Is Sharing Caring?

Is Sharing Caring?

- A Quantitative Study of Consumers Intention

to Write Online Product Reviews.

Bachelor Thesis

Tutor: Viktor Magnusson Term: Spring 2019 Course: Independent

Degree Project of the Marketing Programme With Integrated Method

Acknowledgment

The authors of this paper would particularly like to thank a few involved people who have helped and contributed to the completion of this thesis.

First, we would like to direct our appreciation to our tutor Viktor Magnusson for his support during the whole process. Thank you for your confidence in us and for your advice. You have

contributed crucial material for us in order to complete this thesis.

In supplement to this, we would like to thank our examiner Åsa Devine for the valuable guidance we received during the seminars.

Likewise, we would like to present our gratefulness to Setayesh Sattari who have contributed with appreciated and valuable knowledge regarding the conduction of this quantitative

research.

Finally, we would like to thank all the respondents who participated in the questionnaire for this thesis. Without your valuable contribution, we would not have been able to complete this

thesis.

Simon Huléen Vincent Lind Philip Warntoft

Abstract

Background

increasing relevance and importance when it comes to consumers purchase decisions. Online reviews are considered to be the most important and reliable form of eWOM and it refers to when consumers communicate with other consumers online regarding their previous

shopping experiences. In order to increase the generalizability and understanding of motives that influence consumers intention to write online reviews, authors of previous studies have suggested that future research should intend to focus on motives that influence intention to write online reviews in other contexts that are not aimed towards the service industry.

Purpose

consumers intention to write online product reviews.

Methodology

research design were used in order to test the model founded on two major hypotheses. The data was gathered through a self-completed questionnaire with a sample of 222 respondents.

Findings

personal motives and social motives influence consumers intention to write online product reviews. After analysing the data that emerged from the survey, a conclusion can be drawn that the hypothesis concerning personal motives and its influence on consumers intention to write online product reviews was accepted whereas the hypothesis concerning social motives was rejected.

Conclusion

suggestions found in this study with a more diverse sample in order to increase the

generalizability of the theoretical implications. Furthermore, directions for future research will also recommend testing the modified model that emerged in this paper in order to detect if there are any additions that need to be added to increase the predictability of the findings.

Keywords

Table of Content

3. Conceptual Framework 9

3.1 Personal Motives 9

3.2 Social Motives 10

3.3 Proposed Research Model 11

4. Methodology 12 4.1 Research Approach 12 4.1.1 Deductive Approach 13 4.1.2 Quantitative Research 13 4.2 Research Design 14 4.3 Data Sources 14

4.4 Data Collection Method 15

4.5 Data Collection Instrument 15

4.5.1 Operationalization and Measurement of Variables 16 4.5.2 Questionnaire Design 20 4.5.3 Implementation of Questionnaire and Data Collection 22 4.5.4 Pretesting 23

4.6 Sampling 24

4.6.1 Sample Selection and Data Collection Procedure 25

4.7 Data Analysis Method 25

4.8.4 Reliability 29 4.8.5 Replication 30

4.9 Ethical and Societal Considerations 30

5. Results 32

5.1 Control Questions and Descriptives 32

5.2 Reliability Analysis 35

5.3 Correlation Analysis 35

5.4 Hypothesis Testing 36

5.4.1 ANOVA Table 38

6. Discussion 39

6.1 Discussion About Personal Motives 39

6.2 Discussion About Social Motives 40

6.3 Discussion About Previous Online Review Behavior 41

6.4 Modified Model 42

7. Conclusion 43

7.1 Theoretical Implications 43

7.2 Managerial Implications 44

8. Limitations and Future Research 45

8.1 Limitations 45

8.2 For Future Research 46

Reference List

Appendix 1 Categorization of Artciles in the Literature Review Appendix 2 Selection Process of Articles

1

1. Introduction

In the introduction chapter, this study will introduce and explain the object of this investigation. Furthermore, a problematization regarding the certain object will be conducted and the purpose of this study will be stated in the final section of this chapter.

1.1 Background

With the rapid development of technology and social media, electronic word of mouth (eWOM) has become more relevant than ever. The rapid advancements in technology have made it easier than ever for people to share their thoughts and opinions with other people through social networking. As a result of this, entire websites dedicated to eWOM, such as TripAdvisor, are receiving large amounts of traffic from consumers seeking assistance in their purchase decision process (Chen & Law 2016). eWOM is described as a recent

extension of the well established concept of Word of Mouth (WOM). Both concepts revolve around spreading information through verbal communication, however, the key difference between the two concepts is that eWOM refers to WOM activities communicated through the internet (Anubhav & Satish 2016). In short, eWOM communication refers to statements shared through online platforms and includes both positive and negative statements (Hennig-Thurau et.al. 2004). When it comes to the consumer's expression of eWOM, it can be

communicated in several different ways, such as through comments, online feedback, and posts. However, one of the most common and most trusted forms of eWOM communication is online reviews (Anubhav & Satish 2016).

Online reviews, is described as communication of a consumer's consumption

experience, which is often in combination with a rating system. Consumers can utilize online reviews in their post-purchase process in order to share their experiences with other

2

by individuals with a certain level of expertise of the product (Gonçalves, Silva & Martins 2018). Through reading particular online reviews properly, consumers can receive a more profound insight concerning a certain product, which in turn can come to have a significant impact when it comes to making their final purchase decision (Thakur 2018).

Through reviewing previous research on the motives that influence consumers intention of writing online reviews. The authors of this study found that numerous of the articles focus on influential motives in a service industry context (See Appendix 1a). Only 4 of the 18 articles presented in Appendix 1a were focused on motives influencing consumers intention of writing online product reviews.

1.2 Problem Discussion

As described in the previous chapter, most of the previous studies within the field of which motives influence consumers intention of writing online reviews have been conducted within the context of the service industry (See Appendix 1a). Authors such as Labsomboonsiri, Mathews and Luck and Xiang et. al., argues that it is difficult to generalize the antecedents of online reviews since most of the motives have only been tested in the scope of the service industry. Thus, the direction for future research is aimed towards the application of the identified antecedents in industries other than the service industry in order to achieve a greater level of generalization for the identified motives influencing consumers intention of writing online reviews (Labsomboonsiri, Mathews & Luck 2017)(Xiang et al. 2018).

