It’s a small, commercial & happy world after
all!
A study on the consequences of Social Network Sites on
mood state and brand preference
Elisabeth de Mol van Otterloo
6386954
Thesis Business Studies
H.H. Lee
Amsterdam Business School - University of Amsterdam
October 2013
Preface
It is done, I am finished. Making this thesis was an interesting and eventful path with the necessary ups and downs. I have come to the conclusion that this ‘project’ is a process, in which I do not only get to know more and more about the background of the content of my thesis, but I also did get to know myself a bit better.
After three years of studying History of Art (it made SPSS a tough part of the project…) at the University Leiden and earning a Bachelor of Arts-degree, and now finishing a master in Business Studies at the University of Amsterdam, the time has come to find a proper job.
Due to my creative background and some experiences in the commercial and
organizational world, I decided to specialize in Marketing. I chose to write my thesis about social network sites and the mood states and brand preferences from my own experience and troubles with SNS. I wanted to know more about this subject.
I could not have written and completed this thesis without the support and help others. First, I would like to thank my supervisor Meg Lee. In the end, I realized she actually was my coach and has helped me find my way in this process. Thank you for the time and effort and I wish you all the best now that your life is about to change. I would like to thank everyone that has completed this survey. And finally, I would like to thank my family and friends for their support and involvement.
Elisabeth de Mol van Otterloo
Abstract
The Social Network Sites (SNS) has become a growing area of interest for marketers and researchers. For the consumers, the purpose of the SNS is to publish personal information and communication online and share this with other people. The marketers have become aware that the SNS is an important source of creating awareness for the brand and a way to address and influence the potential customers directly by creating communities on the SNS and post messages about the brand and its products. The members can subsequently like and share this with their online friends.
In this study, the consequences of a positive message on the SNS on the receiver’s mood state and brand preference are measured. It is examined whether the receiver of the message follows the path of social comparison or social contagion when getting in a certain mood state and if tie-strength with the sender of the message plays a moderating role in the effect size of the mood state of the receiver.
Experimental research is done to get the proper data to draw conclusions. The results show that positive messages do not significantly create a positive or negative mood state at the receiver. Social comparison and social contagion do not play a significant role in the paths towards a particular mood state. Tie-strength plays a moderating role only in the positive mood state, and finally, the mood state, obtained by seeing the message on the SNS, does not significantly influence brand preference.
Table of Contents
Preface 2
Abstract 3
Table of Contents 5
Chapter 1: Introduction 7
1.1 The development of Social Network Sites 7
1.2 The impact of SNS on consumer behaviour 8
1.2.1 Consumer Behaviour and Social Comparison & Social Contagion 8
1.2.2 Mood State 9
1.2.3 Tie Strength 10
1.3 Problem Statement 10
1.4 Delimitations of the study 11
1.5 Contribution 11
1.5.1 Theoretical Contributions 11
1.5.2 Managerial Contributions 12
1.6 Structure of the research 12
Chapter 2: SNS and Social Comparison & Social Contagion 14
2.1 SNS and Social Comparison 14
2.1.1 Theory of Social Comparison 14
2.1.2 SNS and Social Comparison 15
2.2 SNS and Social Contagion 16
2.2.1 Theory of Social Contagion 16
2.2.2 SNS and Social Contagion 17
Chapter 3: SNS and Tie-strength 19
3.1 In-group versus Out-group 19
3.1.1 In-group and out-group 19
3.1.2 Tie-strength and Social Comparison 19
3.1.3 Tie-strength and Social Contagion 20
Chapter 4: Mood State & Brand Preference 21
4.1 Brand Preference 21
4.2 Mood State and Brand Preference 21
4.2.2 Mood State and Brand Preference 22
Chapter 5: Conceptual Model 23
Chapter 6: Methodology 24 6.1 Research Design 24 6.1.1 Questionnaire 25 6.2 Conduct of Experiment 26 6.2.1 Participants 26 6.2.2 Procedure 27 6.2.3 Measures 28 6.2.4 Pre-test 31
6.3 Data Preparation & SPSS-tests 32
Chapter 7: Results 36 7.1 Cronbach’s Alpha 36 7.2 Sample Profile 36 7.3 Control check 37 7.3.1 Realistic situation 38 7.3.2 Strength relationship 38 7.4 Hypotheses testing 38 7.4.1 Hypotheses 1 & 2 38 7.4.2 Hypothesis 3 39 7.4.3 Hypotheses 4a & 4b 39 Chapter 8: Discussion 41 8.1 Research Results 41 8.1.1 Mood states 41 8.1.2 Tie-strength 42 8.1.3 Brand preference 43 8.2 Marketing implications 43 Chapter 9: Conclusion 45 9.1 Research question 45
9.2 Limitations and future research 45
References 48
Appendices 59
Appendix 2: Control check – SPSS results 65
2.1 Realistic situation: One Sample t-test 65
2.2 Strength relationship: One Sample t-test 66
2.3 Control check: One Sample t-test 67
Appendix 3: Hypotheses testing 68
3.1 Control checks: One Sample t-test 68
3.2 Hypotheses 1 & 2 69
3.3 Hypothesis 3 70
Chapter 1: Introduction
The introduction section presents a short summary of the emergence of Social Network Sites (SNS) and the impact of SNS on consumer behaviour. Also, the research question, delimitations and theoretical and managerial contributions are presented. This chapter ends with the outline of the research.
1.1 The development of Social Network Sites
Social Network Sites (SNS) have become a considerable part of the modern life of the consumer. On these websites, members of the SNS spend their time maintaining social relationships and posting and sharing information, messages and pictures (Buechel & Berger, 2012). The first page members see when logging in on the SNS is their so-called wall. On this page, members get an overview of the latest posts of and about their online friends. It is directly seen what information and communication is published and shared with other people. Sharing means copying a message from another person or a website and put this on your own wall.
The concept of privacy has changed these recent years now that everyone is aware of the fact that there is a possibility that snapshots of their lives can be published online (Dwyer, et al., 2007). People do not always have control of the information that is posted about them, as others can also post and share information about that particular person.
The content of online messages is generally more positive than negative (Berger & Milkman, 2011) and the information people publish about themselves is more self-enhancing than self-destructing (Mukesh & Gonçalves, 2012). According to Hutto, et al. (2013), expressing positive sentiment has a significant positive effect on the follower gain, whereas expressing negative sentiment has an adverse effect (p. 1). The advantage of a large audience of followers is that access to a network of social ties, influence, and resources are created (p. 1). Therefore, the user posts and shares three times more likely positive messages rather than negative messages (Hutto, et al., 2013; Gruzd, et al., 2011).
The members of the SNS are not the only ones that stimulate this positivity. This is also done by the organisation of the SNS itself. Below every posted and shared message, there is the ‘like’ button. If a member likes that message, he or she can show this by clicking on the corresponding icon. There is not the option to dislike a message.
