1
Siegrid Reitsma
2
Siegrid Reitsma
University of Groningen
Faculty of Economics & Business
Department of Marketing
Msc Business Administration: Marketing Management
Master Thesis
Completion date August 2011
Author Siegrid Reitsma
Student number 1799290
Address Jacobijnerkerkhof 12 8911 EN Leeuwarden Phone number 0610468096
E-mail address siegridreitsma@hotmail.com Rijksuniversiteit Groningen
First supervisor dr. W. Jager Second supervisor dr. S. Gensler
3
ABSTRACT
This study answers the question what the influence of consumer innovativeness is on social influence, when mediated by product involvement and the number of connections. Also, the influence of homophily on the strength of normative social influence relative to informative social influence is tested. To answer these questions a literature study has been performed to see what is already known about each concept included in this study. The concepts discussed in the theoretical framework are consumer innovativeness, product involvement, number of connections, homophily, normative and informative social influence and ingoing and outgoing social influence. The literature that comes close to this study is the literature that touches upon the relationship between product leadership and social influence. The literature in this field was helpful to establish the expectations for the direct effect of the independent variable on the mediators and dependent variable, and the direct effects of the mediators on the dependent variable. None of the studies we found included a mediating effect in the form of product involvement and number of connections in their models. Therefore, this study attributes by adding the mediators product involvement and number of connections to the relationship of consumer innovativeness on social influence. To test our hypotheses an online survey was conducted. The information obtained from 212 respondents was found useful to test if there is a relationship between homophily and the strength of normative social influence relative to informative social influence. The information obtained from 154 respondents was used to test the relationship between consumer innovativeness and social influence. A simple linear regression was used to investigate the relationship between homophily and relative normative social influence. The more advanced mediation analysis called ‘bootstrapping test’ was used to investigate the relationship between consumer innovativeness and social influence mediated by product involvement and number of connections. The results show that taken as a set the variables do not mediate the relationship between consumer innovativeness and outgoing social influence and ingoing social influence. However, taken separately product involvement and number of connections do seem to mediate the relationship between consumer innovativeness and outgoing and ingoing social influence. No significant results were found for mediation of product involvement and number of connections on the relationship between consumer innovativeness and the relative normative social influence. However, a significant direct effect was found of consumer innovativeness on relative normative outgoing social influence. No significant results were found for there to be a relationship between homophily and relative normative social influence. For marketers these results suggest that when launching a new product it pays to get this new product under the attention of innovators. As innovators are in contact with a large number of potential customers and they are more product involvement than others they are very strong on exerting social influence on others. If the new product meets the expectations of the innovators these innovators should be seen as a low-cost promotion tool for new products.
Key words: Consumer Innovativeness, Product Involvement, Number of Connections, Homophily,
Normative Social Influence, Informative Social Influence, Ingoing Social Influence
4
PREFACE
This thesis is the final step towards completing the Master of Science in Business Administration. When starting this study my goal was to specialize in the field of marketing. After having finished my bachelor-degree in International Business and Languages I felt that there was more for me to learn about marketing and if I was to become a marketer I needed to delve into marketing on the academic level. Now, 2 years later, I learned more than I had hoped for and I feel that the study gave to me new insights that will be of help in my future career. Also, concerning my personal development I feel that this study on the academic level has taught me to think in a more advanced manner.
Special thanks go out to my thesis supervisors dr. Sonja Gensler and dr. Wander Jager. Without their constructive criticism and directions I could not have completed this job. Their quick responses on whatever question I had were appreciated very much.
Also, I would like to thank my friends and family for their support during the process of making this master thesis and the study as a whole. Special thanks go out to my parents and my fellow-student and true friend, Karin Grijpma. My parents have always trusted in my ability to successfully finish this study, even when I (sometimes) doubted this. Their comforting and inspiring words have helped me tremendously. Furthermore, Karin Grijpma and I have formed a great team during these last two years. By helping and motivating each other we have brought out the best in each other.
With this thesis I hope to have highlighted an interesting topic. I hope the findings will be helpful in practice and that it will serve as a basis for future research.
Leeuwarden,
5
TABLE OF CONTENTS
1. INTRODUCTION ... 7
2. THEORETICAL FRAMEWORK ... 10
2.1 THE INFLUENCE OF INNOVATIVENESS ... 10
2.2 THE INFLUENCE OF INVOLVEMENT ... 11
2.3 THE INFLUENCE OF HOMOPHILY ... 11
2.3.1VALUE- AND STATUS HOMOPHILY ... 12
2.3.2INDUCED HOMOPHILY ... 12
2.3.3HOMOPHILY AND SOCIAL INFLUENCE ... 12
2.3.4REFERENCE GROUPS ... 12
2.4 THE INFLUENCE OF A PERSON’S NUMBER OF CONNECTIONS ... 13
2.5 SOCIAL INFLUENCE ... 14
2.5.1INFORMATIVE SOCIAL INFLUENCE VERSUS NORMATIVE SOCIAL INFLUENCE ... 14
2.5.2PROCESS-RELATED DIFFERENCES ... 15
2.6 SOCIAL INFLUENCE: INGOING VERSUS OUTGOING SOCIAL INFLUENCE ... 20
2.7 CONCEPTUAL FRAMEWORK AND HYPOTHESES ... 20
2.7.1CONSUMER INNOVATIVENESS ON INGOING AND OUTGOING SOCIAL INFLUENCE ... 21
2.7.2CONSUMER INNOVATIVENESS ON THE STRENGTH OF NORMATIVE SOCIAL INFLUENCE RELATIVE TO INFORMATIVE SOCIAL INFLUENCE ... 23
2.7.3DEGREE OF HOMOPHILY ON THE STRENGTH OF NORMATIVE SOCIAL INFLUENCE RELATIVE TO INFORMATIVE SOCIAL INFLUENCE... 24
3. RESEARCH DESIGN ... 25
3.1 EXPERIMENTAL DESIGN AND SUBJECTS ... 25
3.1.1PRODUCT CHOICE ... 25 3.1.2SUBJECTS ... 25 3.1.3QUESTIONNAIRE ... 26 3.1.4MEASURING INSTRUMENTS ... 26 4. RESULTS ... 30 4.1 SAMPLE ... 30 4.2 MEDIATION ANALYSIS ... 30 4.2.1SOBEL Z-TEST ... 31 4.2.2BOOTSTRAP TEST ... 32
4.3 THE INFLUENCE OF CONSUMER INNOVATIVENESS ON OUTGOING SOCIAL INFLUENCE ... 33
4.3.1DIRECT EFFECTS ... 35
4.3.2MEDIATION ... 36
4.3.3VALIDITY CHECK ... 37
4.4 THE INFLUENCE OF CONSUMER INNOVATIVENESS ON INGOING SOCIAL INFLUENCE ... 38
6
4.6 THE INFLUENCE OF HOMOPHILY ON THE STRENGTH OF NORMATIVE SOCIAL INFLUENCE RELATIVE TO
INFORMATIVE SOCIAL INFLUENCE ... 42
4.6.1MOST IMPORTANT INFLUENCERS... 42
4.6.2DIRECT EFFECT OF HOMOPHILY ON THE RELATIVE STRENGTH OF NORMATIVE SOCIAL INFLUENCE ... 43
5. DISCUSSION ... 44
6. CONCLUSION ... 49
7. LIMITATIONS AND FURTHER RESEARCH ... 50
8. REFERENCES ... 51
7
1. INTRODUCTION
In July 2011 one of the biggest and most leading companies in the field of search engine marketing, Google, decided on launches the social network site ‘Google Plus’. The access to the Google Plus network was restricted to those that were invited by Google only. Others could only join the social network if they had received an invite from a person that was already on Google Plus. The persons being the first to join Google Plus turned out to be mostly innovators in the field of social media, most of them were software and/or website engineers. These persons had a lot of on others social network sites like Facebook and Twitter. By giving these innovators access to Google Plus, Google tried to create a buzz and in the days following the introduction of Google Plus most of these innovators were indeed asked by their followers to send them an invite. Moreover, Google Plus was discussed among online and offline contacts explosively. With this introduction of Google Plus, Google seemed to want to take advantage of the social influence skills of innovators in social media. This underlying process of consumer innovativeness on social influence where Google seemed to have focused on will be discussed in this study.