3

has found that the nature of the review, i.e. positive or negative, only matters for weak brands, whereas for strong brands it does not necessarily influence consumers purchase behavior. Due to these findings, it becomes important for managers to understand which motives influence consumers intention of writing online product reviews. Through understanding which motives influence consumers intention of writing online product reviews, managers can seek to enhance these motives and thus achieve an edge over competitors (Ho-Dac, Carson & Moore 2013).

Through reviewing the literature within the field of motives influencing consumers intention of writing online reviews, this study was able to identify several different predictors that had been found to have a significant influence on consumers intention of writing online reviews. After reviewing the literature, a total of 44 different antecedents that had all been found to influence consumers intention of writing online reviews (See Appendix 1, Table 1a). Since most of the motives that had been identified had only found to be significant in one, or two studies, the authors choose to focus on the more central and well studied motives that had been found to be significant predictors in at least three or more studies. Through applying this criterion to the list of articles, this study can have the ability to narrow down the number of antecedents to seven different motives, which were, Economic incentives, Venting, Attitude,

Subjective Norm, Helping others, Concern for others and Sense of community. Based on the

nature of these concepts, they could be grouped into either personal or social motives (See Appendix 1, Table 1b). Thus, in line with Labsomboonsiri, Matthews and Luck’s and Xiang et. al’s suggestion, an interesting contribution for this study would be to test these motives which have been identified to influence consumers intention to write online reviews, within the context of online product reviews, thus achieving a greater level of generalization for the motives (Labsomboonsiri, Mathews & Luck 2017)(Xiang et al. 2018).

1.3 Purpose

4

2. Literature Review

This chapter will provide a framework concerning the theoretical concepts that are applied to this study. This chapter will be introduced with theoretical information regarding

behavioral intention. Subsequently, this study will explain personal motives when it comes to writing online reviews including the concepts economic incentives, venting, and attitude. Afterwards, social motives for writing online reviews will be clarified including the concepts subjective norm, helping others, concern for others and sense of community.

2.1 Behavioral Intention

Fishbein and Ajzen describe behavioral intention as indications of an individual's readiness of engaging in a certain type of behaviors (Fishbein & Ajzen 2010). Behavioral intention is a measure that can be applied to virtually any type of behavior in order to predict how likely people are to perform said behavior (Bock et al. 2005). Furthermore, when individuals portray a higher sense of intention to perform a certain behavior, an increase in actual compliance can be seen (Ajzen, Czasch, & Flood 2009).

Fishbein and Ajzen go on to describe that behavioral intention can be expressed by statements such as “I intend to perform the behavior” or “I expect to perform the behavior”. Meaning that behavioral intention at its core is measured based on a person's conscious estimation of their own likeliness of performing said behavior. Thus, the higher the

individual's subjective perception is, the higher the predictability of the construct becomes (Fishbein & Ajzen 2010).

2.2 Personal Motives

5 2.2.1 Economic Incentives

Economic incentives basically refer to any monetary benefit that consumers can receive from writing an online review, such as discount codes, reward points, coupons, etc. Meaning that consumers who contribute by writing online reviews are reimbursed in some manner for writing the online review (Hussain et al. 2018). In the study conducted by Kumar and Purbey, it was found that economic incentives are a major motive for why consumers determine to get involved and spread their opinions through writing online reviews. Suggesting that

consumers choose to write online reviews because of the reward they receive from doing so (Kumar & Purbey 2018). This is further corroborated by Hennig-Thurau et al. which also found that economic incentives are one of the key motives influencing consumers intention of writing online product reviews (Hennig-Thurau et.al. 2004).

2.2.2 Venting

The desire for consumers to ventilate their emotions following a consumption experience has been found to be one of the primary predictors influencing consumers intention of writing online reviews. Venting, in an eWOM context, refers to consumers intention of writing online reviews as the result of strong emotional motives, such as frustration, anger or perhaps

disappointment (Verhagen, Nauta & Feldberg 2013). Such motives are predominantly present when consumers are sharing their experience with strangers. Following a negative

consumption experience, consumers have been found to be more prone to write online

reviews in order to vent their frustration to strangers (Lopez-Lopez, Ruiz-de-Maya & Warlop 2014). In addition to this, Wetzer, Zeelenberg, and Pieters found that emotional motives, such as venting are primarily related to the communication of negative eWOM, meaning that consumers are more prone to write negative online reviews as a result of emotional reasons from negative consumption experiences (Wetzer, Zeelenberg & Pieters 2007). When the sender perceives the receiver as a stranger, emotional motives tend to take on a more

6 2.2.3 Attitude

Consumers attitude is another concept that has been found to play a significant role in regards to influencing consumers intention of writing online reviews. Consumers attitude relates to the general outcome of goodness or badness associated with writing an online review. In other words, it can be interpreted as how the consumer adopts the behavior of writing online reviews (Yusuf, Che Hussin & Busalim 2018) This statement is supported by Zainal Harun and Lily, who also found that attitude positively affects consumers intention of writing online reviews (Zainal, Harun & Lily 2017).

Individuals attitudes can be denominated into different terms, which can serve as strong motivators for consumers inclination towards writing online reviews. The concept of attitude revolves around cognitive-, and affective attitude (Aghakhani, Karimi & Salehan 2018). Cognitive attitude refers to the extent that individuals tend to advance their beliefs that is related to a certain object. Affective attitude refers to the degree of excitement that is perceived when an individual participates in a certain type of behavior, such as writing an online review (Zainal, Harun & Lily 2017). When it comes to cognitive attitude, it is

perceived to have a higher influence when consumers are contributing to online reviews that are exposed through a written text. On the other hand, affective attitude is found to have a higher influence when it comes to adoption of online reviews through rating systems

(Aghakhani, Karimi & Salehan 2018). In addition to this, Liang et al., found that consumers commitment to write online reviews largely depends on consumers general attitude towards communicating their experiences online, meaning that it largely depends on how consumers adjust to writing online reviews (Liang et al. 2013).

2.3 Social Motives

Previous studies have found that social motives have the ability to positively influence

7

(Hennig-Thurau et al. 2004). Finally, individuals desire to feel a sense of belonging, or sense of community is yet another factor that greatly influences consumers intention of writing online reviews due to how consumers feel connected with one another (Balaji, Khong & Chong 2016).