Also, the sites only send out ‘positive’ notifications to the users. Unpleasantness, such as rejection and failed friendships is avoided by not reporting these negative reports (Ramanigopal, et al., 2012). If an individual has requested an online friendship with someone and this request has been declined, the individual will not receive a message of the SNS about this rejection. If a friendship request has been accepted, the individual will than receive a notification about this acceptance.
Finally, people are frequently subjected to these positive incentives, as it is not only possible to see the ‘wall’ on the website of the SNS on a computer. With the smartphones and the associated applications, people are online throughout the day. Users of SNS can be continuously inundated with these – mainly positive - incentives, because they receive a notification on their phone once somebody has left a message or once someone has ‘liked’ a picture on their wall. In this thesis, it is about messages with a positive content (i.e. a message that expresses a positive sentiment).
1.2 The impact of SNS on consumer behaviour
The question is what all this positivity does to the behaviour of the receiver of the message (in this thesis called ‘receiver’) and if it causes a positive mood state at the receiver. Also, the source of the message, in this thesis called ‘sender’, may be of influence on the effect size of the mood state of the receiver. A person will react differently and will get in a different mood state when the sender is a close friend instead of a distant acquaintance. These two issues are discussed in the next sections.
1.2.1 Consumer Behaviour and Social Comparison & Social Contagion
All these online messages, positive and negative, have an impact on the consumer behaviour. In general, messages on the SNS are of influence on the consumers’ attitude (Aral & Walker, 2012; Romero, et al., 2010), on its degree of socialization and communication, on their decision-making process (Gatautis & Kazakevičiūtė, 2012), and on their preferences and consumption (Durukan, et al., 2012).
Social comparison and social contagion are two mechanisms of consumer behaviour that occur while looking at other people’s online messages (Bello, 2008; Aral & Walker, 2012; Hinz, et al., 2012). Social comparison means people have a tendency to compare themselves with others (Wu & Lee, 2008; Knobloch-Westerwick & Westerwick,
2011; Festinger, 1954). Social contagion is the extent to which people have the tendencies to copy or adopt behaviour from others (Barash, 2011).
These two mechanisms can influence the impact online messages have on the consumer behaviour, thus referring to the consequences mentioned in the first paragraph of this section. In the context of social contagion, the consumer can be afraid of social exclusion, which influences his or her degree of socialization and communication (Barash, 2011). The consumer will do its best to ‘fit in’ within his or her environment by copying the behaviour of someone else. Or they want to learn from other people’s messages through comparison, which influences the consumers’ attitude and their decision-making process (Wood, 1989; Festinger, 1954). By making a comparison with someone else, the captured knowledge might change the attitude towards, for example, a certain product. This development can also be of influence on the decision-making process, as this knowledge can change the choices the consumer makes regarding the purchase of a product.
In this thesis, it is examined if the receiver is following either one of the paths of social comparison or social contagion before getting in a particular mood state. Making a comparison with a person can have a different effect on the mood state than copying someone else’s behaviour. Social comparison and social contagion are further explained in Chapter 2 of this thesis (p. 14).
1.2.2 Mood State
Mood state is a set of affective factors, subjectively perceived by individuals that have an influence on the consumer behaviour, e.g. brand selection and advertisement exposure (Gardner, 1985). They are transient, but result from a specific situation and time (Hirschman & Stern, 1999).
Seeing other people’s information on the SNS can influence the mood state of the receiver. According to Bollen et al. (2010), the reports of social events on the SNS do have an influence on the fluctuations in mood levels. In both the physical and in the virtual world the negative emotions are stronger and more intense experienced than positive emotions (Tugend, 2012; Wu & Lee, 2008). This mood state, either positive or negative (Sahney, 2012), can have a direct effect on the receiver’s ideas and preferences towards the brand or event that they were initially affected by (Gardner & Hill, 1989).
In this thesis, the focus is on the effect on the mood state of the receiver when seeing a message on the SNS and if this mood state affects the brand preference towards the brand mentioned in the message of the case. The concepts of mood state and brand preference are further explained in Chapter 4 (p. 22).
1.2.3 Tie Strength
The sender of the message may also be of influence on the behaviour of the receiver of the message, in particular on the effect size of the mood state the receiver gets into when seeing the positive message of the sender.
In the social environment of an individual, a distinction can be made between in-group and out-in-group (Knobloch-Westerwick & Westerwick, 2011; Barash, 2011). The in-group is a social category or group and the members of this group are the people the individual identifies strongly with (Giles & Giles, 2013; Schubert & Otten, 2002). Out-group is a social category with which the individual does not identify (Giles & Giles, 2013; Schubert & Otten, 2002). In this thesis, it is examined if tie-strength between the sender and the receiver plays a moderating role in the effect of mood state. This effect is tested with the mechanisms social comparison and social contagion. The concept of tie-strength is discussed extensively in Chapter 3 (p. 19).
1.3 Problem statement
This paper examines how positive messages, posted on the SNS, impact the mood state and the brand preference of the receiver. Regarding the relationship between the sender and receiver of the message, a distinction is made between in-group and out-group, as this factor may be of influence on the strength of the mood states and brand preference as perceived. Furthermore, the theories of social comparison and social contagion are researched, as they may be determinative in the mood state and brand preference of the receiver.
The central research question is:
What is the influence of positive messages posted on SNS on the mood state and brand preference of the receiver of these messages?
The research question is divided in the following sub questions:
- In what mood state does the receiver come when seeing positive messages posted on the SNS?
- Does the receiver of the message follow the path of social comparison or the path of social contagion when getting in that particular mood state?
- Does tie strength have a moderating role in the effect size of the mood states? - Are both a positive and a negative mood state beneficial for the brand
preference after seeing a particular post online that involves a brand?
1.4 Delimitations of the study
There are a few delimitations in this study. First, this research is only about positive messages, as the content of the majority of the posted and shared messages on the SNS is positive in nature. The negative message is excluded from this study. Secondly, there are a limited number of positive and negative mood states used for measurement in this study, as there is no previous research done about the relationship between SNS and mood states. The mood states that are used for measurement in this study are scientifically positive and negative mood states that do have a relationship with online marketing and branding activities. The latest fact is necessary for this study, as SNS has become an essential part of online marketing and branding activities of organisations.
Thirdly, this study is about messages that are only about the sender. No other people are subsequently involved in this message. Normally, there are messages on the SNS that are not only about the sender, but also about, for example, the sender and their friends, such as a group photo. Due to practical problems it is not possible to test the mood state and brand preference of the receiver when looking at a message that is about a group of people.
1.5 Contribution
1.5.1 Theoretical contributions
This study makes a methodological contribution to the debate on how SNS affects the consumers. Is the development of sharing (positive) snapshots of private lives with the public a positive one and does it also make everyone truly happier? Secondly, this study is about the thinking process of the receiver when looking at a message, with regards to the mood state and brand preference, and whether the receiver is going through the
processes of social comparison or social contagion. Finally, it is also measured if the degree of influence on mood state and brand preference is determined by tie-strength between the sender of the message and the receiver.