As today’s society is highly connected with all of its high-tech communication possibilities, there is no escaping to social influence. Communication lines are short and physical interaction is not necessary anymore. We do not have to take a ride to a mortgage bank to get financial advice, or go to the doctor or pharmacy to get diagnosed or get advice on medicine use, we do not have to go to the store to get updated with the latest trends in clothing, decorating or gadgets. Just one simple click on our mobile phone –at any time, where ever we are- can connect us with thousands of people that are happy to give us advice on what kind of issue we have. As many researchers (Burnkrant and Cousineau, 1975; Bearden, Netemeyer and Teel, 1989) stated before, one of the most important determinants of an individual’s behavior is the influence of those around him.
Especially, the rise of the social media on the World Wide Web has made society more susceptible to social influence. People can relatively easy share their opinion and potentially influence a great number of people. Nowadays, there are online social networks focused on our professional lives, friends, and brand preferences and there are blogs and forums on medical-, technical- and beauty advice and so on. Just think of anything and it can be found on the internet. This high level of connectivity due to advanced communication channels and tools can be a positive thing when we for example need advice on a purchase decision our friends cannot help us with. But it can also have negative consequences from a company’s perspective, when someone places a bad review on brand X or product Y on a review site or a negative comment on one of the many social network sites.
8
now belong to. Both informative and normative social influence are found to affect consumer purchasing and use processes (Bearden and Etzel, 1982; Burnkrant and Cousineau, 1975; Park and Lessig, 1977). In this study, the social influence a person exerts and experiences is believed to be affected by a person’s level of consumer innovativeness, with the help of the mediators product involvement and the number of connections a person has.
Consumer innovativeness is the tendency to buy new products more often and more quickly than other people (Midgley and Dowling, 1978). This construct has been well-researched over the years (Daneels and Kleinsmith, 2001; Goldsmith and Hofacker, 1991; Hurley and Hult, 1998; Lyons and Henderson, 2005). A person with a high level of consumer innovativeness will be one of the firsts to buy a just-launched product, while a person with a low level of consumer innovativeness will wait for the product to have proven itself before he goes to buy it. In this study, consumer innovativeness is believed to affect product involvement and the number of connections a person has.
Product involvement is a person’s relevance of the object based on inherent needs, values and interests (Zaichkowsky, 1985, 1994). The more product involved a person is the more knowledge he has of the product. Involvement leads one to search for more information and spend more time searching for the right selection (Clarke and Belk, 1978).In order to be innovative one should be up-to-date with the developments in the market, this enhanced information absorption about market developments might lead to enhanced product knowledge which eventually will lead to higher levels of product involvement. Searching for information seems to lead to more opportunities to be socially influenced. Moreover, with all this product knowledge this involved person might become a product expert and therefore become a helpful information source for others. Therefore, the researcher expects there to be a link between consumer innovativeness, product involvement and social influence.
The other (expected) mediator in this study is the number of connections one has. With number of connections the researcher means the number of people a person is in contact with. Enhanced consumer innovativeness is believed to lead to being in contact with more people, because an innovative person is believed to manifest itself as a sort of expert in a certain product (category) and because an innovator searches for information on a specific product (category) more active.
With this study we hope to create a more clear picture of the relationship between consumer innovativeness, number of connections, product involvement and social influence. The current study investigates in more detail how a person’s level of innovativeness affects the relative strength of the normative and informative influence this person exerts or experiences through the mediators product involvement and number of connections. The research question of this study will be the following:
How does a person’s innovativeness influence the number of connections he has in a social network and his level of product involvement, and how do number of connections and product involvement
affect the social influence he exerts and experiences.
9
An answer on this research question could be useful for businesses that are launching a new product for example. By knowing how to target those persons that are in contact with an even larger number of potential customers, with what informative or normative cues to persuade them and with what informative and normative cues these influencers are able to persuade others, you can make a good start with making the new product successful.
As our cell phone does not only play an important role in the accessibility to social influence, whether you are the receiving or the providing party, but it alone is the subject of many discussions on birthdays, lunch breaks, product review sites, discussion forums and social media, we will use this market to test our hypotheses. Not only are there many variations in type of device, there are also many reasons why people buy a certain mobile phone and many differences in the use of the mobile phone.
