well as normative commitment and the items used in the questionnaire were designed using their descriptions.
To measure commitment, several descriptive items to describe affective and normative organizational commitment were used. (den Hartog, de Hoogh & Belschak, 2014). These items can be found in appendix II and were used to describe the cases that were presented to the respondents. In the cases, two different levels of organizational commitment were described: full commitment and no commitment, making a total of four cases presented to the respondents. The actual texts that were used in the questionnaire can be found in appendix III. Table 1: The commitment types that were described per case in the questionnaire.
Case Type of commitment used
1 High normative and affective commitment 2 High normative and low affective commitment 3 Low normative and high affective commitment 4 Low normative and affective commitment
3.3.3 Personality
To assess the influence of personality on co-creative behaviour, four personality dimensions were used as a measurement scale. As described earlier, curiosity, citizenship, extraversion
and generosity were measured. These items were put on a five point Likert-scale ranging from strongly disagree, disagree, neither agree nor disagree, agree to strongly agree where higher scores indicate higher levels of that specific dimension. Half of the 39 statements were reversed and the questions were randomly put in order in de questionnaire. The actual statements that were used in the questionnaire can be found in appendix II.
The means of the four personality dimensions extraversion ‘PEXMean’, curiosity ‘PCUMean’, generosity ‘PGEMean’ and citizenship ‘PCIMean’ were used in this research, combined with the means for the three co-creative dimensions for each of the four cases, making 12 means in total for these dimensions. These 12 items were coded as ‘cxXXMean’, where cx stands for the case number (c1,c2,c3,c4) and the XX was used for the three dimensions: feedback (FB), information sharing (IS) and effort (EF). Furthermore, the control variables gender, age, education and profession were also used to analyse the data.
3.3.3.1 Curiosity
The personality dimension ‘Curiosity’ was assessed using the ten-item scale developed by Peterson and Seligman (2004). The nine items with an original Cronbach’s alpha of .78 were derived from the International Personality Item Pool (IPIP, 2014). Each items represents a statement related to curiosity, with items some items as reversed statements.
The average score for this dimension per respondent was obtained by calculating the mean of all the statements combined, after changing the score of the reversed items. In the present study the obtained Cronbach’s alpha reliability estimate was: .72, indicating a strong and reliable scale to measure curiosity (Field, 2009).
3.3.3.2 Citizenship
The personality dimension ‘Citizenship’ was assessed using the nine-item scale developed by Peterson and Seligman (2004). The ten items with an original Cronbach’s alpha of .78 were
derived from the International Personality Item Pool (IPIP, 2014). Each items represents a statement related to citizenship, with some items as reversed statements.
The average score per respondent for this dimension was obtained by calculating the mean of all the statements combined, after changing the score of the reversed items. After inspection, the calculated mean value was found to be moderate negatively skewed. Therefore, the mean values were transformed (Xt = √(K-X)) and recoded back in order to undo the reversing of the variable, caused by transformation of a negative skew (Field, 2009). In the present study the obtained Cronbach’s alpha reliability estimate was: .67, indicating a moderately strong and reliable scale to measure citizenship (Field, 2009).
3.3.3.3 Extraversion
The personality dimension ‘Extraversion’ was assessed using the ten-item scale from the Big-Five domains (Goldberg, 1990). The ten items with an original Cronbach’s alpha of .87 were derived from the International Personality Item Pool (IPIP, 2014). Each items represents a statement related to extraversion, with some items as reversed statements.
The total score for this dimension was obtained by calculating the mean of all the statements combined, after changing the score of the reversed items. In the present study the obtained Cronbach’s alpha reliability estimate was: .90, indicating a very strong and reliable scale to measure extraversion (Field, 2009).
3.3.3.4 Generosity
The personality dimension ‘Generosity’ was assessed using the ten-item scale developed by Peterson and Seligman (2004). The ten items with an original Cronbach’s alpha of .72 were derived from the International Personality Item Pool (IPIP, 2014). Each items represents a statement related to generosity, with some items as reversed statements.
The total score for this dimension was obtained by calculating the mean of all the statements combined, after changing the score of the reversed items. In the present study the obtained Cronbach’s alpha reliability estimate was: .60, indicating some issues with the strength and reliability of this scale (Field, 2009).
3.3.4 Control variables
Lastly, four demographics were asked from the respondent at the start of the survey. Each respondent was asked to indicate his or her gender, age, education and current profession. The actual questions can be found in appendix II.