This study will provide a reality check in this virtual social world of positivity.
1.5.2 Managerial contributions
This study will also make a few important contributions for the marketing department of organizations that are active on the SNS. The results will show if there are effects of online social comparison and social contagion on the mood state of an individual. Online social comparison and online social contagion are two mechanisms in which the consumer creates an opinion and a possible preference towards a good (i.e. tangible and intangible goods). Former research was focused on consequences of SNS on consumers, mainly in the context of the changes in people’s decision-making process and their purchasing behaviour (Wu & Lee, 2008; Gatautis & Kazakevičiūtė, 2012). This research focuses on brand preference of the individual through SNS, but with a link to the mood state of the individual and the mechanisms social comparison and social contagion. The results of this research can help organizations in their online marketing activities regarding influencing the consumer and their brand preference.
The degree of influence on mood state and brand preference is in this study measured by tie-strength between the sender and the receiver, which can be of importance in the organizations marketing strategies. Can the organizations keep their marketing activities general on the SNS or does tie-strength have a significant influence on the effects of mood state that it necessary to create a campaign in which people are personally addressed to get a significantly more positive word-of-mouth and brand preference?
1.6 Structure of the research
In this thesis, chapter two till four will examine the past literature on (online) consumer behaviour, the theories of social comparison and social contagion, tie-strength, and brand preference. All these subjects are studied in the general context and in the online context. Chapter five will outline the conceptual model to give an overview how these variables are connected to and with each other in this research, based on the literature review. Chapter six will explain the methodology of this research, with an explanation of
the questionnaires and the necessary methods from SPSS to conduct the experiment. In chapter seven, the results of these calculations will be given, and whether the hypotheses are supported or not. Chapter eight will contain a discussion around the results and finally in chapter nine an answer will be given to the research question, also on the basis of the sub-questions, with an explanation of the limitations and implications for future research.
Chapter 2: SNS and Social Comparison & Social Contagion
In this chapter, the theories of social comparison and social contagion are described. A short and general description of both theories is given. Furthermore, this chapter contains also an explanation of the relationships of both theories with SNS. The hypotheses regarding these issues are presented in this section.
2.1 SNS and Social Comparison
2.1.1 Theory of Social Comparison
In general, humans have a drive to evaluate their own opinions and abilities by comparing themselves with other people (Festinger, 1952; Wood, 1989; Wu, et al., 2012). Social comparison is the process of thinking and reflecting the information about other people in relation to the self (Wood, 1996), and it is a phenomenon that comes across everyone from time to time (Gibbons & Buunk, 1999). These comparison processes occur especially when individuals are confronted with the performance of another person that is on a domain of high relevance to our self and when this person is psychologically close (Smith, 2000). There are three underlying motives that move people to compare themselves with others: improvement, evaluation, and self-enhancement (Wood, 1989; Corcoran, et al., 2011). With self-improvement as a motive, people would like to learn more about their abilities and in doing so improve these skills by comparing themselves with and learning from other people (Wood, 1989). The second motive, self-evaluation, is the process by which the self-concept is socially negotiated and modified (Sedikides & Strube, 1997). If self-enhancement is the drive of social comparison, people are motivated to protect themselves from any negativity and to elevate the positivity of their self-conceptions (Sedikides & Strube, 1997). Uncertainty is also associated with social comparison, increasing the amount of comparison during periods of stress, change or novelty, situations that cause uncertainty (Gibbons & Buunk, 1999).
Consequences of social comparison are, among other things, the fact that the information individuals compare themselves with, will be considered as a standard. Social comparison may result in assimilation if the standard is a moderate one, and in contrastive consequences if this standard is not moderate (Corcoran, et al., 2011). This standard is used as a basis for self-improvement, self-evaluation (Corcoran, et al., 2011), and self-enhancement. Emotional consequences are also involved in social comparison
(Wu, et al., 2012). Individuals have a tendency to make a downward or upward comparison (Wu, et al., 2012; Wheeler & Miyake, 1992; Smith, 2000). Downward comparison occurs when an individual compares itself with someone who has less (Wu, et al., 2012), and he or she will have the tendency to interpret others as worse-off to give its own well being a boost (Wheeler & Miyake, 1992; Suls, et al., 2002). Thus, this is positive for their mood state. According to Taylor & Lobel (1989), upward comparison takes place when people have a strong desire to have contact with more fortunate others as a preference to interact with or to gain knowledge about, as these people are better off (p. 569). The individual compares itself with someone who has more and has the tendency to interpret others as superior, hoping to become sanguine and inspired (Wheeler & Miyake, 1992; Suls, et al., 2002). This can intensify the negative self-evaluations and lead to a negative mood state (Harper, et al., 2007).
2.1.2 SNS and Social Comparison
The online profiles and messages offered on the SNS are also the basis for social comparison processes, and the motives improvement, evaluation and self-enhancement are also applicable in the context of SNS. After being confronted with online information about or from others, people have the tendency to relate themselves to that information, to evaluate their own attitude and to learn from other people’s messages (Haferkamp & Kramer, 2011). Members focus their energy on improving their relative standing (Harper, et al., 2007).
Although people have the tendency to relate themselves to information published online and make comparisons, former research did not indicate what the mood state of the viewer was directly after making that comparison.
As this thesis is about positive messages on the SNS, the assumption is made that the sender is having a better time or a more positive experience than the receiver has. In this case, the receiver will compare itself with someone who is in a better position and is more fortunate at that moment (Taylor & Lobel, 1989). Thus, upward social comparison will take place. Reading the positive message will decrease the receivers’ feelings of well being, as he or she is having a worse time than the sender. The assumption is made this will lead to a negative mood state at the receiver.
This leads to the following hypothesis:
H1: A positive message on the SNS will trigger negative feelings at the receiver, as he or she is going through the route of social comparison.
2.2 SNS and Social Contagion
2.2.1 Theory of Social Contagion
Watching other people’s messages and pictures online can also make individuals want to copy or adopt behaviour people show in the message. Social contagion is the extent to which people have the tendency to copy or adopt behaviours from others, the so-called copycat behaviour (Barash, 2011). The key mechanisms regarding social contagion are normative influence, social learning, social identity, and social exclusion. (Barash, 2011; Iyengar, et al., 2012; Aral, et al., 2009). Normative influence implies that if someone adopts a certain product or embraces a belief, the other will likely follow that person and will do the same thing, hoping to conform to the expectations of that person (Barash, 2011; Iyengar, et al., 2012). With social exclusion, non-adopters are discouraged to continue non-adopting, enabling the diffusion of social contagion (Barash, 2011). In fact, social exclusion acts in an opposite way of normative influence, but with the similar results. Social learning refers to social influence in which information from others serves as evidence about reality, changing the belief about the truth and convincing the individual the innovation is worth adopting (Aral, et al., 2009; Peyton Young, 2007).