10
2. THEORETICAL FRAMEWORK
2.1 The influence of innovativeness
There is a variety of terms in which innovativeness is defined. Product innovativeness is the degree of newness of a product (Daneels and Kleinsmith, 2001). Firm innovativeness indicates a firm’s ability to develop and launch new products at a fast rate (Hurley and Hult, 1998). Consumer innovativeness is the tendency to buy new products more often and more quickly than other people (Midgley and Dowling, 1978). In this paper, the word ‘innovativeness’ will refer to consumer innovativeness only. Consumer innovativeness can be measured in different ways. The tree most conventional methods are: measuring the time-of-adoption, the cross-sectional method and the self-report method (Kohn and Jacoby, 1973). The first strategy measures the time between introduction and adoption of the product, the second looks at how many of a prespecified list of new products a particular individual has bought at the time of the survey (Midgley and Dowling, 1978), and the last strategy asks consumers to define their own innovative behavior. However, these strategies were subject to faulty memory of respondents, misconceptions of past events, interviewer bias and were not domain and- or product specific or turned out to not be valid. Therefore Goldsmith and Hofacker (1991) developed a new scale that treats the before-mentioned scales as complementary methods and that is valid and reliable, their scale measures domain-specific innovativeness which is the tendency to learn about and adopt innovations (new products) within a specific domain of interest. We will elaborate on this scale later in this paper.
Whether a person is an innovator or not is market- and personality dependent. Being an innovator in a certain product category does not guarantee being an innovator in other product categories; one may be an innovator in the field of telecommunication devices but be a follower in the field of television devices (Lyons and Henderson, 2005).
Earlier studies showed that consumer innovators tend to have higher levels of income and education, are younger, have greater social mobility and favorable attitudes towards risks, and have greater social participation and higher opinion leadership (Dickerson and Gentry, 1983; Im, Bayus and Mason, 2003; Uhl, Andres and Poulson, 1970). This last finding that innovators have a higher degree of opinion leadership is very useful for this study as opinion leadership and social influence have been linked before (Eck et al., 2011; Lyons and Henderson, 2005; Rogers and Cantano, 1962). If the level of innovativeness is connected to the level of product leadership than the findings of studies that investigate the effect of product leadership on social influence might indirectly apply to the link researched in this study, the relationship between innovativeness and social influence. Therefore, it is useful to take a look at the studies that investigated the relationship between opinion leadership and social influence.
11
They play an important role in the information acquisition process of consumer decision making and consumers find the information they receive from communications with domain-specific opinion leaders more credible than advertising messages (Lyons and Henderson, 2005). In a recent study of Eck et al. (2011) the effect of opinion leadership on social influence has been investigated. Although the subjects in the research were children and thus the results might not hold for adults it did give insights about opinion leadership and social influence. According to their research, opinion leaders have a lower sensitivity to normative influence and a better ability to judge product quality. Furthermore, opinion leaders exert both normative and informative influence. Compared to a follower, which is the opposite of an opinion leader, normative influence is less important to an opinion leader. This suggests that opinion leaders can only be convinced of something when they are presented with hard facts and figures (informative influence). The difference between these types of social influences and the concept of normative and informative social influence will be discussed in paragraph 2.5 in more detail.
2.2 The influence of involvement
In this research we expect there to be a mediating role for the construct involvement in the link between consumer innovativeness and social influence.
Product involvement is a person’s relevance of the object based on inherent needs, values and interests (Zaichkowsky, 1985, 1994). Involvement with products has been hypothesized to lead to greater perception of attribute differences, perception of greater product importance, and greater commitment to brand choice (Howard and Sheth, 1969). Involvement with purchases leads one to search for more information and spend more time searching for the right selection (Clarke and Belk, 1978). An involved person thus has more product knowledge than a person with low level of product involvement. Mangleburg et al. (1998) found in their study that people with little product knowledge tend to look for more normative cues and those with (some) experience with the product already, tend to look for factual and credible information.
As discussed before, being innovative means that a person is one of the firsts to buy new products. To be able to be innovative this means that a person is up-to-date with the latest developments within a product (category). To be able to buy a new product as one of the firsts requires a person to know what is going on the market. Being up-to-date would require active information search which will result in more product knowledge than less innovative people. This connectedness with a certain product (category) is expected to be related to more product involvement.
2.3 The influence of homophily
12
2.3.1 Value- and status homophily
Value homophily refers to the fact that people feel justified in their opinions when they are surrounded by others that share the same values, attitudes and beliefs (Centola et al., 2007; McPershon et al., 2001). Thus, a person surrounds itself with people that think alike. These persons are strengthened in their opinions by the others, they for example think in the same way about politics, marrying before living together, finishing one study before hopping onto another etc..
Status homophily refers to the fact that people feel more comfortable when they interact with others that share the same cultural background (Centola et al., 2007). According to McPershon et al. (2001) status homophily is based on interacting with people with shared socio-demographic characteristics with, like having the same race, ethnicity, sex, age, religion, education and occupation.
Both value- and status homophily are forms of so-called ‘choice homophily’, where patterns of interaction are driven by preferences for similarity.
2.3.2 Induced homophily
Induced homophily refers to the fact that due to influence dynamics, individuals become more similar overtime. Other than with value and status homophily, induced homophily emerges not from individual choice. The disproportionate interaction of likes with likes is not a result of a psychological tendency - the preference for similarity - but rather the result of continuous interaction.
2.3.3 Homophily and social influence
Stemming from homophily a person’s social network can be divided into weak- and strong ties. According to Centola and Macy (2007) weak ties connect acquaintances who interact less frequently, are less invested in the relationship, and are less readily influenced by others. Strong ties on the other hand, connect close friends or kin whose interactions are frequent, affectively charged, and highly salient to each other. Strong ties increase the trust we place in close informants, the exposure we incur from contagious intimates, and the influence of close friends.
Caildini and Goldstein (2004) have stated that greater perceived similarity or homophily has been demonstrated to lead to enhanced compliance, even when the apparent similarities are based on superficial matches such as shared names, birthdays and fingerprint type. Thus, when it concerns interaction with people that are similar to us we tend to comply to social influence more easily. Also, according to Rogers (1955) the influence potential of network ties with an individual’s intimate friends is stronger than the opportunity for influence with an individual’s weak ties.
2.3.4 Reference groups
13
clique of friends, and extended family, a particular work group, a club or a school (close-links). The gender, ethnic, geographic and age group to which one belongs are also associative reference groups. Other types of reference groups are aspirational and dissociative reference groups. An aspirational reference group is a group that one admires and desires to be like; this mostly concerns distant-links like celebrities, such as TV-personalities and pop singers. A dissociative reference group is the complete opposite of an associative reference group; it is a group one does not want to emulate. This group’s behavior is disapproved of and not wished to copy.