4. Results
In the following sections the results of this research are described. The results are discussed in order, giving an overview of each step that was taken in order to test the hypotheses.
4.1 Correlation
In order to analyse the data, firstly a correlation matrix of this research was made. This matrix is summarized in Table 2.
! !
Table 2: Means, Standard Deviations and Correlations. Variables M SD 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 01. Gender 1.62 0.49 - 02. Age 41.41 14.02 .10 - 03. Education 5.45 0.79 .06 .35** - - 04. Profession 1.78 1.50 .00 .36** -.21 - 05. Citizenship 1.66 0.16 -.01 .30** - .20 -.14 (.67) 06. Extraversion 3.46 0.63 -.16 .34** - .08 -.19 .56** (.90) 07. Generosity 3.66 0.42 -.03 -.17 -.09 -.21 .48** .41** (.60) 08. Curiosity 3.88 0.45 -.15 -.01 -.02 -.16 .31** .43** .35** (.72) 09. C1 Feedback 4.34 0.43 .03 -.13 .13 -.18 .20 .13 .10 .34** (.83) 10. C1 Effort 0.67 0.12 -.07 -.20 .12 .03 .21 .02 .10 .29** .42** (.86) 11. C1 Information Sharing 4.22 0.48 .10 -.02 .13 -.08 .16 .03 .05 .34** .80** .55** (.74) 12. C2 Feedback 4.19 0.51 .13 -.17 .11 -.16 .24* .13 .19 .32** .74** .31** .65** (.70) 13. C2 Effort 1.71 .22 -.03 -.19 .12 -.09 .18 .00 .12 .17 .21 .61** .28* .26* (.86) 14. C2 Information Sharing 3.97 0.60 .08 -.10 .17 -.21 .36** .19 .21 .32** .55** .42** .58** .70** .53** (.82) 15. C3 Feedback 1.60 0.21 .08 -.00 .02 .05 .30** .19 .24* .27* .58** .20 .51** .62** .03 .51** (.81) 16. C3 Effort 2.76 0.72 -.09 .16 .03 .12 -.02 -.10 -.01 .09 .02 .33** .20 .11 .23* .11 .34** (.86) 17. C3 Information Sharing 3.73 0.63 .10 .12 -.04 -.03 .26* .18 .18 .28* .38** .22* .46** .46** .03 .51** .75** .42** (.74) 18. C4 Feedback 0.61 0.15 -.05 -.03 -.04 .03 .22 .13 .28 .35** .32** .16 .30** .53** .16 .44** .64** .26* .49** (.83) 19. C4 Effort 1.32 0.22 -.10 .22* -.11 .17 -.07 -.12 -.01 -.06 .46** - -.05 .30** - -.26* .29** -.03 -.01 .48** .13 .15 (.85) 20. C4 Information Sharing 2.99 0.81 .06 .13 -.05 -.09 .15 .05 .17 .30** .20 .06 .21 .33* .20 .48** .53** .26* .63** .59** .36** (.76)
**. Correlation is significant at the 0.01 level (2-tailed) *. Correlation is significant at the 0.05 level (2-tailed)
!
Table 2 shows several significant correlations. Age is negatively correlated with Education, Citizenship and Extraversion, indicating that in this sample, elder people are found to be less educated, they show less citizenship and are more introvert than younger people (p < 0.05). Furthermore, Age is positively correlated with effort in the fourth case, indicating that in this sample, elder people tend to put more effort in that case, compared to younger people (p < 0.01).
Significant correlations were found between all the four personality dimensions. All the dimensions are positively correlated among each other (p < 0.05). Indicating that, in this sample, the four dimensions chosen for this research reinforce each other.
When looking at each personality dimension compared to the co-creative dimensions, citizenship was found to be positively correlated to feedback (p < 0.05) and information sharing (p < 0.05) in the second case as well as in the third case, indicating that in this sample, people that show more citizenship tend to give more feedback and share more information, compared to people with less citizenship.
Generosity was found to be positively correlated to feedback in the third case (p < 0.05), indicating that, in this sample, people who are more generous, tend to give more feedback when they do not know the starting entrepreneur, but are interested in the idea of the entrepreneur, compared to people that are less generous.