2.3.1 Subjective Norm

Subjective norm is defined as a process by which individuals within a group are influenced by other members and through social pressure their beliefs and behavior is influenced. This process has been found to influence consumers intention of writing online reviews (Sijoria, Mukherjee & Datta 2018). Where Fu, Ju, and Hsu found it to be one of the primary predictors of consumers intention of writing online reviews, stating that is has a positive impact on intention of communication both positive and negative eWOM (Fu, Ju & Hsu 2015). In addition to this, Liang et al. describe that subjective norm is such a strong predictor of the intention of writing online reviews since it possesses the ability to create a ripple effect. Liang et al. describe this as individuals becoming more prone to communicate eWOM if they notice their friends doing so, suggesting that these types of individuals intention of

performing a behavior is influenced by the subjective norm revolving said behavior (Liang et al. 2013).

2.3.2 Helping Others

Another key motive that influences consumers to write online reviews revolves around consumers inclination of helping others (Verhagen, Nauta & Feldberg 2013). Consumers can become more motivated to share their consumption experience founded on how it can help other consumers to make purchase decisions (Labsomboonsiri, Mathews & Luck 2017). This basically means that consumers choose to write online reviews in order to help the

community they belong to (Verhagen, Nauta & Feldberg 2013). Helping others is argued by Yang to be a significant reason for why individuals want to share their opinion in an online context (Yang 2017). This is corroborated by Cheung and Lee who continues to argue that individuals can also find enjoyment in helping others by sharing their experiences on

8 2.3.3 Concern for Others

Concern for other consumers is considered to be an essential determinant for writing online reviews. The concept refers to how consumers write online reviews since it contributes to decent opportunities for other consumers to make good decisions before purchasing (Xiang et al. 2018). The concepts have been found to have a high level of significance on the influence of consumers intention of writing online reviews (Hennig-Thurau et al. 2004). As social concern to be a significant influential factor to consumers intention to write online reviews is something that is supported in the study conducted by Gonçalves, Silva & Martins. It has also been stated that concern for other consumers is something that is mostly valued among consumers when it comes to the contribution of writing online reviews (Gonçalves, Silva & Martins 2018). If an individual has experienced something stated as uncertain or

disappointing is the individual more likely to have a concern for others and warn them about these undesirable events. The reason for this is described as this individual do not want others to experience the same negative consumption experience as they have. Preventing others from experiencing these negative feelings is stimulating for these individuals (Wetzer, Zeelenberg & Pieters 2007).

2.3.4 Sense of Community

When individuals are interacting and discussing common interests with others, a sense of community can be experienced. This is described to have a positive influence on individuals reasons to share opinions online. Meaning that if consumers feel a sense of community, they become more inclined to sharing their opinion (Hennig-Thurau et al. 2004). This is supported by Cheung and Lee and Balaji, Khong and Chong, stating that sense of belonging and

community has been found to increase consumers intention of writing online reviews

(Cheung & Lee 2012)(Balaji, Khong & Chong 2016). Thus, the sense of belonging motivates consumers to write online reviews and share their experience in order to assist their fellow community members. If consumers feel a strong sense of community towards a particular website, consumers become more inclined to communicate their experience with other community members (Cheung & Lee 2012). Hennig-Thurau et al. also found that when personal information is shared throughout social networking sites, individuals tend to

9

3. Conceptual Framework

The conceptual framework for this paper is founded on the premises of personal-, and social motives influence on behavioral intention. The hypotheses founded on personal and social motives will be presented as well as the proposed research model for this paper.

3.1 Personal Motives

The first category of motives encouraging consumers to write online product reviews revolves around their own personal motives. In the study conducted by Lopez-Lopez, Ruiz-de-Maya, and Warlop, personal motives revolve around motives centered around the consumers own interest. For instance, if a consumer has recently experienced a poor

consumption experience of a product, then the consumer may decide to ventilate the negative consumption experience to other through writing an online product review (Lopez-Lopez, Ruiz-de-Maya & Warlop 2014). Previous studies have also found that economic incentives have the potential of increasing consumers intention to write online product reviews through receiving economic rewards, i.e. a discount code, from publishing a review (Kumar & Purbey 2018). Another factor related to personal motives influence on behavioral intention of writing online product reviews relates to the consumer's personal attitude, where both cognitive and affective attitudes were found to increase consumers intention of writing online product reviews (Yusuf, Che Hussin, & Busalim 2018).

Since all of the motives discussed in the previous paragraph has been found to

motivate consumers to write online reviews for services, such as for hotels or restaurants, the authors of this thesis aim to investigate whether these motives also have an influence on consumers intention of writing online product reviews. Thus, in relation to the behavioral intention, the hypotheses H1 was developed which describes the positive influence personal motives has on consumers intention of writing online product reviews.

H1: Personal motives positively influence consumers intention of writing online product

10 3.2 Social Motives

The second category of motives that affect consumers intention to write online product reviews refers to social motives. Previous studies have found that social motives affect consumers intention of writing online product reviews within the context of the service industry. Social motives are described in this paper as motives that include reasons of doing something for others, such as when consumers intention of writing online reviews is

influenced by factors such as subjective norm (Liang et al. 2013). It has been found that consumers are primarily motivated to write online reviews based on social motives. In addition to this, consumers have also been found to write online product reviews because they are concerned that other consumers make poor purchase decisions (Xiang et al. 2018). Other social motives that have been found to influence consumers intention of writing online product reviews are the subjective norm, also known as social pressure. Meaning that

consumers write online product reviews because they feel that it is expected of them as consumers to contribute by writing online product reviews (Liang et al. 2013). Finally, the sense of community has also been found to be a factor that motivates consumers to write online product reviews. Suggesting that consumers might feel a sense of community where consumers feel connected with each other, and thus write online product reviews to interact and socialize with each other (Cheung & Lee 2015).

Thus, as described in chapter 2.3 Social motives, in regards to the service industry context, social motives have been found to influence consumers intention of writing online product reviews through appealing to consumers inclination to perform a certain behavior for reasons other than one's own personal benefit. Which in relation to the behavioral intention leads to how consumers would write online product reviews in order to benefit others. Thus this study proposes the hypothesis H2.

11 3.3 Proposed Research Model

Based on the proposed hypotheses, this study has developed the conceptual research model presented in figure 3. Which aims to visualize the hypothesised relationship between the independent variables with the dependent variable.

12

4. Methodology

The methodology is the chapter where this study discusses the research approach and design in order to collect the data for this study. Thereafter, a discussion regarding data sources and the data collection method will take place. Furthermore, an operationalisation table is included in this chapter where this study exhibits the questions included in the questionnaire followed by the design of a questionnaire, pretest and sample selection. Thereafter, the method chapter continues with a discussion concerning the methods used in order to analyse the data that emerged from the survey. Subsequently, quality criteria will be conducted where the authors discuss the importance of validity, reliability, and replication when it comes to the collection of data. Furthermore, ethical and societal principles will be included as well since this is a central part to consider when it comes to data collection.