The fact that persons use different reference groups for different kind of issues should be kept in mind for this study. The research should be designed in such a way that it not automatically applies to a specific set of people like family, friends or colleagues. It should be designed in a way that it is possible to be influenced by several people each belonging to a different reference group.
2.4 The influence of a person’s number of connections
The number of connections a person is believed and/or being able to have depends on the social network topology that is used. A distinction exists between random and scale-free networks.
In a random network each person (called a node) has approximately the same number of contacts. However, in a scale-free network each person is believed to have varying number of contacts. This type of network follows a power law, where the probability for each node of having number of neighbors is:
~
such that agents with only a few contacts are numerous, whereas a few agents have a very large number of contacts. The agents with a very large number of contacts are called hubs (Barabási et al., 2003; Delre et al. 2010).
Other than in the random network model, in the scale-free network model networks are believed to grow through the addition of new agents that are already present in the system, with a preference for agents with a large number of connections (hubs). This preference is often referred to as ‘preferential attachment’ (Barabási et al., 2003; Eck et al., 2011).
14
As innovativeness is related opinion leadership, this finding is important for this study. It might give indications of what the relationship between innovativeness and number of connections is, thus that innovativeness is positively related with number of connections.
2.5 Social Influence
2.5.1 Informative social influence versus normative social influence
As mentioned earlier we can distinguish between two types of social influences, a normative social influence and an informative social influence. Normative social influence is focused on obtaining social approval from others; it refers to influence to conform with certain expectations held by others (Cohen and Golden, 1972). Informative social influence is about obtaining truthful information through efficient information acquisition, sharing experiences and valuing expertise, not with the goal to conform to others but to behave correctly by forming an accurate interpretation of reality (Caildini and Goldstein, 2004; Burnkrant and Cousineau, 1975). Informative social influence refers to accepting information provided by others which is taken as evidence about reality (Cohen and Golden, 1972). Informative influence may be sought either directly, by seeking information from knowledgeable people and/or opinion leaders, or indirectly, by making inferences by observing the behavior of significant others (Bearden, Netemeyer and Teel, 1989; Park and Parker, 1977). In summary, normative influence is based on concerns about the group and one’s position in it, whereas informative influence is based on concerns about being accurate and arriving at correct solutions (Kaplan and Miller, 1987).
15
everyone has that same one is ‘not cool’ and it can make a person seem one of the herd and not having an own opinion which in turn in some social environments can be very important, having an own opinion. Third, in several situations it may take only one interaction to change behavior for a long period of times (Jager, 2010). Someone’s advice to use a certain toothpaste brand can lead to a habit for life. But as this has become a habit and the interaction took place a long time ago one may not be aware of the social influence anymore and consequently undervalue or neglect it.
2.5.2 Process-related differences
Previous research into this topic has mainly focused on revealing the conditions under which each influence is manifested and the process-related differences between normative and informative influence (Kaplan and Miller, 1987; Bearden, Netemeyer and Teel 1989, Park and Parker, 1977). A more recent study on social influence of Katona et al. (2011) focuses on social influence in online social networks. Companies are discovering the possibilities and efficiency of online marketing, the demand for online marketers is huge. Therefore, in order to make good online marketing campaigns the underlying processes of social influence in the online world become more important. The literature stream on this topic is expected to grow.
2.5.2.1 Social influence and judgmental and intellective issues
Kaplan and Miller (1987) discovered that the type of issue and decision rule affect the mode of influence used during group decision. They used two situational variables, type of issue and type of decision rule. The word issue can refer to a question one has concerning a certain product (category). At one end of the issue continuum is the intellective issue, these are issues for which there is considered to be a correct answer. For example, which mortgage is the best one to take. At the other end of the continuum is the judgmental issue, which involves behavioral, ethical or esthetical judgments for which there are no correct answers. For example, which wedding dress a person should buy depends on personal taste. The second situational variable used in the study is the group decision rule. Majority rule suggests that a majority of the group members has to agree on the issue and the unanimity rule suggest that there has to be a unanimous agreement among group members.
The study showed that when it concerns intellective issues there is a relative greater use of informative influence, because these issues tend to focus attention on factual information and accuracy. As where judgmental issues, that tend to elicit concern about the positions of group member and attainment of consensus, result in relatively more use of normative influence. The decision rule, the use of a unanimity rule as opposed to a majority rule appears to enhance this effect of issue type on mode of influence. A possible explanation for the moderating effect of decision rule is that unanimous agreement is often more difficult to reach than majority agreement. It may be that greater difficulty in agreement leads to more use of the type of influence that is most relevant to the issue type that is under discussion. When it concerns consumer behavior this might implicate that judgmental or intellective might also be applicable to product type, where for example clothing is more judgmental and mortgage more intellective. Thus, it seems that some products are by nature more likely to engage in informative than normative influence and the other way around.
16
indicate that in that case the normative or informative influence is even stronger than is the case where the opinion of others isn’t that important. Thus, when doing a study into social influence you should keep in mind that the nature of the product can bias your results. Therefore, choosing a ‘neutral’ product that is not dominated by intellective nor judgmental issues is the best option. In this study the cell phone is believed to be such a neutral product.
2.5.2.2 Personal characteristics and susceptibility to social influence
Bearden, Netemeyer and Teel (1989) focused on another condition that manifests process-related differences between normative and informative influence, namely consumer susceptibility to interpersonal influence. Consumer susceptibility to interpersonal influence is defined as the need to identify or enhance one’s image with significant others through the acquisition and use of products and brands, the willingness to conform to the expectations of others regarding purchase decisions, and/or the tendency to learn about products and services by observing others and/or seeking information from others. This susceptibility to social influences differs per person and is related to personal characteristics like e.g. self-esteem and intelligence.
Bearden et al. (1989) developed a 12-item questionnaire that measures consumers' susceptibility to social influence on two scales: normative (eight items) and informative (four items) influence. The questions have to be answered on seven-point bipolar agree–disagree scales. The normative scale includes items that assess the tendency to conform to the expectations of others (Burnkrant and Cousineau, 1975); the informative influence scale contains items to assess the tendency to accept the information from others as facts (Deutsch and Gerard, 1955). It seemed that people with a low self-esteem are more likely to comply with others as they have a greater fear for social disapproval and are keener on getting social approval. Insecurity caused by a lack of product experience can lead to higher susceptibility to social influence as well (Cohen and Golden, 1972). Also, people that are somewhat inexperienced with a product, as is the case when quality is ambiguous because of a lack of standards and one’s own ability to discriminate is not thought enough satisfactory, are more susceptible to social influence.