Social contagion can occur on the behavioural level, also called behavioural contagion (Marsden, 1998). This means that the behaviour of the receiver has changed to become more like that of the sender by copying their behaviour. This change occurs in a social interaction, but the sender has not communicated with the intention to evoke such a change (p. 179).
Social Contagion can also occur on the emotional level, which is called emotional contagion (Marsden, 1998). This is the process by which individuals copy the emotions of the people around them. This process starts with inducing consciously or unconsciously emotion states and behavioural attitudes by the sender (Barsade, 2002; Neumann & Starck, 2000). People have the tendency to automatically mimic and synchronize their facial expressions with those of other persons (Tamietto & de Gelder,
2008). According to Pugh (2001), exposure to an individual who expresses positive or negative emotions can produce a corresponding change in the mood state of the observer, in this case the receiver of the message (p. 8). Positive emotional contagion means seeing someone else’s good mood and, as a consequence, experiencing also a positive mood (Barsade, 2002). The display of positive emotions is positively related to consumer’s positive affect and tends to a greater emotional contagion than negative emotions do (Pugh, 2001; Belkin, 2006).
2.2.2 SNS and Social Contagion
It is still scientifically unknown if people get in a positive mood state when seeing a positive message on the SNS, and if the receiver than actually follows the path of social contagion.
However, it is known that, in general, a minority belief can become a dominant one as more people choose to adopt statements and opinions from friends and share these on the social networks (Barash, 2011). Online social contagion is also of influence on people’s purchase decisions, as people are willing to undertake physical action to copy the exact behaviour or image of others that is visible on the walls of SNS (Ghose, et al., 2011; Aral & Walker, 2012; Iyengar, et al., 2011).
A function on the SNS that can cause social contagion is sharing. People can copy other people’s messages and paste this on their own wall, because they find these interesting. Sharing can work contagious, as the information is continuously forwarded and spreads itself over the SNS. Previous research on user motivations of SNS indicates that the four main reasons why people like to share - often personal - information with others are the need to be part of a group, a need to be individualistic, a need to be altruistic, or a need for personal growth (Ho & Dempsey, 2009).
The consequences of social contagion can be positive for the receiver and give a boost to the level of self-confidence as they ‘fit in’ with other people (Barash, 2011), have spread awareness for something they like (Ghose, et al., 2011), or have demonstrated their knowledge (Owyang & Li, 2012). Also, as mentioned in the final paragraph of section 2.2.1, the display of positive emotions are in general positively related to the consumer’s positive affect. On the basis of these facts, assumptions can be made about the mood state of the receiver when seeing a positive message on the SNS.
These facts lead to the following hypothesis:
H2: A positive message posted on the SNS will trigger positive feelings at the receiver, as he or she is going through the route of social contagion.
Chapter 3: SNS and Tie-strength
This chapter focuses on the difference between in-group and out-group. Also, the relationship between tie-strength and SNS is explained. The established hypotheses regarding these relationships are presented in this section.
3.1 In-group versus Out-group
3.1.1 In-group and out-group
There is a distinction between in-group and out-group members and the way these people react to each other, also with regards to social comparison and of social contagion (Knobloch-Westerwick & Westerwick, 2011; Barash, 2011). In-group is a social category or group with which you identify strongly (Giles & Giles, 2013; Schubert & Otten, 2002). This identification and selection can be based on homophily, which states that connected individuals, people that are in some respect similar to each other, are likely to develop social relationships (Aral, et al., 2010; Barash, 2011). Social identity is the mechanism in this case (Barash, 2011). This identification can be based on different categories and levels, such as language and speech styles that are created and used within a group, a certain religion, the dress code that is adopted in the group or certain traditions and rituals that are highlighted with particular festivals (Giles & Giles, 2013). It can also be based on sport teams and school classes (Stenstrom, et al., 2008). Out-group is a social category and these people are ones with whom you do not identify (Giles & Giles, 2013; Schubert & Otten, 2002).
3.1.2 Tie-strength and Social Comparison
In general, comparing with members of the in-group often has a greater effect on self-evaluations than making a comparison with members of the out-group (Blanton, et al, 2000). According to Blanton, et al. (2000), the reasons for this are, among other things, the fact that a member of the in-group is more likely to be viewed as someone who is similar to the receiver and thus a more informative standard of comparison, the fact that members of the in-group feel an emotional bond with each other, and the fact that the abilities of one person are often defined in relation to the standards of their in-group.
Upward comparison with in-group members affects the self-esteem of the receiver and their feelings of depression more than when making this comparison with members of the out-group (Major, et al., 1993). People consider upward comparison
with out-group members as non-self-relevant (Martinot, et al., 2002). This moderating effect is non-existent with downward comparison.
In the context of the SNS, after seeing a message on the SNS from a member of the in-group, the receiver may face a sense of loneliness (Keen, 2012; Turkle, 2011) and discouragement, because the receiver realizes he or she not having that good of a time as their close friends on the picture. When seeing a similar message on the SNS of a person from the out-group, these feelings may be less strong, as there is no emotional, cultural or cognitive connection with him or her (Giles & Giles, 2013).
3.1.3 Tie-strength and Social Contagion
According to Ghose, et al. (2012), the impact of social contagion is in general moderated by the tie-strength people have with each other. Ties with the members of the in-group causes a stronger effect of social contagion than ties with members of the out-group.
An in-group member can also influence the mood state of a friend consciously or unconsciously, thus the individual emotions become eventually a collective emotion that at least two people share (Bosse, et al., 2009). According to Coenen and Broekens (2012), there is in general significantly more positive emotional contagion between in-group members than there is towards an out-in-group member. This is not the case with negative emotional contagion; there is no difference in effects of contagion regarding tie-strength.
A message posted on the SNS is a direct source of information regarding the mood state or well being of the sender. Based on the findings from researches regarding social contagion and tie-strength, the assumption can be made that seeing a positive message may cause a positive mood state at the receiver. This effect may be larger compared to the mood state captured when seeing a message from someone from the out-group.
Based on the findings from past literature regarding the relationship between mood states, social comparison and social contagion and the role of tie-strength in these relationships, it can be assumed that this role is a moderating one.
These statements lead to the following hypothesis:
Chapter 4: Mood State & Brand Preference
This chapter gives an explanation of the concept brand preference. The relationship between mood state and brand preference is clarified, as is the way people in certain mood states process information to form an opinion about a brand. Also, hypotheses are presented regarding the effects of mood states on brand preference.
4.1 Brand Preference
A brand name signifies a level of quality and consistency consumers can trust on (Azzawi & Nzube Ezeh, 2012), and it is an important attribute in determining preference (Cobb-Walgren, et al., 1995). Brand preference is the reflection of the biased behavioral tendencies on the consumer’s predisposition towards a brand (Ebrahim, 2011). According to Ebrahim (2011), a brand preference facilitates consumer decision-making and activates brand purchase. Next to the importance of the analysis and communication of brand and product attributes, also other, more subjective, factors as emotional responses and social influences have an effect on consumer’s brand preference (p. 2). This means that not only the mention of a brand name, but also the receiver’s mood state and network ties affects their brand preference. In this thesis, the focus is on the relationship between mood state and brand preference in the context of SNS.