4.1 Research Approach

The first steps of any type of marketing research start off with formulating a clear definition of the research problem. According to Malhotra, the problem formulation should consist of problematization from both a managerial perspective, as well as from a theoretical one. Meaning that the research should have a clear understanding of what its managerial implication can be, as well as how the study further advances the field from a theoretical perspective. Once the problematized area has been defined, the researcher should develop an approach to the problem. The research approach should be developed in regards to the problematized area, using the potential implications of the research as a guide for how to approach the research. Another important step in formulating the research approach is reviewing the existing literature within the field (Malhotra 2010). Bryman and Bell describe that the most important reason for conducting a thorough literature review is for the

researcher to gain a deeper understanding of which areas within the field that already have been studied. Meaning that the aim of the literature review is for the research to make sure that their research is worthwhile, and not a repetition of previous studies (Bryman & Bell 2015).

13

of the service industry. From this, the theoretical implication of testing the influencing motives within the scope of online product reviews was reached.

4.1.1 Deductive Approach

As described in the previous chapter, the defined problematization of the research stems from a strong theoretical foundation. In regards to the managerial and theoretical implications, the idea to test these findings within another context was adopted. Thus, the approach to the research stems from testing the findings of previous research, which suggests that the

research approach is of a deductive nature. A deductive approach is associated with a certain conceptual and theoretical structure, which is further developed and tested through empirical observations. A deductive approach is denoted to move from a general perspective to a more specific perspective. An example of this can be that there is already existing theories which explains certain factors that motivate consumers to make actions in different ways.

Researcher's intention is then to test these theories in another context to determine if these theories can be confirmed or rejected on that specific occasion. In order to make this possible, the researcher then has to gather certain significant data of all the variables that the particular theories have determined as being important (Collis & Hussey 2014). When it comes to a research strategy in a deductive approach, a study is primarily founded around existing theories. Based on these theories, hypotheses are formed and the intention with these hypotheses is to test them and determine if they are confirmed or rejected. It is further

described that a deductive approach is a relationship between research and theory. It refers to a certain approach that includes ideas and hypotheses in the research that are founded on already existing theories (Bryman & Bell 2015).

4.1.2 Quantitative Research

14

reason when using the quantitative method as a research strategy is to get an understanding regarding specific attitudes, behaviors, opinions and other defined variables from a larger sample of a certain population (Bryman & Bell 2015)

4.2 Research Design

Research design can be defined as a framework intended for the collection and analysis of data in order to provide coherent justification regarding method and particular analysis techniques (Saunders, Lewis & Thornhill 2016). The intention of a research design is that it should guide the study towards its objectives. The reason for this simply depends on that the research approach explains for the researcher how the data to gather will be acquired (Aaker, Kumar, Day & Leone 2011). Suggesting that depending on the aim of the study, the research design will vary. Since the aim of this study is to investigate if personal- and social factors influence consumers intention of writing online product reviews, an explanatory (casual) research design is suitable. Explanatory research is used in order to investigate the cause-and-effect relationship of variables, i.e. an explanation of how a change in one variable influences another variable (Malhotra 2010). Considering that this study takes on an explanatory nature for its purpose, a cross-sectional research design becomes a suitable design for the study to adopt. When using a cross-sectional research design, the researchers' intention is to search for variation among different variables (Saunders, Lewis & Thornhill 2016). Malhotra goes on to argue for the wide range of applicability of the cross-sectional design in marketing research. The author describes the design as a way of collecting information from a sample of the population only once, in order to analyse the variance in the answers (Malhotra 2010). Meaning that the researchers' aim is to search for variable variance in the responses of the sample. For a variation to occur, it is necessary to investigate several cases simultaneously. Through the investigation of several cases simultaneously, the researcher can identify distinctions between different variables and the influence these variables can have on each other (Saunders, Lewis & Thornhill 2016).

4.3 Data Sources

This research applies a cross-sectional research design, the empirical data will be gathered from surveying a sample of the population, meaning that the data will be gathered

15

researchers in another particular context (Malhotra 2010). The major differences between these two data sets are that primary data is collected through a direct connection to the stated purpose of the intendant study, while secondary data have individual purposes that further can be used in other studies (Bryman & Bell 2015). However, primary data is required in order to have the ability to measure the stated purpose of the study and the reason for this is that the nature of the purposes is not exactly the same (Malhotra 2010).

4.4 Data Collection Method

The method used for collecting the empirical data which is to be analysed is an important decision for any type of research. Data collection can be defined as the process of gathering information for a certain study and the data can be gathered through several different approaches depending on what sort of study that intends to be conducted (Bryman & Bell 2015). Since this study applies a cross-sectional design, the data should be gathered directly from the respondents (Malhotra 2010). In addition to this, the study also takes on a

quantitative approach, which suggests that the data must be quantifiable, suggesting that a large amount of responses is required in order to generalize it (Bryman & Bell 2010). In order to efficiently gather quantifiable data directly from the respondents, a self-completed

questionnaire has been chosen as the method for collecting the data. Malhotra defines a questionnaire as a way of gathering data from respondents through using a formalized set of questions. The key objective of a questionnaire is to translate theories or information into the form of a question so that the respondent is able to provide an answer. Thus, through

responding to the questions the researchers are provided with information about the constructs the questions have been formulated to measure. A self-completed questionnaire simply means that the researcher takes no active role in the respondents answering process, thus reducing any influence researchers might have on the respondents (Malhotra 2010).

4.5 Data Collection Instrument

The gathering of data with the help of a questionnaire can be made in various different ways. Malhotra explains that the distribution of a questionnaire can be made through telephone, personal meetings with the respondents, by mail or electronically through the Internet (Malhotra 2010). Bryman and Bell further argue that a self-completed questionnaire is specifically beneficial for the researcher when the selected sample is geographically

16

also in order to reach out to a large number of respondents in a comprehensible way (Bryman & Bell 2015). In accordance with Malhotra, this study will partly use online platforms such as Facebook but also personal meetings with the respondents when distributing the

questionnaire. Taking the arguments concerning time and cost saving in to account, which has been claimed by Bryman and Bell (2015), personal meetings and the electronic

distribution of the questionnaire will be considered as a reasonable approach when it comes to the distribution of the questionnaire in this study.