Furthermore, the older we get the less susceptible we are to social influence (Park and Parker, 1977), this is (a) a result of the learning we do in life and the (product)experiences we gain, (b) the accommodation of information (Jacoby et al., 1974) and (c) risk handling (Kogan and Wallach, 1967). A person’s dependence on social referents is inversely related to his confidence in his competence to judge the issue. In other words, does the person think that he is able to make a good decision? The learning in life and the experiences one gains will enhance this confidence. Furthermore, young people may be more susceptible to social influence since they would be expected to have more limited capacity to cope with uncertainty and risk than more mature individuals, due to limited financial resources for example.
2.5.2.3 Social influence and type of product (category)
17
the purchasing process of the second group. While for some product categories, like automobiles both normative and informative influences are important. This also seems to apply the product used in this study, cell phones. Though consumers obviously search for information on the functions and abilities of a certain cell phone, the decision on which cell phone to buy is often also accompanied by looking at what type of cell phone family and friends are using.
2.5.2.4 Social influence and the visibility and information density of the product
Besides dividing products in necessity and luxury items, the type of influence might also depend on the visibility of the product and its information density. Information density refers to the complexity of the decision task, it depends on how many attributes and brands there are available. A product scores high on information density when the product has many product attributes, thus if there is many choice within the product, or if there are many brands available in the product category. It seems plausible that for products that are visible in the daily environment and that are not information dense, such as clothing, normative influence dominates as normative influence is focused on obtaining social approval or avoid punishment from others. On issues that are more information dense and less visible, such as mortgages and insurances, it is more likely that informative influence dominates as the goal is to make a correct and intelligent decision. Privately consumed goods prioritize the informative influence, whereas for publicly consumed goods both types of influence are critical (Grewal, Metha and Kardes, 2000). The product used in this study, the cell phone, seems to belong to the type of product where both normative and informative influence are critical. A cell phone is visible in the daily environment, thus consumed publicly. The information density of the product is high as there are many choices within the product category: many brands, many operating systems, many functions etcetera.
2.5.2.5 Social influence in online social networks
As the popularity of social networks on the internet increases marketers see the potential to reach a lot of consumer and the possibility to gain al lot of information on their (potential) customer. For example, Google recently filed a patent for an algorithm that identifies so-called influencers on social networks. Several other firms (e.g., Idiro, Xtract) provide network analysis for the telecommunications sector and for various Web 2.0 platforms to assist in viral marketing campaigns (Katona, 2011). Companies see chances in identifying the influencers and to being able to measuring the probability that a certain person will adopt a certain (new) product.
18
Table 1 Literature overview: social influence
Study Results
Burnkrant and Cousineau (1975) People use others’ product evaluations as a source of information about the product. After observing others evaluating a product favorably, people perceive the product more favorably themselves than they would have in the absence of this observation. The evaluations of others are used as a basis for inferring that product is, indeed, a better product. This represents a departure from prior research which suggests that the responses of others establish a norm with which the subjects comply. People may frequently buy products that others in the groups buy, not to establish some self fulfilling role relationship to the others nor to obtain some reward or avoid punishment mediated by the others, but rather to acquire what they perceive as a good product. The mere demonstration of unanimity or consensus does not warrant an inference that the basis for that agreement is normative.
Bearden, Netemeyer and Teel (1989)
The authors developed a two-dimensional measure of informative and normative interpersonal influence and examined its reliability and validity. The construct is defined as the need to identify with or enhance one’s image in the opinion of significant others through the acquisition and use of products and brands, the willingness to conform to the expectations of others regarding purchase decisions, and/or the tendency to learn about products and services by observing others and/or seeking information from others.
Cohen and Golden (1972) Social influence is operative in situations not characterized by strong normative pressures. Buying decisions, even when the product or brand being judged is not novel or unfamiliar, seem to be characterized by uncertainty. This may stem in part from a lack of objective standards and a lack of reliable comparative brand information. Such conditions should tend to produce a heightened readiness to respond to apparently competent information from others. Thus, even when conditions favoring normative compliance are largely absent, there’s the pervasiveness of informative influence.
Deutsch and Gerard (1955) 1) Normative social influence upon individual judgments will be greater among individuals forming a group than among individuals who do not compose a group.
2) Normative social influence upon individual judgment will be reduced when the individual perceives that his judgment cannot be identified or, more generally, when the individual perceives no pressure to conform directed at him from others.
3) Normative social influence to conform to one’s own judgment will reduce the impact of the normative influence to conform to the judgment of others. 4) Normative social influence to conform to one’s own judgment from another as well as from oneself will be stronger than normative social influence from
oneself.
5) The more uncertain the individual is about the correctness of his judgment the more likely is to be susceptible to both normative and informative social influence in making his judgment.
Higby and Mascarenhas (1993) 1) Parent informative influence significantly dominates all other influences teens receive during ordinary and special food shopping. 2) Parent plays a significant higher informative and normative role in special than in ordinary food shopping.
3) Media informative influence is the next dominant influence among teens during food shopping.
Kantona et al. (2011) 1) The more friends of a person that already join the social network site the higher the probability that this person will join this social network site as well. 2) The higher the interconnectedness of these friends on the social network the higher the probability that you will join the social network site as well. Kaplan and Miller (1987) Intellective issues tend to focus attention on factual information and accuracy, leading to relatively heavier use of informative influence during group
19
normative influence. However, type of issue also appears to be moderated by the imposed decision rule. The relatively greater use of normative influence, when the issue is judgmental, and of informative influenced, when the issue is intellective, tends to increase when the decision rule is unanimity rather than majority. Kelman (1961) Social influence operates through one or more of the three distinct processes:
1. Internalization: individual accepts influence because the content is perceived as being inherently instrumental to the attainment of his goals. (informative influence)
2. Identification: individual adopts behavior or opinion derived from another because the role relationship between the individual and the other is beneficial to some portion of the individual’s self concept. (normative influence)
3. Compliance: individual conforms to the expectations of another in order to receive a reward or punishment mediated by that other. (normative influence) McGuire (1968) Most people are influenced by one’s society or by one’s significant others, although people differ in their responses to social influence. Susceptibility to
interpersonal influence is a general trait that varies across persons. In general, a person’s relative influenceability in one situation tends to positively affect that person’s influenceability in a range of other situations; people who conform to one influence source on one issue will be likely to conform to others sources on others issues.