Interestingly, positive significant correlations were found between curiosity and feedback as well as information sharing in all the four cases (in the first, second and fourth case p < 0.01 and in the third case p < 0.05), indicating that, in this sample, people who are more curious, tend to give more feedback and share more information, regardless of their commitment to the starting entrepreneur or the idea of the entrepreneur, compared to people that are less curious.
!
Significant correlations were also found between the three dimensions measuring co-creative behaviour. Per case, feedback, information sharing and effort are all correlated in this sample (p < 0.05) indicating that they are all positively connected to each other. Although the sample used for this research is small, this result is very encouraging because the three dimensions measuring ‘Co-creative Behaviour’ were brought together by the researcher after intensively looking for constructs to measure ‘Co-creative Behaviour’, because no existing set of questions was found in the literature to measure this specific behaviour. This correlation at least shows an indication that the chosen dimensions are related to each other. It however does not indicate that they actually do measure ‘Co-creative Behaviour’ as intended by the researcher. In order to test this assumption, a factor analysis was done. The results of this test can be found in the next section of this research.
Also, some significant positive correlations were found between the three dimensions measuring co-creative behaviour per case and one of these dimensions in another case (p < 0.05). The implications of these will be tested in the following sections of this research.
4.2 Factor Analysis
Next a factor analysis was done in order to check if the items for each case can be reduced to one single item. A principal component analysis (PCA) was conducted on the 11 items for each of the four cases. This analysis was done to test if there are similarities between the variance of these variables in order to reduce these 11 items to one that measures co-creative behaviour (Field, 2009).
In order to be successful, the result of the factor analysis would be to find only one component that explains the variance of the underlying items (Field, 2009). The PCA analysis was done using the parallel analysis developed by Hayton, Allen and Scarpello (2004). For this analysis, the eigenvalues for the 11 items were compared to eigenvalues that were
!
calculated from 50 separate random sets of data, each containing 11 items and 82 cases. Out of these eigenvalues per item, the mean and 95th percentile of the underlying eigenvalues was generated and compared to the measures eigenvalue of the actual item. According to Hayton, Allen and Scarpello (2004), each component with an actual eigenvalue score that is lower than the corresponding random score can be left out as individual component (Hayton, Allen & Scarpello 2004).
Table 3: Average eigenvalues and actual eigenvalues per case. Mean 95th Percentile Actual
Case 1 Actual Case 2 Actual Case 3 Actual Case 4 1,687 1,844 5,268* 4,983* 4,897* 4,357* 1,487 1,619 1,989* 2,030* 2,279* 2,609* 1,340 1,448 0,911 0,975 0,862 0,792 1,218 1,289 0,654 0,707 0,571 0,722 1,110 1,192 0,520 0,576 0,535 0,514 1,008 1,061 0,445 0,499 0,507 0,465 0,926 0,978 0,384 0,399 0,391 0,394 0,830 0,887 0,300 0,299 0,318 0,382 0,737 0,814 0,205 0,251 0,279 0,302 0,637 0,697 0,174 0,177 0,206 0,259 0,558 0,627 0,150 0,105 0,155 0,202
* = Actual eigenvalue is higher than the random eigenvalue
In the following section, the results of the PCA per case are discussed, followed by an overall conclusion for the PCA for this sample.
4.2.1 Case 1
The PCA for the items of the first case was conducted with orthogonal rotation (varimax). The Kaiser-Meyer-Olkin (KMO) measure verified the sampling adequacy for the analysis, KMO = .84, which is ‘great’ according to Field (2009). All the individual KMO values for the individual items were > .70, which is above the acceptable limit of .5 (Field, 2009). Bartelett’s test of sphericity resulted in an Chi-Square (55) = 537.373, p < .000, indicating that correlations between the items were sufficiently large for the PCA (Field, 2009). The
!
PCA analysis was done using the parallel analysis developed by Hayton, Allen and Scarpello (2004). Examination of the results for case one in table 3 indicates that the first two actual eigenvalues are greater than those generated by the random PCA. In combination they explain 65.97% of the variance. The results of the exploratory factor analysis results for case one can be found appendix V. The results indicate that information sharing and feedback can be put together as one single component, but effort cannot.