4.2 Mood State and Brand Preference
4.2.1 Branding and SNS
There has not been any research done yet about brand mentions on the SNS and the relationship between the receiver’s mood state and its possible brand preference through this mention. It is clear though, that with the emergence of SNS, marketing strategies of organizations are changing. New opportunities have come to create brand awareness and brand preference. Key is to stand out from the crowd and create an own identity, as there is the possibility to reach the consumers more directly and personally.
Organizations can advertise on the SNS and create online communities to keep their potential customers up to date regarding the latest products, but there is also the possibility that customers mention the brand themselves on the SNS. Recommendations and information about brands, given by friends, might be considered reliable, as these people are trustworthy and known to the receiver of the message.
4.2.2 Mood State and Brand Preference
For organizations, it might also be useful to know what messages on the SNS do to the mood state of potential customers, as this might influence the brand perceptions, brand awareness and even the brand preference. According to Myers and Sal (2013), people in a positive mood tend to focus on the overall meaning of the persuasive message and make meaningful mental associations (p. 10). A positive mood generally increases the positive affective state and the subsequent brand attitude (p. 9), thus the consumer will think positive about the brand. Individuals in a negative mood generally process the information in a detail-oriented analytical way and think in a more focused manner (p. 12). Such processing can have a hedonic value to their mood state, so that they imagine their future consumption to eventually become positive about a brand (Myers & Sal, 2013; Hill, 1988; Martin, 2003). This process is intense, as the individual wants to repair the negative mood through heuristic processing (Martin, 2003).
Based on the findings of past literature on mood state and brand preference in general, assumptions can be made regarding this relationship in the context of SNS. It is thought that also in the context of SNS, when the receiver processes a brand name in a particular mood state, they eventually will have a positive brand preference, as the receiver might go through similar thinking processes, as just mentioned.
This leads to the following hypotheses:
H4a: Individuals in a positive mood have a positive brand preference toward the brand mentioned in the sender’s message on the SNS.
H4b: Individuals in a negative mood have a positive brand preference toward the brand mentioned in the sender’s message on the SNS.
Chapter 5: Conceptual Model
Based on the literature review and the hypotheses that are established in the previous chapters, a conceptual model is developed to get a clear overview of the relationships that will be examined and of which the results will be discussed in the next chapters of this thesis. Social Comparison ---Social Contagion Tie-Strength
Chapter 6: Methodology
This chapter describes the used methods for this research. First, a description of the research design is given, with a focus on the questionnaires and pre-test. Also, the implementation of the experiment and the different variables and moderator are explained. Finally, it is clarified which methods of analysis of SPSS are used in this research.
6.1 Research Design
The purpose of this study is to examine the effects of positive messages on SNS on the mood state of the receiver. It is tested if tie-strength between the sender and the receiver has a moderating role in the effect of the mood state. Finally, it is examined whether the mood state may affect the brand preference towards the brand that is mentioned in the message.
To test the hypothesis, a questionnaire was designed, as experimental research is the most suitable method to test hypotheses (PAEC, 2013). Also, the questionnaire was designed in a way that the generated hypotheses could be tested and the results would give the information needed to either reject or accept the hypotheses. The variable tie-strength will be manipulated. Two versions of the questionnaire are developed, an in-group version (questionnaire 1) and an out-in-group version (questionnaire 2). Participants will receive either one of these questionnaires, without the knowledge of the existence of a second version of the questionnaire.
To determine which attributes should be included in this questionnaire, the existing literature is studied. It is of importance that the variables and items of this study are results of previous studies that are conducted in the context of marketing- and branding-activities, and if possible in the context of SNS, as this is a relative new area of research. SNS has become part of marketing- and branding-activities. The details of the questionnaire are explained in section 6.1.1 (p. 26).
First, a pre-test is done to check whether the questions of the survey are clear and comprehensible and to consider the reliability and validity of the questions and the variables. Also, the pre-test serves as an indication for the direction this experiment is heading. This test is done with 30 participants that made questionnaire 1. Further explanation and the results of this pre-test are given in section 6.2.4 (p. 32).
6.1.1 Questionnaire
Case
In the questionnaire, the presented case is about a posted status update on Facebook in the form of a check-in message (i.e. a person shows he or she is on a specific site), as it is possible to include a personal note and it includes the name of the place (Facebook, 2013). This function comes along with the presentation of a map underneath the message that shows the exact location of the place. Also, the name of the place, which functions in this study as the brand name, is a hyperlink so that the receiver can be directed to the Facebook-page of that place if he or she wants more information. Thus, this feature gives the receiver an idea of the brand.
In this study a reference is made to Facebook and the layout of the message is derived from Facebook, as this SNS is one of the largest and most well known SNS at the moment (Stenovec, 2013). The respondent of the questionnaire is the receiver of the message.
Message
The place that is mentioned in the case is a café & restaurant in Amsterdam, called De Ysbreeker. There are a few reasons why this place was chosen. First, this café & restaurant has an actual proper name so that it can also function as the brand name. Secondly, De Ysbreeker is open from 8.00 o’clock in the morning until 1.00 o’clock in the evening. On Fridays and Saturdays, this place is open until 2.00 o’clock in the evening (De Ysbreeker, 2013). Not only is it possible to drink some coffee or beer, but also to have breakfast, lunch or dinner there. It is possible to sit outside, as there is a terrace at the waterfront. The sender can be at De Ysbreeker on any moment of the day, also when the respondent is making the questionnaire, regardless what time it is or what the weather conditions are at that moment. Thirdly, this place exists since 1702 and it has just been renovated. It is a modern place, but it still has the classic and original elements. De Ysbreeker is a classic for the elderly and it is a trendy place for the young respondents. This information about De Ysbreeker, more or less just as described as in this paragraph, is given to the respondent in the questionnaire in case the respondent has not heard about De Ysbreeker before.
The personal note in this case is: ‘whooehooe having a blast!!’. The tone of this note is enthusiastic, thus positive, but the note itself is kept general, so it can be used in any situation while visiting this café & restaurant. It is unknown if the sender is with someone else the moment the message is posted. Thus, it only is clear that the sender is having a blast at the moment.
Questionnaires 1 & 2
Both questionnaires were similar to each other, except in the first version the participant had to have a close friend in mind while making the questionnaire. This thought was emphasized by naming the initials or nickname of that close friend at the beginning of the questionnaire. These initials or nickname were automatically processed in some of the questions of the survey, making the questionnaire personal for the respondent and making sure the respondent is constantly reminded by the name of the sender. In the second version the participant had to fill in a nickname or the initials of a distant acquaintance. Both questionnaires were randomly and equally distributed among the participants of the survey.