Park and Parker (1977) Students when compared to housewives are (a) more likely to be receptive to reference group influence given the particular product, or (b) more receptive to reference group influence for a larger number of product cases. Possible reasons for this are:
1) Differences in age distribution. This age difference is often accompanied with a difference in (a) the amount of learning (e.g. degree of familiarity with the product), (b) the accommodation of information and (c) risk handling.
2) Need differences due to differences in their immediate social surroundings and daily activity patterns. Students are expected to have more frequent social contacts than housewives, and the group structure we expect students to interact within is expected to impose more rules and norms than is the case of groups in which housewives interact.
3) Need differences may exist due to differences in their stages in the socialization process. Venkatesan (1966)
1) In a consumer decision-making situation where no objective standards are present, individuals who are exposed to a group norm will tend to conform to that group norm.
20
2.6 Social influence: ingoing versus outgoing social influence
Both normative and informative influence can be in- and outgoing, thus an influence can be exerted or experienced. Little research has been done into this part of social influence yet.
What we assume is that when we are more susceptible to social influence because of insecurity we are less likely to exert social influence as it does not seem reasonable that an insecure person is able to exert social influence onto others. Thus, the level of ingoing social influence will be higher than the level of outgoing social influence. However, for being able to exert informative influence one needs to be accurate and correct. This does not depend on a person’s self-confidence. Therefore, if an insecure person is able to exert some kind of social influence it is most likely that this will be informative social influence instead of normative social influence. If an insecure person is able to exert social influence it seems likely this will be informative social influence rather than normative social influence. If one knows that the ‘insecure source’ is for example smart and experienced with the product category he might still accept the insecure person’s informative social influence. On the other hand, for normative influence this is less likely as getting social approval from one that is constantly seeking for social approval themselves seems less obvious
The opposite of an insecure person is a confident person, someone that trusts its one competence to make correct and accurate decisions. Previous studies (Bearden et al., 1989) already showed that these persons are expected to be less susceptible to social influence, thus it will take more effort to influence them which means that they probably experience a low level of ingoing social influence. We assume these persons are able to manifest themselves as innovators or as opinion leaders and expect them to be very good at exercising a social influence onto other people with their confident appearance, thus their level of outgoing social influence is expected to be high.
2.7 Conceptual framework and hypotheses
Above-mentioned studies do not answer the question whether a person’s innovativeness affects the relative strength of ingoing and outgoing social influence and the relative strength of normative and informative social influence. Also, no results were found yet for product involvement and number of connections to be mediating the effect of innovativeness on social influence.
21
2.7.1 Consumer innovativeness on ingoing and outgoing social influence
Figure 1 The effect of consumer innovativeness on outgoing and ingoing social influence and the mediating effect of product involvement and number of connections.
As consumer innovativeness is related with higher opinion leadership (Im, Bayus and Mason, 2003) and opinion leadership has been related to studies on social influence earlier it seems, though with caution, we may assume that the results of these studies might also apply to the relationship of consumer innovativeness on social influence discussed in this study. According to King and Summer (1970) opinion leaders articulate extensively about certain issues they are concerned with, this could be a certain product (category) for example. Articulating more about a certain product implies also more outgoing social influence, because there simply are more occasions to socially influence others. Also, Rogers and Cantona (1962) found that opinion leader exert an unequal amount of influence on others. They are often believed to be more credible than advertising messages (Lyon and Henderson, 2005). As informative social influence is more important to them than normative social influence and they are less susceptible to normative social influence, opinion leaders seem to be more difficult to convince. Before being influenced they need hard facts and figures, they will not comply to superficial arguments. This filter they apply to ingoing social influence seems to lead to less susceptibility to ingoing social influence. Therefore,
22
According to Midgley and Dowling (1978) innovators are persons that are among the firsts to buy a new product or innovative product. Knowing about these latest product or market developments seems to require to actively monitor the information stream within this product category or market. This would eventually lead to higher product knowledge. With all this information passing it seems that an innovator would know every little detail of the product (category) or market. Of product involvement we now that is related to a higher perception of attribute differences (Howard and Sheth, 1969). Attribute differences are the differences in the details of the product. Therefore, we believe that:
H2: a person’s level of innovativeness is positively related with the level of product involvement.
Product involvement leads one to search for more information on the product (Clarke and Belk, 1978). In time this will make this person experienced with the product and having gained high levels of product knowledge. The active search for information also exposes this person to ingoing social influence, as a more active information search creates more opportunities to be influenced. Furthermore, the product knowledge they gain with the information makes them an attractive information source for other people who have less product knowledge. Therefore, product involved people are expected to articulate more on the product (category) they are involved with. Therefore,
H3: a person’s level of involvement is positively related with (a) ingoing social influence and (b) outgoing social influence.
Im, Bayus and Mason (2003) stated that innovators have greater social ability. Eck et al. (2011) found that opinion leaders are centrally positioned in our social network. This means that information travelling through the network almost always travels through opinion leaders. For this to be possible it seems that opinion leaders have many connections otherwise it would not be able for them to touch upon all these information flows. As innovativeness is linked with high level of opinion leadership these findings result in the following hypotheses,
H4: a person’s level of innovativeness is positively related with the number of connections he has.
The connections we have is often a collection of persons we met or had contact with over the years. Park and Parker (1977) state that the older we get the less susceptible we are to social influence, this means the level of ingoing social influence will be low. Although no evidence for this is found, it seems also logical that with me more connections we are better to filter the ingoing social influence as we have many references. On the other hand, having many connections results in having access to many information sources, according to Barabási et al. (2003) this makes a person a desirable information source as well according to the ‘preferential-attachment’-theory. This person will exert social influence more often.
H5: the number of connections a person has is (a) positively related with outgoing social influence and (b) negatively related with ingoing social influence.
23
H6: the effect of consumer innovativeness on (a) the ingoing and (b) the outgoing social influence is mediated by product involvement and the number of connections one has.
2.7.2 Consumer innovativeness on the strength of normative social influence relative to informative social influence
Figure 2 The effect of consumer innovativeness on the strength of normative social influence relative to informative social influence and the mediating effect of product involvement and number of connections.