4.2.2 Cases 2, 3 and 4
The PCA for the items of the other cases were also conducted with orthogonal rotation (varimax). The Kaiser-Meyer-Olkin (KMO) measure verified the sampling adequacy for the analysis for all the cases was above .84 , which is ‘great’ according to Field (2009). All the individual KMO values for the individual items were > .74, which is above the acceptable limit of .5 (Field, 2009). Bartelett’s test of sphericity resulted in an Chi-Square (55) = 520.508,
p < .000, in case two, 490.291, p < .000 in case three and 428.236, p < .000 in case four,
indicating that correlations between the items were sufficiently large for the PCA. The PCA analysis was done using the parallel analysis developed by Hayton, Allen and Scarpello (2004). In combination thy all explain a minimum of 63.33% of the variance. The results of the exploratory factor analysis results for case two, three and four can be found appendix V. The results indicate, as in case one, that information sharing and feedback can be put together as one single component, but effort cannot.
4.2.3 Combined results
After inspection the Factor analyses of the four cases it can be concluded that, for this sample, it is not possible to combine the 11 items for feedback, effort and information sharing into one item, measuring ‘Co-Creative Behaviour’. The PCA shows that for this sample, the 11 items can be brought back to two components: one measuring feedback and information sharing
!
together and the other measuring effort. Instead of testing the hypotheses with a single item, they will be tested using the separate scores for feedback, effort and information sharing per case.
In order to do test the results of the three factors per case, the mean value per factor per case was calculated. After inspection for skewness and kurtosis for these mean values, several data transformations were necessary in order to solve skewness and/or kurtosis. The mean values for effort in case two and four were found to be moderate positively skewed. Those values were transformed (Xt = √X) and stored in the dataset (Field, 2009). The mean value for feedback in case three was found to be moderate negatively skewed. This value was transformed (Xt = √(K-X)) and recoded in order to undo the reversing of the variable, caused by transformation of a negative skew (Field, 2009). The mean value for feedback in case four as well as the mean value for effort in case one was found to be substantial negatively skewed. These values were transformed (Xt = Log10(K-X)) and recoded in order to undo the reversing of the variable, caused by transformation of a negative skew (Field, 2009).
With all the calculated means for both the personality as well as the co-creative items in order, the hypotheses were tested.
4.3 Paired-Samples T-test
In order to test the effect of commitment on the co-creative behaviour of the respondents in the four different cases, a Paired-Samples T-test was done. The samples are dependent because they are taken from the same participants, which all took part in the four experimental cases of this research (Field, 2009). The Paired-samples t-test was used to test whether the group means of each case was different than the case before, comparing case one with two, case two with three and case three with four (Field, 2009). In the following section, the results of these tests are described.
!
On average, the participants scored higher in the first case, which described both high normative as well as affective commitment, on all three dimensions compared to the second case, which described high normative commitment and low affective commitment. The participants tend to give less feedback in the second case (M = 4.19, SE = 0.06) compared to the first case (M = 4.34, SE = 0.05), t(81) = 3.91, p = .00, r = .40. They also tend to share less information in the second case (M = 3.97, SE = 0.07) compared to the first case (M = 4.22, SE = 0.05), t(81) = 4.45, p = .00, r = .44. Furthermore, the participants would give less effort in the second case (M = 2.67, SE = 0.08) compared to the first case (M = 3.80, SE = 0.07), t(81) = 11.61, p = .00, r = .79. In short, when the first and second case were compared, significant effects were found for all the three dimensions. In order to assess if these differences are substantial, the effect size of the difference (r-value) was calculated (Field, 2009). In this sample, only the effect size for effort is substantive, given the threshold of .5 (Field, 2009).
On average, the participants scored higher in the second case, which described high normative commitment and low affective commitment, on all three dimensions compared to the third case, which described low normative commitment and high affective commitment. The participants tend to give less feedback in the third case (M = 3.99, SE = 0.07) compared to the second case (M = 4.19, SE = 0.06), t(81) = 3.69, p = .00, r = .38. They also tend to share less information in the third case (M = 3.73, SE = 0.07) compared to the second case (M = 3.97,
SE = 0.07), t(81) = 3.59, p = .00, r = .37. Furthermore, the participants would give less effort
in the third case (M = 2.76, SE = 0.08) compared to the second case (M =2.97, SE = 0.07),
t(81) = 2.00, p = .05, r = .22. In short, when the second and third case were compared,
significant effects were found for all the three dimensions. In order to assess if these differences are substantial, the effect size of the difference (r-value) was calculated (Field, 2009). Interestingly, in this sample, although all the comparisons were significant, no