4.5.1 Operationalization and Measurement of Variables

An operationalization can be defined as the process of explaining how concepts are about to be measured in a particular study. Researchers convert investigated theories to measurable variables in order to see if these theories have a relation to each other or not. This is described as an operational definition. In addition to this, it is described that the operationalization needs to have a number of indicators stated. The indicator’s part in an operationalization could be stated as questions or statements that should provide an understanding of how a concept is about to be measured. Meaning that operationalization is a process of breaking theories down into individual items and questions which can be understood by the

respondents (Bryman & Bell 2015).

Interval and nominal are two certain types of variables which have been applied to this questionnaire. Interval variables investigate the relationship between different variables (Bryman & Bell 2015). What is further described by Malhotra is that interval variables can be put into scales. The distance between the interval variables in a scale should be of the same difference, explaining that there is just as a significant difference between variable one and variable two as for variable four and five (Malhotra 2010). Bryman and Bell state that the name of these scales are likert scales and they are added in order to measure the interval variables connected to a statement. Likert scales can be further described as a tool that measures perceptions about something in an interval context. Nominal variables can be described as a categorization of data, arguing that nominal variables should not rank answers in comparison to others (Bryman & Bell 2015). Nominal variables could be used in order to identify what type of respondents that are participating in an investigation (Malhotra 2010).

17

The nominal variables have been used in order to find out respondents habits with creating product reviews, age, occupation, gender, and online purchasing habits. These questions are not included in the stated operationalization.

Theory Nr Indicator Item Measurement Definition Question

18 (Aghakhani, Karimi & Salehan 2018). Personal Motives 4 Affective Attitude Pers 4 Likert scale 1-5 Affective attitude refers to the degree of excitement that is perceived when an individual participates in a certain type of behavior, such as writing an online review (Aghakhani, Karimi & Salehan 2018). I think it is exciting to talk about a product with others. Social Motives 5 Subjective Norm

Soc 1 Likert scale 1-5 Subjective norm refers to the social pressure that is experienced when performing or not performing a certain behavior (Liang et. al. 2013). Others people's expectations influence my behavior. Social Motives 6 Helping Others

Soc 2 Likert scale 1-5

Helping others refers to when consumers write online reviews in order to help the community they belong to (Verhagen, Nauta, & Feldberg 2013). I enjoy helping other people to make better purchase decisions. Social Motives 7 Concern for Others

Soc 3 Likert scale 1-5

19 make a better purchasing decision (Xiang et. al. 2018). Social Motives 8 Sense of Community

20 estimation of their own likeness and willingness of performing said behavior ( Fishbein & Ajzen. 2010). Table 4.5.1 Operationalization Table

4.5.2 Questionnaire Design

When it comes to the design of a questionnaire, it is important to take the formulation of the questions into consideration since the question should be formulated in such a way that they can assist in solving the purpose of the study. This is partly of importance since it helps to prevent non-relevant questions from being included in the study. In order to facilitate the construction of a questionnaire, the researcher should first and foremost determine the purpose for why the research needs to be conducted, in this case in relation to what problem you want to solve through the application of the questionnaire (Bryman & Bell 2015). Determining the purpose of the questionnaire is something that Saunders, Lewis, and Thornhill also state as an important step when it comes to designing the questionnaire. However, it is also of great importance to explain the purpose of the survey to the respondents so that they have a clear and profound understanding regarding what the questionnaire is investigating. The researcher should, therefore, consider conducting a covering letter for the questionnaire which describes the purpose of the study to the

respondents since this is the section of the questionnaire the respondents will look at first. In addition to this, it is essential to also explain to the respondents why their contribution to the study is important. This is something that should be included in the covering letter and it is also a decent opportunity for the researcher to convince the respondents to reply to the survey which in the end can result in an increase in the questionnaire response rate (Saunders, Lewis & Thornhill 2016).

Bryman and Bell continue to explain that it is important to become familiarized with the position of the respondent. Suggesting that through understanding the respondent's perception of the questions will help to determine if there are any questions which are

21

general questions tend to result in the respondents misunderstanding the context of the questionnaire due to a lack of attention (Bryman & Bell 2015).

The researcher should also construct simple and specific questions and avoid long and unspecified questions (Lietz 2010). If this is not considered, there is a chance that the

questions might become leading or double-barreled, which can greatly influence the respondent's responses to the questionnaire. Leading questions tend to push the respondent towards a certain direction which means that it sort of suggests a certain reply to the respondents. Double-barrelled questions refer to when one question consists of several questions, formulated in such a way that the question actually measures several things, for example, “Are you satisfied with your employer and co-workers?”. The problem with double-barrelled questions is that it leaves the respondent in a certain position that can cause

uncertainty on how to respond best to the questions presented. In place of double-barrelled questions, the researcher should consider emphasizing one question at a time in order to ensure that the respondents focus only on that specific question, which results in a better outcome quality (Bryman & Bell 2015).

Malhotra describes the importance of choosing the right wording when it comes to the formulation of questions to a survey. The reason for this simply depends on that the questions should be clear and easy for the respondents to understand (Malhotra 2010). Bryman and Bell suggested that questions including negative wording such as ‘not’ or ‘would you rather’ should be avoided when formulating the questions. The reason for this is that the respondents tend to miss these words out when reading the questions, which in the end can come to affect the outcome of the questionnaire since they perceive it to ask something different then what it actually does. Instead, the questions should intend to be formulated with a positive design, since this will make the questionnaire easier to respond to. However, in certain contexts, it can be difficult to determine when a respondent answer ‘yes’ or ‘no’ and its meaning in relation to the questions. This is a common problem when conducting Likert scales since here it can be difficult to avoid questions that are formulated with a negative format (Bryman & Bell 2015).

22

‘one or two times’, ‘more than four times’, is more appropriate wording since it does not provide the respondent with different meanings to each answer (Malhotra 2010).

In regards to the design of a questionnaire, the respondents are likely to see statements positioned in a certain way, depending on how information is ordered. It is stated that

respondents often read the first or the last statement in a questionnaire, which can result in information being missed. If the questions in the questionnaire are placed in an order that is not possible to predict, it helps to form the perception that questions that measure the same concepts are placed in random order throughout the survey. Due to this, a randomized order can be applied to the questionnaire in order to encourage the respondent to read each question carefully and not simply fill out the same answer to similar questions (Malhotra 2010).