As already discussed in the beginning of this section opinion leaders find informative social influence more important than normative social influence (Eck et al., 2011). As they are less susceptible to normative social influence, opinion leaders seem to be more difficult to convince. Therefore, influencing an opinion leader is a hard job which involves being able to come up with accurate and credible arguments. Because of the product knowledge gained innovators are also believed to exert more informative social influence than normative social influence.
H7: for persons that are innovative the importance of normative social influence relative to informative social influence is lower than for persons who are less innovative.
People that are product involved are believed to have more product knowledge than others (Howard and Sheth, 1969). Mangleburg et al. (1998) found that little product knowledge tends to look for more normative cues while people that are already experienced with the product tend to look for more factual and credible information. Thus, informative social influence is more important for product involved persons. Also, Bearden et al. (1989) stated that confidence in a person’s own judgments makes this person less susceptible to social influence. Again, this implicates that in this situation a person seeks for factual and credible information. Having product knowledge makes a person more confident of their own judgment, therefore:
H8: for persons with a high level of product involvement the importance of normative social influence relative to informative social influence is lower than for persons who are less product
24
Earlier in this section we mentioned that having many connections makes a person a desirable information source according to the ‘preferential-attachment’-theory (Barabási et al., (2003). While innovativeness is expected to be positively related to the number of connections, persons with a high number of connections are believed to be expert in a certain product (category). In comparison to persons with lesser connections who do not have expertise in this field these experts are believed to e easier persuade others, with more superficial arguments, therefore,
H9: for persons with a high number of connections the importance of normative social influence relative to informative social influence is higher than for persons with a lower number of
connections.
Last, we hypothesize there to be a mediating effect of product innovativeness and number of connections on the relative normative strength, therefore:
H10: the effect of consumer innovativeness on the strength of normative social influence relative to informative social influence is mediated by product involvement and the number of
connections one has.
2.7.3 Degree of homophily on the strength of normative social influence relative to informative social influence
Figure 3 The effect of homophily on the strength of normative social influence relative to informative social influence
According to Rogers (1955) the influence potential of our friends is stronger than the influence potential of individuals that are not our friends. The people we hang out with most are called the strong ties in our social network, people we (almost) do not hangout with are called weak ties. According to Centola and Macy (2007) we do not place trust into weak ties therefore it is likely that these sources need to prove themselves by showing their credibility and accurateness, thus the informative influence is expected to be relatively greater than the normative influence here. As our intimate friends are often those we share a high degree of homophily we seem to engage in more normative influence than informative social influence, we hypothesized that the degree of homophily is positively related with the strength of normative social influence relative to informative social influence.
H11: for persons that share a high degree of homophily the strength of normative social influence relative to informative social influence is stronger than for persons that share a low of
25
3. RESEARCH DESIGN
3.1 Experimental design and subjects 3.1.1 Product choice
In this study the respondents will be asked about the purchase and use behavior in the mobile phone market. We focus on this market for several reasons. First, we needed a product that applies for both types of social influence. In the mobile phone market social influence – normative as well as informative – is inevitable as people use their phones in many ways. For some a mobile telephone means purely functionality, for others it is a fashion statement. Therefore, both normative and informative influence play a role in the purchase process. Besides, both influences are easy to access. Cell phones are very ‘transparent’ products in the way that a lot of information can be found on each brand and model in an informative way, like on the internet, and also a normative way. A cell phone is a product that can be seen everywhere, it is consumed in public. The presence of the cell phone in our daily lives is overwhelming. Just think of how many people you see using their cell phones when walking down the street. Or think of all the cell phones we see on the tables when walking through a restaurant. With this in mind it seems likely that people are influenced by what they see. Moreover, this makes cell phone users also aware of the fact that others see which cell phone they use.
Furthermore, cell phones are a very popular conversation topic. The rapid developments in cell phone technology are watched by millions of people who all have an opinion about this. The search function ‘which cell phone to choose’ (in Dutch) on Google brings along over 1.5 million hits. These hits mostly refer to a topic on an online discussion platform were people ask for advice of their fellow internet users.
3.1.2 Subjects
Suitable subjects for this research are persons that bought a cell phone over the last 12 months. First, the limit was set to six months because the questionnaire required doing some recalling on the purchase process of a cell phone. However, after some pre-testing it seemed that persons who bought a telephone over the last 12 months were also able to recall specific details on the purchasing process. Therefore, eventually the questionnaire was appointed at persons that bought a cell phone over the last 12 months.
The Dutch research company GfK expects that almost 6 million cell phones will be sold in the Netherlands in 2011, from which 2.8 million smartphones and 3 million normal cell phones1. The goal was to reach at least 100 consumers that bought a new cell phone over the last 12 months. We tried to reach especially consumers in the age of 18 till 30 years old, as we expected that within this group there’s the biggest chance of filled out questionnaires. We expected these people switch phone more
26
often than older people do. However, because of the snowballing effect of spreading the questionnaire electronically we might have also reached middle-aged and older people.
3.1.3 Questionnaire
An electronic questionnaire (appendix 1) was designed to collect the data that will help us to test the proposed hypotheses. The questionnaire was spread by sending a link to the researcher’s family, friends and acquaintances who were also asked to send the link to their family, friends and acquaintances. Also, an appeal for filling in the questionnaire and forwarding the appeal was posted on the researcher’s Hyves-, Facebook- and Twitter account. As the survey was spread through the internet we have to state that we made use of a convenience sample.
The questionnaire consisted of 42 questions spread over six different parts. To make sure only people who had bought a new cell phone in the last 12 months would fill in the questionnaire the first question was ‘when did you buy your latest cell phone?’. Those that answered ‘more than 12 months ago’ were redirected to the end of the questionnaire.
The first part of the questionnaire measured the subject’s level of innovativeness within the product group cell phones. Respondents were asked to rate to what extent they agreed with the six statements on level of innovativeness with cell phones. Second part of the questionnaire consisted of ten contradictive word-pairs that measured the respondent’s product involvement with the cell phone. Respondents were asked to scale which word of the word-pair was most applicable to them. The third part of the questionnaire measured the respondent’s network position, to see whether the respondent had a more central or less central network position. Questions were asked on interaction with the respondent’s friends and acquaintances. In the fourth part of the questionnaire people were asked to scale to what extent the persons on a prespecified list had been of influence in the purchase process of their last cell phone. Next, respondents were asked to answer questions about demographics and perceived similarity on those persons that, after scaling, turned out to be the three most important influencers. Part five of the questionnaire measured the social influence the respondent experiences and exerts. Respondents were asked to scale to what extent they agreed with the 24 statements. In the final part of the questionnaire the respondents were asked to fill in some details on their demographics.