6.2 Conduct of Experiment
6.2.1 Participants
The participants of the pre-test and the main study were gathered from a random group of people. Preferably, the test had to be conducted by participants of 17 years and older. They can empathize in the situation, as they have the ability and resources to buy themselves a cup of coffee or a lunch. They are also in the position to visit a place like De Ysbreeker by themselves. In the end, all the participants were 17 years and older, but one. The people who received the e-mail with the link of the survey were allowed to forward this link to their friends and family. The only condition was that the respondents had to be at least familiar with the functioning of the SNS. If they were not familiar with it, they did not have to make the questionnaire. Half of the participants received the link of questionnaire 1 and the other half received the link of questionnaire 2. In the end, 180 participants filled in the questionnaires; after correction this number was 133. More information about the sample size is given in chapter 7.2 of this thesis (p. 36).
6.2.2 Procedure
The survey was made with the program Qualtrics, which is a web-based tool for building surveys (Qualtrics.com, 2013). Because it was not possible to activate both questionnaires at the same time, questionnaire 1 was first distributed to a random group of people. When the number of 75 participants was reached, questionnaire 2 was activated and send to another random group of people. If people of the first questionnaire had mistakenly received also this link, they were asked to ignore this e-mail.
Once the participants had opened the link, they first read the introduction page, which included a brief encounter, a thank you and the instructions of making this survey, the estimated time it will take to complete this survey and contact information in case there were ambiguities. On the next page, the participant was asked to either write down the nickname or initials of a close friend or a distant acquaintance in the white bar shown on that page. On the following page was the presentation of the case. In the message, the nickname or initials of the sender were processed, so that it actually looked as if their friend or acquaintance had posted the message.
After reading the case the participants were asked in what mood state they were now they had seen this message. They had to indicate in what degree they felt the four given positive mood states and the four given negative mood states. On the next page, the following questions measured the brand preference towards De Ysbreeker. This brand was not repeated in the question, as it was the intention that once a brand name was given in the message, this already could influence the brand preference. Repeating it could possibly influence the preference.
The following questions, on the next pages, were respectively to measure social comparison and social contagion, a reality check regarding the truthfulness of the case, and a reliability check regarding the relationship with the sender of the message. The closing questions were about some demographics. Please see Appendix 1 (p. 60) for the distributed surveys.
6.2.3 Measures
Independent variables • Positive message
Positivity is the key aspect of the SNS. Users share three times more likely a positive message than a negative message (Hutto, et al., 2013; Gruzd, et al., 2011). Also, positivity is greatly stimulated from the organisation of the SNS itself. To stay in the positive sphere of the SNS and of the users, only a positive message is used in this case.
Dependent variables • Mood states
Mood states either have a positive or a negative valence, which means that they are experienced as either good or bad (Sahney, 2012).
For this survey some emotions were chosen that could be generated after seeing such a message. There has not been a study before about the mood states that were generated after seeing other people’s personal messages on the SNS. Moods were chosen that were scientifically established as either positive or negative. In former studies, certain brands or other marketing stimuli caused these emotions (Sahney, 2012), and they are valid reactions consumers having on decision making and purchasing (Bagozzi, Gopinath & Nyer, 1999).
In this study it is examined if these emotions will also be evoked in the context of SNS, given indirectly a possible link between marketing stimuli and SNS. There are four positive and four negative emotions mentioned. Participants were asked to what extent they felt the following emotions: sad, happy, unhappy, depressed, joyful, warm, angry and amused. These emotions are all measured on a five point Likert scale (1 = Disagree, 5 = Agree).
Positive mood states
Happy is a positive mood state. It is suggested that a person is happy in a positive environment composed of a positive object or outcome (Tamir & Robinson, 2004; Bagozzi, Gopinath & Nyer, 1999). Consumers spent more time watching an advertisement with a positive content, which creates happy moods in the viewer (Hirschman & Stern, 1999).
Joyful is a positive emotion and this emotion is experienced once an outcome-desire fulfilment happens (Bagozzi, Gopinath & Nyer, 1999; Sahney, 2012). Joyful people are optimistic about consumption endeavours and enthusiastic about acquiring pleasurable experiences (Hirschman & Stern, 1999; Sahney, 2012).
Warmth is a positive emotion and a feeling of sentiment (Gardner, 1985). This mood state is positively related to liking for ads and purchase intentions (Xie, et al., 2004).
And finally, amused is a positive mood state and it means a person is occupied in a pleasing and entertaining way (Hirschman & Stern, 1999; Wiles & Cornwell, 1991).
Negative mood states
Sad is a negative mood state and people in this state experience emotions of hopelessness and dread (Hirschman & Stern, 1999; Sahney, 2012). People are pessimistic about consumption activities and they believe that things will never get any better than this (Hirschman & Stern, 1999).
Unhappy is the opposite happiness, but although a person feels unhappy, it still has the physical and mental energy to undertake action, unlike sad persons (Hirschman & Stern, 1999; Smallwood & O’Connor, 2011).
Depressed is a negative feeling (Gardner, 1985; Hirschman & Stern, 1999). Part of the group of consumers who are depressed are feeling this way, because the enormous propensity of the last decades have made them forgotten what the essentials are in life. These people are endlessly consuming more with less effort, in the hope to find these essentials (Voinea & Filip, 2011).
Anger is a negative mood state and people in the state of anger are likely to lose their temper, because they have failed to achieve a certain goal or if there is a case of injustice (Hirschman & Stern, 1999; Bagozzi, Gopinath & Nyer, 1999).
• Brand preference
Information about the brand De Ysbreeker is given in the questionnaire. This information is objective. If the respondent was unfamiliar with this place, this short description and a photograph of De Ysbreeker ought to make the case a bit more clear. Four questions regarding brand preference were asked. Three of these questions are in favour of this brand; the fourth question is not in favour of this
brand (i.e. ‘It does not meet my preferences of a café & restaurant’) (Hossain, 2006; Lee, 2000; Khan & Bozzo, 2012).
Control Variables
• Social Comparison
The questions regarding social comparison are derived from or based on the INCOM, the Iowa-Netherlands Comparison Orientation Measure, which is a scale for social comparison orientation created by Gibbons and Buunk (1999). The basis of this scale goes back to the original ideas of social comparison of Festinger (1954), founder of this theory.
• Social Contagion
One question regarding social contagion refers to recommendation system for individuals. This system is based on social contagion, because contagion is usually proportional to trust between individuals (Quijano, et al., 2009; Shang, et al., 2011). Closer friends an influence on each other, also through making recommendations (Quijano, et al., 2009; Shang, et al., 2011). Two other questions refer to sharing messages others have posted on the SNS (Appendix 1; questions 6.4 & 6.6). In case of
sharing, that participant actively participates and spreads the message that is posted by the sender, causing contagion (Durukan, et al., 2012). The fourth question is about copying other people’s behaviour on the SNS in general. This means that behaviour is copied, without the fact that anyone else is necessarily aware of it (i.e. no sharing) (Barash, 2011).