It is also important for the researcher to consider how the design of the rating scale influences the responses. First of all, it is of great importance to include every possible alternative a respondent could pick according to the context of the question. So that the respondent finds a suitable answer according to the statement. Further, it is also important that the respondent only chooses one alternative in the context of the rating scale if no other instructions are given. The reason for this is to avoid confusion for the respondent while also making the questionnaire easier to answer (Malhotra 2010).

4.5.3 Implementation of Questionnaire and Data Collection

The self-completed questionnaire for this study was created in Google forms. The reason for this is simply that the authors of this study possess prior knowledge regarding the use of Google forms. In accordance to Saunders, Lewis, and Thornhill (2016), the questionnaire was introduced with a clear covering letter as an introduction to the questionnaire explaining what the purpose of the study is, and also to inform the respondents what the survey is about. The questionnaire was conducted in Swedish. The reason for this was to eliminate any language barriers that might arise, which could have influenced the respondents understanding of the questions in the questionnaire. According to Malhotra, in order for the respondents to have a clear understanding of the questions and its meaning (Malhotra 2010). This also leads to the sample being limited to Swedish speakers only.

The questionnaire was introduced with a covering letter including a clear description of the purpose for sending out the questionnaire, as well as any other information the

23

previously were asked. These questions were asked to form a perception of how many of the respondents had previously written online reviews, and if it having performed the behavior previously has any influence on their intention of writing online product reviews in the future.

In the second chapter of the questionnaire, the respondents were assigned particular statements founded on the social and personal motives presented in the theoretical chapter. In accordance to Malhotra, the questions of the questionnaire were provided in a randomized order in order to prevent that the answers should affect each other since there were a few questions that could be considered as similar (Malhotra 2010). The respondents were asked to respond to the questions to the best of their ability, and thorough responding to the provided interval scale, ranging from 1 to 5. Where 1 meant that the respondent strongly disagreed with the statement and 5 meant that they strongly agreed. The fourth and final chapter of the questionnaire covered general questions regarding the respondents age, sex, employment but also how often they purchase a product online during a regularly month in order to get an understanding regarding how often the respondents to this survey purchase products online, which in turn can contribute to submitting online product reviews. The questionnaire used in order to gather the data for this paper can be found in Appendix 3.

4.5.4 Pretesting

As suggested by Bryman and Bell, pretesting a questionnaire is an important step in order to make sure the questionnaire is formulated in the correct way (Bryman & Bell 2015). For questionnaires, in particular, pretesting is an important step to perform. Since there is no interviewer present to solve any confusion that might arise when the respondents are answering the questionnaire, it is crucial to have questions that are clear and easy to understand. The purpose of doing a pretesting is simply to investigate whether there are specific uncertainties in the study that can be difficult for the respondent to understand. By detecting these uncertainties in time, this can result in a better understanding of the content in the questionnaire, which increases the likelihood that the answers from the respondents contribute to valid data (Malhotra 2010).

It is important to include all the aspects when doing a pretest of a questionnaire. This means that inclusion of the questions content, wording, layout, question difficulty as well as instructions are necessary in order to detect and remove potential problems with the

24

be done in several different ways, such as through, e-mail, telephone, or through other electronic channels. However, when it comes to the pretesting of a questionnaire, it is suggested to be completed through personal interviews due to the fact that it allows the researcher to observe the reactions and attitudes among the respondents when pretesting the questionnaire (Malhotra 2010).

For this paper, the authors conducted a pretest of the questionnaire in order to detect eventual areas in the questionnaire that can be considered as unclear and that should be clarified. During this process, it was found that language barriers occurred and affected the respondents understanding of some of the questions incorporated in the questionnaire. Therefore, the authors suggested conducting the questionnaire in Swedish in order to

eliminate any difficulties in understanding the questions. Furthermore, there were a couple of questions included in the questionnaire that were considered as unclear and had to be

corrected before sending out the final version of the self-completed questionnaire.

4.6 Sampling

The definition of a sample can be stated as a group of individuals that all are a part of the population in a society. These individuals are chosen in order to complete an investigation. The reason when it comes to determining a sample from the population would be that not all members of a society have the skillset to fulfill the problematization of the study and to select the whole society would also be rather time and money consuming (Bryman & Bell 2015).

There are two primary approaches when it comes to selecting a sample group, probability sampling, and non probability sampling. In this study, a non probability sample will be selected. A non probability sample is defined as a sample that has not been chosen to be totally random. Meaning that everyone in society will not have an equal chance of

participating in the study. Within the non probability sampling techniques, a convenience sample has been selected to be applied when collecting the data. This basically means that the participants are relatively accessible for the researchers. Further, sampling easily accessed participants can help to reduce the research cost as well as saving time since the sampling technique is not as time consuming as other sampling methods (Bryman & Bell 2015).

25

determining the studies sample size, it is important for the researcher to conduct a thorough review of the literature and understand any potential errors with the other studies sample size selection and be careful not to incorporate the same flaws (Israel 1992). When analysing similar studies within the eWOM context, the authors found the study by Cheung and Lee, which has incorporated a sample of 203 respondents (Cheung & Lee 2012). The study conducted by Yusuf, Che Hussin & Busalim which had applied a sample sellection of 218 respondents (Yusuf, Che Hussin & Busalim 2018). The study by Aghakhani, Karimi, and Salehan used a sample selection of 202 respondents (Karimi & Salehan 2018). Since all of these studies have a sample selection of around 200 (±18), a sample size of at least 200 respondent was applied to this study.

4.6.2 Sample Selection and Data Collection Procedure

This study collected data through the application of two different approaches. The data were collected both by sharing the questionnaire on social media, as well as through physical questionnaires. A total of 126 answers were collected by sharing the questionnaire on social media. An additional 97 responses were gathered by asking visitors of the university library at Linnaeus university in Växjö to fill out a physical paper questionnaire. The physical questionnaire followed the same design as its online counterpart, which can be seen in Appendix 3. Through this mixed approach of data collection, a total of 223 answered were collected. However, one of the answers was not considered valid since the respondent had not completed the entire questionnaire and was thus disregarded in the analysis of the data. This resulted in a total of 222 valid responses, which is larger than the minimum requirement for this study (200).

4.7 Data Analysis Method

26

the researcher. Furthermore, it can also provide support for avoiding false findings and assumptions about the emerged data. Moreover, it offers a foundation in order to understand analysis guided by others and the use of particular techniques when it comes to data analysis, which can affect the research objectives and the design of a certain study (Aaker et al. 2011). For this study, IBM SPSS has been used as a tool for analyzing the statistical data that has emerged from the survey. According to Malhotra, the use of SPSS plays a crucial role in various stages in the process of marketing research. SPSS usage will contribute to the preparation and analysis of data, which in turn will come to be presented (Malhotra 2010).