3.1.4 Measuring instruments
To see whether our hypotheses hold the questionnaire consists of items that measure (1) consumer innovativeness, (2) product involvement, (3) network position, (4) homophily and (5) social influence. Each item is listed in table 3 with a short explanation.
27
Table 2 Means and (standard deviations)
Product Innovativeness 3.34
(.87)
Product Involvement 3.93
(.59)
# connections before purchase 4.31
(3.61)
# connections after purchase 10.15
(7.93)
Ingoing social influence 2.31
(.60)
Outgoing social influence 2.15
(.80)
Ingoing normative social influence 1.86 (.75)
Ingoing informative social influence 3.23 (.86)
Outgoing normative social influence 1.87 (.78)
28
Table 3 Measuring instruments
Item Scale Reference(s) Cronbach
Alpha
Consumer innovativeness
Six-item self-reporting scale that measures domain-specific innovativeness; which is the tendency to learn about and adopt innovations (new products) within a specific domain of interest (Roehrich, 2004), in our case the cell phone market.
Consumers have to answer on a 5-point Likert-scale to what degree they agree or disagree with a specific statement (1= strongly agree, 5= is strongly disagree).
Goldsmith and Hofacker (1991)
.831
Product involvement
Ten bipolar adjective scales (important vs. unimportant, means nothing vs. means a lot to me, not needed vs. needed, etc). The respondent has to rate on a 5-point Likert scale which side of the scale is most applicable to them.
Zaichkowsky (1994) .870
Number of connections
The respondent is asked to fill in the approximate number of people he talked with before buying the new cell phone and the approximate number of people he talked with after buying the new cell phone.
.372
Homophily The respondent is asked after the demographics of their influencers; gender, age, marital status, education and income. Next, the respondent has to rate on a 5-point Likert scale (1= very similar, 5= not similar) how similar their influencers is on the perceptual level (influencer is like me, thinks like me, behaves like me, has the same status as me, is from the same social class like me, culturally alike and is economically alike).
As the demographics are not measured in a scale-manner a new ordinal scale is developed by giving one point if gender, age, marital status and income of the influencer were similar to that of the respondent and zero points if this was not the case.
The average score on the perceptual homophily items was added to the average score on the demographic homophily item, which resulted in one homophily variable.
Jager (GfK questionnaire)
.668
Ingoing social influence
The respondent is asked to rate a total of 12 scales on a 5-point Likert scale (1= strongly agree, 5= strongly disagree). Eight of these scales measure the ingoing normative influence and 4 scales measure ingoing informative influence.
Bearden et al. (1989) .825
Outgoing social influence
The respondent is asked to rate a total of 12 scales on a 5-point Likert scale (1= strongly agree, 5= strongly disagree). Eight of these scales measure the outgoing normative influence and 4 scales measure the outgoing informative influence,
Jager (2009) .938
Normative social influence
The total of sixteen scales under ‘ingoing social influence’ and ‘outgoing social influence’ that measure normative social influence are summed and an average is calculated.
Bearden et al. 1989 Jager (2009)
29
Informative social influence
The total of eight scales under ‘ingoing social influence’ and ‘outgoing social influence’ that measure informative social influence are summed and an average is calculated.
Bearden et al. 1989 Jager (2009)
.732
30
4. RESULTS
4.1 Sample
In total 154 respondents completed the questionnaire. Approximately 178 respondents did not complete the questionnaire. This number was determined by looking at who made it to the end of the questionnaire. This showed that most respondents left the questionnaire when they had to fill in the demographics of their influencers (part 4), assumingly respondents found this was too much work and taking too long.
The sample consisted of 77 males and 77 females. The age distribution shows a peak at age 24 and 25 (cumulative 42,9%) and the youngest respondent was 18 years old while the oldest was 59 years old. The marital status of these respondents is fairly evenly distributed, 28.6% is single, 17.5% has a relationship but does not live together, 28.6% is living together and 25.3% is married. Almost half of the respondents has completed a ‘HBO’ education and most of the respondents (cumulative 62.3%) earns up to € 2,500 a month.
Looking at the characteristics and different backgrounds of the respondent it seems that it is a fairly representative sample.
4.2 Mediation analysis
In this section the analysis results will be shown of the models and their corresponding hypotheses. Two of the three models are multiple mediator models that are examined with the help of a mediation analysis.
To see whether there is mediation the bootstrap test of Preacher and Hayes (2004, 2008) is used in this study. The study of Zhao et al. (2010) will serve as a guide of how to perform this specific mediation analysis, this article describes step-by-step how to perform the bootstrap test and it tells how the results of the test should be interpreted.
To illustrate the mediation analysis used in this study works, an example is given with the help of part of our own model (figure 4).
31
Path ‘a’ represents the direct effect of independent variable consumer innovativeness on the mediator product involvement. Path ‘b’ shows the direct effect of the mediator product involvement on the dependent variable ingoing social influence. Path ’c’ is the direct effect of the independent variable consumer innovativeness on the dependent variable ingoing social influence. A path ‘c’” is the indirect effect, thus the effect of consumer innovativeness on ingoing social influence when mediated by product involvement.
There are several reasons why in this study the bootstrap test is used and not the Sobel z-test designed by Baron and Kenny (1986), which is today’s most commonly used mediation test.
4.2.1 Sobel z-test
Baron and Kenny‘s (1986) mediation analysis is focused on the statistical significance of the a and b paths (figure 4).
According to their causal steps approach a variable functions as a mediator when it meets the following conditions: (a) variations in levels of the independent variables significantly account for variations in the presumed mediator, (b) variations in the mediator significantly account for variations in the dependent variable and (c) when paths a and b are controlled, a previously significant relation between the independent and dependent variables is no longer significant, with the strongest demonstration of mediation occurring when Path c is zero.
Thus, one needs to test the following equations to see whether there is mediation:
(1) regress the mediators product involvement and number of connections before and after purchase on the independent variable consumer innovativeness (path a):
Involv = i + *Innov + e
(2) regress the dependent variable ingoing social influence on the independent variable consumer innovativeness (path c’), like:
Ingoing = i + c’*Innov + e
(3) regress the dependent variable ingoing social influence on both the independent variable consumer innovativeness and the mediator product involvement (paths b and c)
Ingoing = i + c*Innov + b*Involv + e