Both the questions to measure social comparison and social contagion were mixed and asked under one heading on a five point Likert scale (1 = Disagree, 5 = Agree).
Reliability check
To verify if the case presented in this thesis is a reflection of a realistic situation, some questions in this questionnaire function as a reliability check. If these questions are not properly answered, the analysis is not useful.
• Tie-strength
The strength of the relationship with the mentioned friend or acquaintance, the tie-strength, is checked through two questions (Appendix 1; 9.1 & 10.1) and measured
on a five point Likert scale (respectively 1 = Very weak, 5 = Very strong; 1 = Very upset, 5 = Very relieved).
• Brand name
The respondent is asked to recall the brand name (De Ysbreeker) that is mentioned in the example, in order to check if the respondent is focused while making the questionnaire and if such mention is recognized.
• Realistic Situation
The participants were asked if the situation described in the case was a realistic situation. Four questions are given. The first three questions are measured on a five point Likert scale (1 = Disagree; 5 = Agree). With the final question it was only possible to answer with a ‘yes’ or ‘no’.
Moderator
• In-group and out-group
In this study the label in-group and out-group is assumed to be the moderator. It is the question whether the effects on mood state of seeing a positive message on the SNS, is dependent on the relationship between the sender and the receiver. Two versions of the questionnaire are designed to measure the difference in effects, which are an in-group and an out-group version.
6.2.4 Pre-test
During the pre-test, it turned out a few questions had to be more extensive or different for the respondent in order to understand the situation or the question correctly.
To check in what direction the results of this experiment are heading, the means of the variables are measured. As all variables were measured on a five point Likert scale in the questionnaire, a mean of higher than 3 indicates that the participants overall agrees with the statements regarding these variables. One item measured on a two point Likert scale. This item is part of the variable realistic situation. In combination with the other items of this variable, which are measured on the 5 point Likert scale, the standard of this mean is set on 2.
The results (Table 1) indicate that the respondents get in a positive mood state when seeing the message (µ = 3.42) and not in a negative mood state (µ = 2.12), which should mean that the receiver copies the behaviour of the sender. However, the results
so far indicate that the receiver is more into social comparison than into social contagion (µcomp = 2.89; µcont = 1.89). Also, the respondents have a positive brand preference towards the brand mentioned in the case so far (µ = 3.52).
Variable Mean Standard Deviation
Positive MS 3.42 0.828
Negative MS 2.12 0.814
Social Comparison 2.89 0.827
Social Contagion 1.98 0.707
Brand Preference 3.52 0.804
Table 1: Means and standard deviations of variables
Control check pre-test
To determine if the respondents had made the questionnaire truthfully, some control questions were added. The participants were asked about their relationship with the sender of the message, to verify the tie-strength with the sender (i.e. strength relationship). The participants were also asked about the brand name that was mentioned in the case to check if they had read the case properly, but also to check if at least brand awareness was created with this single mention. Finally, it was checked if the participants thought the situation described in the case was a reflection of reality and if this is a type of message that the participant has seen before.
A one paired samples t-test was conducted to compare means. The control variables are significant supported on a 0.05 level (Appendix 2, section 2.3). The results indicate that the case itself, the situation described in the case and the questions regarding tie-strength (µ = 4.23, σ = .46; t(21) = 43.5, p < .05) , reality check (µ = 2.53, σ = .58; t (29) = 24.04, p < .05) and brand recognition (µ = .77, σ = .68; t(29) = 6.19, p < .05) are reliable and do not have to be changed.
6.3 Data Preparation & SPSS-tests
There are three datasets in this study. Next to the datasets of both questionnaire 1 and questionnaire 2 separately, there is also the third dataset that consists of the results of both questionnaires to get a clear overview of the data and to facilitate the testing of some hypotheses in case the results of both questionnaires are necessary. Also, new
variables are computed which contain the mean value of item scores concerning the same variable.
Cronbach’s Alpha
Before testing the hypotheses, other measurements have to be made with SPSS. First, the Cronbach’s Alphs needs to be calculated. With the Cronbach’s Alpha, the reliability of the variables is checked. According to Field (2005), when the score is α > 0.7, the score is internal consistent and therefore reliable. The score is not reliable if α < 0.6. However, this all depends on the number of items that are included in the calculation. With a scale of less than ten items, an Alpha of 0,5 is adequate (Field, 2005). In this case, most variables consist of four items and a score of α > 0.5 as adequate is used. See Chapter 7.1 (p. 36) for the results.
Control checks
For the actual test there are two control points. The first is about the question whether the situation described in the case is a realistic situation or not.
The strength of the relationship (i.e. tie-strength) between the sender and the receiver is also checked. To control the strength of the relationship between the participant and the mentioned friend or acquaintance, both questionnaires need to be checked separately. The strength of the relationship between a participant and its close friend has to be different from the relationship between a participant and the distant acquaintance. For questionnaire 1, the mean of the two measured items should be µ > 3, because than the receiver agrees with the statements and thinks the relationship is strong. The mean of both items of questionnaire 2 should be µ < 3, because the receiver does not agree with the given statements.
Both control checks are measured with a one pair sample t-test.
Hypotheses 1 & 2
Regarding the first two hypotheses, it is tested if people are in a positive or negative mood state after seeing the posted message. To test these hypotheses, a multiple regression analysis is conducted. The multiple regression analysis is used to study the relationship between a dependent variable and several independent variables (Hair,
2010). Results of this analysis indicate how well a set of independent variables is able to predict the dependent variable.
For the analysis of the negative mood states, the dependent variable is negative mood state (negative MS) and the independent variables are social comparison, social contagion, gender, age and education. The variables gender, age and education are the control variables. To analyze the positive mood states in the multiple regression analysis, the dependent variable is positive mood state (positive MS), and the independent variables are social comparison, social contagion, gender, age and education.
Hypothesis 3
The third hypothesis is about the moderating role of the label in-group/out-group and whether it has an effect on the mood state when seeing the receiver’s message. A multiple regression analysis is used to test the moderating effect of tie-strength in the relationship between the variables social comparison & social contagion and the mood states.
Before the multiple regression analysis, first the multicollinearity of the variables social comparison, social contagion and ingroupoutgroup has to be calculated. Ingroupoutgroup is a dichotomous variable, meaning that the data only consists of zero’s (out-group) and one’s (in-group). Multicollinearity is a term to indicate the presence of a linear relationship among independent variables in a linear regression (Silvey, 1969). A regression analysis has to be done to check the multicollinearity. If there is a problem with the multicollinearity, the variables have to be standardized. If the Variance Inflation Factors (VIF) are around 1, than there is no problem with multicollinearity (Mansfield & Helms, 1981). According to the results the variables do not have to be standardized, as all the VIF’s are around 1 (see Appendix 3; section 3.3, p. 71).