4.7.1 Descriptives

Saunders, Lewis and, Thornhill describes descriptive as a generic term for how variables can be described through statistics (Saunders, Lewis, and Thornhill 2016). Two of the most important values presented in the descriptives table is the skewness and kurtosis levels. Skewness refers to how the values are distributed on each side of the mean, the lower the skewness level is, the more evenly distributed the data set is. Malhotra describes that the skewness level should be between -1 and 1, and ideally as close to 0 as possible, since this indicates a lower skewness level. Kurtosis on the other hand measures the distribution peaks of the data set. The normal distribution kurtosis is -1 and 1, where a positive kurtosis level suggests that the data set has a higher peak than normal, and a negative value suggests that the distribution is flatter than normal. For the kurtosis, the skewness value should between -1 and 1, however a value between -2 and 2 can also be accepted according the particular application (George & Mallery 2003).

4.7.2 Correlation Analysis

27

correlation is also heavily influenced by outliers, so the data has to be cleaned thoroughly before conducting the analysis (Moore, McCabe & Craig 2017).

4.7.3 Regression Analysis

Regression analysis is used in order to summarize the relationship pattern between two variables. Through analysing the regression, it becomes possible to predict how the value of an independent variable (x) impacts the value of the dependent variable (y). The intention when it comes to regression analysis is to construct a model or equation which describes the relationship between a dependent variable to one or more independent variable(s). The model or equation can be used in order to describe, predict or control the dependent variable or interest on the root of the independent variable(s). In contrast with other methods of data analysis, which merely deal with only two variables, regression analysis is an effective and beneficial tool since it can handle more than two variables at the same time. Besides the quantification of individual relationships, regression analysis can also provide statistical control of the coverage variables included (Aaker et al. 2011).

Regression analysis is a key step in conducting a significance analysis (hypothesis testing). Significance testing revolves around testing the significance of the regression

28 4.8 Quality Criteria

It is a significant step to include quality criteria when it comes to business research and evaluation of data. When it comes to quality criteria, there are three major criterias that this study will take into consideration when it comes to managing the gathered data. The first criteria refer to validity, which includes content validity, construct validity and criterion validity. Furthermore, the second and third criteria that must be fulfilled refers to reliability and replication (Bryman & Bell 2015).

4.8.1 Content Validity

The validity of the content in a survey is an important aspect to keep in mind when doing an investigation (Bryman & Bell 2015). Content validity, also known as face validity, describes to what extent a survey is measuring what it is supposed to measure. For instance, if one item has the intention to measure pain, then it should measure pain as well. If the item for any reason measure anxiety instead, the study should not be considered as valid (Malhotra 2010). As suggested by Bryman and Bell, content validity can be established by pretesting the questions designed to measure the concepts. Explaining that by asking other people how they interpret the question, the researcher is provided with an understanding of how well the question measures the associated theory (Bryman and Bell 2015). In line with this, as described in chapter 4.5.4, the developed questionnaire was pre-tested in order to determine that the study has content validity.

4.8.2 Construct Validity

29 4.8.3 Criterion Validity

Malhotra explains that criterion validity reveals whether a certain scale presents a certain performance as expected in comparison to other criterion variables with significant criteria. When it comes to criterion validity, these validities can contain particular demographic and psychographic characteristics, factors measuring attitudes and behaviors, but also scores gained from other scales. There are two forms of validity to consider when it comes to criterion validity concerning the time period. These two forms refer to concurrent- and predictive validity. Regarding concurrent validity, this form is measured when the data of the scale is estimated during the same occasion as the collection of the criterion variables. On the other hand, predictive validity occurs when the researcher gathers data during one specific occasion and the data for the criterion variables subsequently (Malhotra 2010). Concurrent validity can be established though measuring how the predictable variables correlate with each other. Predictable validity, on the other hand, is established when a measure can be found to serve as a predictor for any type of future event (Aaker et al. 2011). In order to address the concurrent validity of this study, a correlation analysis was performed on the gathered data. In addition to this, the predictable validity was established through measuring the outcome of the hypothesis testing where the influence, i.e. the predictability, of the independent variables on the dependent one was determined.

4.8.4 Reliability

Reliability associates to the consistency of a measure of a concept. For instance, if it turns out in a survey that all of the answers emerged to point out a similar result, the outcome of the survey can be perceived as reliable. However, if another study shows that the answers point in different directions, the outcome of the survey is interpreted as not reliable (Malhotra 2010). When it comes to the concept of reliability, there are three outstanding factors to consider which determine if a measure is reliable. These factors refer to stability, internal

reliability, and inter-rater reliability. When it comes to stability, this factor refers to retest

30

alpha will be used as the method for establishing internal reliability. Cronbach’s Alpha is described as a statistic that measures the consistency of responses between a set of question designated to measure a specific concept. The Cronbach’s alpha consists of an alpha

coefficient (α) which has a value between 0 and 1 (Saunders, Lewis, & Thornhill 2016).

4.8.5 Replication

Replication is a common concept when working with quantitative research. The concept refers to when reproducing a certain study due to a lack of relevant evidence that has emerged from the original study. Obviously, for replication to take place in a study, the study must simply be replicable. This means that the researcher has to point out the whole procedure in great detail for replication to be possible (Bryman & Bell 2015). For this paper, a quantitative research strategy with a cross-sectional research design has been used in order to collect important data for this paper. When it comes to cross-sectional research design, replication plays a crucial role since the level of replication simply depends on the degree to which the researcher share information regarding the process of selecting respondents to the

questionnaire. Moreover, replication is also founded on how well the researcher provides information about the design to measure concepts, research instruments as well as how to analyse the data that emerged from the survey. Although, when it comes to the cross-sectional design, it is common that the researcher shares this procedure to a large degree, which in turn affects the replicability of the study (Bryman 2016).

4.9 Ethical and Societal Considerations

31

It is important to reflect on anonymity when it comes to the respondents of a

questionnaire, especially for data that is gathered online. Partly, due to legal issues but also ethical considerations. The reason for this is simply to protect the participants of the survey. Furthermore, it is also important to make sure that the participants have received enough informed consent regarding the investigation process in order for them to make the decision to take part in the survey or not. Hence, one should clearly express the importance of their participation, what it involves and how the data that emerges from the survey will be used (Bryman & Bell 2015). In line with how the respondents should be protected, this study has ensured total anonymity among the respondents who participated in the survey. Also that the data that has been collected will only be used for purposes associated with this study.

Furthermore, a clear and extensively introduction to the survey was presented for the

respondents to take part in in order to understand the context as well as the aim of the survey. When the respondents had completed the survey, they received an appreciation for their participation in the survey.

From a societal perspective, there are some requirements to take into consideration when it comes to research and the collection of data. The Social Research Association (SRA) states that it is important for the researchers of a study to handle the work of the study with responsibility and also in line with the permitted order of the society in which the

investigation takes place. The reason for this stem from how the work should not only be beneficial for the investigation but the results should also be beneficial for the society as well (Social Research Association 2003). In accordance to this and when it comes to the

32

5. Results

This is the section of the paper where the authors present the results that emerged from the survey. By way of introducing this chapter, the authors will provide demographic and descriptive information regarding the respondents to the survey. Thereafter, the authors will present the reliability of the variables measured. Furthermore, a correlation analysis will be presented as well as testing of the two hypotheses presented in the conceptual framework.

5.1 Control Questions and Descriptives

For this study, a total of 223 respondents consented and filled out the questionnaire. One of the respondents did not fill out the entire questionnaire, and their response was thus excluded from the study. Out of the 222 responses that were analysed in SPSS, 92,8% of the

respondents were below 34 years of age, 87,4% were students and 84,2% are low to moderate online shoppers. Regarding the gender of the respondents, the frequency is more evenly distributed with a total of 45,5% male respondents, 54,1% females and 0,5% prefered not to share. The responses for the two questions regarding previous review behavior, 67,1% had written an online review before, and 50,5% of them had written an online review about a product before. Since 97 respondents (43,69%) were surveyed at the campus of Linnaeus University in Växjö, it is a reasonable outcome that most of the respondents are the same age group and has the same primary occupation. A full list of the demographics of the

respondents can be seen in table 5.1.A Demographic Table.

33

Prefer not to say 1 0,5%

Total: 222 100% Occupation: Student 194 87,4% Employed 23 10,4% Self-employed 3 1,4% Unemployed 2 0,9% Total: 222 100% Frequency of online purchases per month: Never 18 8,1% 1-5 times 187 84,2% 6-10 times 14 6,3% 10 or more times 3 1,4% Total: 222 100%

Have you ever written an online review before?

Yes 149 67,1%

No 73 32,9%

Total: 222 100%

Have you ever written an online review about a physical product before? Yes 112 50,5% No 110 49,5% Total: 222 100%

34

The descriptives table aims to provide a clear overview of how the respondents responded to each of the items used on average. Since the items for the independent and dependent

variables measured the variable through a 5 point likert scale, the responses vary between 1 and 5. Two of the most important values to look at in the descriptives table is the skewness and kurtosis levels. Skewness refers to how the values are distributed on each side of the mean, the lower the skewness level is, the more evenly distributed the data set is, thus it is ideal for the skewness level to be between -1 and 1, and ideally as close to 0 as possible. Kurtosis, on the other hand, measures the distribution peaks of the data set. The normal distribution kurtosis is 0, where a positive kurtosis level suggests that the data set has a higher peak than normal, and a negative value suggests that the distribution is flatter than normal. For the kurtosis, the value should be as close to 0 as possible, however as long as it is between -2 and 2, it is still considered to be adequate (Malhotra 2010).

Item Mean Median Mode Std.

Deviation Skewness Kurtosis Pers 1 3,93 4 4 1,048 -0,925 0,336 Pers 2 3,31 3 3 1,092 -0,285 -0,519 Pers 3 3,74 4 4 1,008 -0,484 -0,380 Pers 4 3,50 4 4 1,067 -0,451 -0,284 Soc 1 3,16 3 4 1,102 -0,285 -0,623 Soc 2 3,53 4 4 1,014 -0,349 -0,419 Soc 3 3,50 4 4 1,195 -0,535 -0,549 Soc 4 2,15 2 2 1,007 0,663 -0,152 Intent 1 2,76 3 3 1,215 0,223 -0,844 Intent 2 3,68 4 4 1,042 -0,428 -0,505 Intent 3 3,35 3 3 1,193 -0,213 -0,823

35 5.2 Reliability Analysis

Cronbach's Alpha (α) is a measurement of how well the items measure the variable they are supposed to measure. The α level is always between 0 and 1, where a higher α signifies a higher reliability level, thus a more accurate measurement of the variable. As a general rule of thumb, the α level is considered ideal when above 0,8, good when it is above 0,7 (Bryman & Bell 2015) and it can be acceptable when it is above 0,6 (Malhotra 2010). As can be seen in table 5.2 Reliability table, the α level for behavioral intention is ideal (0,848). The α level for personal motives is also acceptable at 0,670. The α level for social motives, on the other hand, is quite low at 0,516. Malhotra describes that such a low α value might be associated with the variable being multidimensional. Describing that certain multi-item scales are developed to measure multidimensional constructs, where each dimension is somewhat independent of each other. Suggesting that when a construct is deemed to be

multidimensional, a reliability analysis is not he most appropriate manner of determining internal consistency (Malhotra 2010).

Variable Cronbachs Alpha (α) # of items (n)

Personal motives 0,670 α 4

Social motives 0,516 α 4

Behavioral intention 0,848 α 3

Table 5.2 Reliability Table

5.3 Correlation Analysis

36

Variable Personal motives Social motives

Personal motives 1 0,439**

Social motives 0,439** 1

** Correlation is significant at the 0.01 level (2-tailed). Table 5.3 Correlation Table

5.4 Hypothesis Testing

Table 5.4 Significance test, demonstrates the level of significance for the two hypotheses. The hypothesis test is done by conducting a regression analysis, initially with the control questions and then conducting another regression analysis with the inclusion of the independent variable, starting with adding one at a time (Malhotra 2010). Finally, the

regression is run with the inclusion of all of the control questions in addition to the inclusion of all of the independent variables, which is demonstrated in M4. Here, the level of

significance can be seen for each of the hypothesis, where significance is demonstrated by stars added to the beta value. Thus, the M1 describes the correlation between the dependent variable and the control question. M2 adds the independent variable Personal motives in to the mix, which is substituted by Social motives in M3. Finally, M4 displays the relationship between all the variables and control questions, thus it provides an indication of whether the hypothesis will be accepted or rejected.