After the calculation of multicollinearity, a few interaction terms have to be made. An interaction characterizes the effect of the moderating variable (Cohen, et al., 2003). Social comparison has to be multiplied with ingroupoutgroup (comp x io) and social contagion also has to be multiplied with ingroupoutgroup (cont x io). All the variables are placed in the regression analysis in two steps. In the first step, only the dependent variable negative MS and the variables social comparison and social contagion, which are in this test the independent variables, are filled in. In the second step the variables
ingroupoutgroup, comp x io and cont x io are the independent variables. The same procedure is done with positive MS as the dependent variable.
Hypotheses 4a & 4b
The fourth set of hypotheses is about the effects of mood state on preference towards the brand mentioned in the posted message. To test these hypotheses, a multiple regression analysis is done. For the analysis of the brand preference and negative mood state, the dependent variable is brand preference and the independent variables are gender, age, education, brand name (i.e. the question whether the participant can remember the brand name mentioned in the message) and negative MS. To analyse the brand preference and positive mood state, the dependent variable is brand preference and the independent variables are gender, age, education, brand name and positive MS.
Chapter 7: Results
An online experimental survey is developed to test the hypotheses. In this chapter the main results of this survey are presented. The reliability of the data, the sample profile and the results of the statistical tests are presented. These results indicate if the hypotheses are significantly supported or not.
7.1 Cronbach’s Alpha
According to the measured Cronbach’s Alpha, all the items are internal consistent, of which two variables, social comparison (α = 0.502) and brand preference (α = 0.560), are sufficiently reliable. However, a check has been done to see if the Cronbach’s Alpha of these two variables will be larger when an item of these variables is deleted. When deleting an item of the variable social comparison, there is a Cronbach's Alpha of 0.614, which makes the variable more internal consistent (see Appendix 1; item 6.8, p. 60). This is not the case with brand preference. If an item of this variable is deleted, the Cronbach's Alpha will not get any higher. Check Table 2 for the Cronbach’s Alpha’s.
Variable Cronbach’s Alpha
Social comparison 0.614 Social contagion 0.661 Positive MS 0.888 Negative MS 0.823 Brand Preference 0.560 Realistic situation 0.899 Strength relationship 0.832
Table 2: Variables and Cronbach’s Alpha
7.2 Sample profile
In the end, 180 questionnaires were filled in. However, not all the participants had finished the survey (more than 30 % was missing) and these results were removed from the dataset. After correction, n1 (in-group) was 68 and n2 (out-group) was 65. The sample consisted of 66 % of women and 34 % of men, ranging in age from 14 to 81, with the majority in the ages 14 to 59 years. A majority of the participants has been graduated from the University (81,1 %).
7.3 Control check
The variables are measured on a five point Likert scale, except one. The respondent had to make an indication in what degree they agreed or disagreed with the statements and questions given in the questionnaire (1 = Disagree; 5 = Agree). As a result of this measurement, a mean of the variables µ > 3 indicate that the participant is familiar and agrees with the statements. One question was a ‘yes or no’ question (1 = yes; 0 = no), which means that in this case the mean of this item has to be μyes > 0.5 in order to be
valuable. For both control checks a one sample t-test is conducted. Please see Appendix 2 for the SPSS results of these measurements (p. 65).
26% 49% 19% 6%
Age
14-24 25-39 40-59 60 and older 2% 5% 2% 9% 29% 45% 8%Education
Primary education High school graduation Vocational education (MBO) College graduation (HBO) Bachelor Master Postgraduate 34% 66%Gender
Male Female7.3.1 Realistic situation
The results of the one sample t-test indicate that the items for realistic situation (µ = 4.26, σ = .79) are significantly supported (t (128) = 61.624, p < .05). According to the results, there is strong evidence that these items do measure the reality of the situation.
7.3.2 Strength relationship
The results of the one sample t-test indicate that the items to measure the strength of the relationship (µ = 3,55, σ = .79) are significantly supported (t (126) = 35.2, p < .05). According to the results, there is strong evidence that the items of this variable do measure the strength of the relationship between the sender and the receiver.
7.4 Hypotheses testing
In this section the results of the hypotheses are presented. The data is analysed using IBM SPSS Statistics 21.0. The regression analysis is conducted and the relationships are considered statistically significant at a level of 0.95 confidence (p < 0.05). The results of the SPSS tests are shown in the Appendix 3 (p. 69).
7.4.1 Hypotheses 1 & 2
H1: A positive message on the SNS will trigger negative feelings at the receiver, as he or she is going through the route of social comparison.
H2: A positive message posted on the SNS will trigger positive feelings at the receiver, as he or she is going through the route of social contagion.
It was found that 8 % of the variance is explained by the model (R² = .08, F (5,117) = 2.025, p > .05). The results of the multiple regression analysis indicate that social comparison does not significantly predict the negative mood state (β = .073, p > .05). These results indicate two facts: the receiver does not significantly get into a negative mood state when seeing a positive message on the SNS and the receiver does not significantly compare itself with the sender. Thus, Hypothesis 1 is not accepted.
It was found that 5.7 % of variance is explained by the model (R² = .057, F (5,116) = 1.391, p > .05). The results of the regression, with positive mood state as the dependent variable, indicate that social cognition does not significantly predict this
mood state (β = -.010, p > .05). These results indicate that the receiver does not significantly get into a positive mood state when seeing a positive message on the SNS and the receiver does not significantly copy or adopt the behaviour of the sender. Therefore, hypothesis 2 is not accepted.
7.4.2 Hypothesis 3
H3: Tie-strength has a moderating role in the effect of the receiver’s mood state. First, the results with negative mood state as dependent variable are checked. The interaction terms comp x io (β = -0.087, p > .05) and cont x io (β = -0.093, p > .05) do both not significantly predict a difference in effect size of the mood state. According to the results, tie-strength does not significantly influence the mood states of the receiver that are experienced through the routes of social comparison and social contagion.
Secondly, the results with positive mood state as dependent variable are checked. The interaction term cont x io (β = .429, p > .05) does not significantly predict the effect size of the mood state. However, the interaction term comp x io (β = - .770, p < .05) does significantly predict the effect size of the mood state. These results indicate that tie-strength does not significantly influence the effect size of positive mood state when the receiver is going through the route of social contagion. There is significant difference in the effect size of positive mood state when the receiver is going through the route of social comparison. Therefore, hypothesis 3 is partially accepted.
7.4.3: Hypotheses 4a & 4b
H4a: Individuals in a positive mood have a positive brand preference toward the brand mentioned in the sender’s message on the SNS.
H4b: Individuals in a negative mood have a positive brand preference toward the brand mentioned in the sender’s message on the SNS.
The results of the regression indicate that the receiver’s positive mood state does not significantly predict a positive brand preference (β = .05, p > .05). Also, 4,5 % of the variance is explained by the model (R² = .045, F (6,115) = 1.084, p > .05). These findings
