I Decide What I Do. Right? : Persuasive design factors and purchase intention

I Decide What I

Do. Right?

MASTER THESIS WITHIN: Informatics NUMBER OF CREDITS: 30

PROGRAMME OF STUDY: Information Architecture and Innovation

AUTHOR: Toma Vlada JÖNKÖPING May 2020

Master Thesis in Informatics

Title: I decide what I do. Right? Persuasive design factors and purchase intention

Author: Vlada Toma

Tutor: Andrea Resmini, Domina Kiunsi

Date: 2020-05-19

Key terms: Persuasive design, Persuasive Factors, E-commerce, Attitude Change, Purchase Intention, Need for Cognition.

Abstract

Online shopping has gained more and more popularity over the years, naturally, leading to an increase in the number of e-commerce websites. From a variety of websites users can choose the one that best suits their needs depending on different factors, such as information or easy navigation. E-commerce websites have the aim to sell, thus besides providing information and easy navigation other strategies such as persuasion are implemented on the website to increase its selling performance. Some of the well-known persuasive design factors and how they influence user’s purchase behaviour will be discussed in this study.

The purpose of this thesis is to investigate the effects of persuasive design factors on purchase intention, as well as to assess the effect of the need for cognition as a user condition on the relationship.

Based on the literature review, a theoretical framework is proposed to accomplish the aim of the thesis. For the empirical evidence, a quantitative approach was taken, with the help of two online questionnaires (one with persuasive factors and one without persuasive factors). A non-probability convenience sampling approach was used for gathering data.

The results of the study provide evidence that using persuasive design factors on an e-commerce website can increase the likelihood of purchase intention. The findings do not conclude the mediating role of attitude towards using the website, yet a relationship between attitude and purchase intention was found. The moderating effect of need for cognition was not established.

Acknowledgements

I would like to thank everyone who supported me and was involved in writing this thesis. Especially, I would like to express my appreciation and gratitude to my supervisor Andrea Resmini, who has provided support and constructive feedback on this thesis throughout the entire process of writing. As well, I would like to thank Domina Kiunsi for providing valuable guidance and advice while writing this thesis.

Furthermore, I would like to thank my group mates who have provided constructive feedback during the seminar sessions.

In addition, many thanks are expressed to the participants of this study for completing the questionnaires. Without their feedback the study could not have been conducted.

Finally, I would like to thank my family and friends for emotional support and motivation. Thank you!

Table of Contents

1 Introduction……….……….………. 1

2 Theoretical Framewrok ... 6

2.1 Approach to literature review ... 6

2.2 Overview of user experience ... 6

2.3 Website persuasion ... 8

2.3.1 Defining persuasion ... 8

2.3.2 Design vs persuasive design ... 9

2.3.3 Persuasive design ... 11

2.3.4 Persuasive design in the context of e-commerce websites ... 13

2.4 Theoretical model background ... 14

2.4.1 Technology acceptance model ... 14

2.4.2 Elaboration likelihood model ... 16

2.4.3 Cialdini principles of persuasion ... 17

2.5 Research framework and hypotheses development ... 20

2.5.1 Behaviorial intentions: purchase ... 21

2.5.2 Attitude towards using the website ... 21

2.5.3 Need for cognition ... 22

3 Methodology ... 24

3.1 Research philosophy ... 24

3.2 Research approach ... 24

3.3 Research design and strategy ... 25

3.4 Data collection method ... 26

3.4.1 Survey design ... 26

3.4.2 Reliability and validity ... 27

3.4.4 Pilot testing ... 28

3.5 Data analysis ... 28

3.6 Methodology and ethical considerations ... 29

4 Empirical Findings and Analysis ... 30

4.1 Demographic profile and sample ... 30

4.2 Descriptive statistics ... 30

4.3 Hypothesis testing ... 32

4.3.1 Performing t-tests ... 32

4.3.2 Correlations ... 33

4.3.3 Regressions ... 35

4.3.3.1 Regression analysis: persuasive design factors ... 36

4.3.3.2 Regression analysis: scarcity ... 38

4.3.3.3 Regression analysis:reciprocity ... 39

4.3.3.4 Regression analysis: social proof ... 40

5 Discussion ... 43

6 Conclusion ... 45

6.1 Research questions ... 45

6.2 Theoretical implications ... 46

6.3 Managerial implications ... 46

6.4 Ethical issues of persuasion design ... 47

6.5 Limitations ... 48 6.6 Future research ... 50

References ... 51

Appendix

Figures

Figure 1 Three types of persuasion ... 9

Figure 2 The Fogg Behavior Model of persuasion ... 12

Figure 3 The revised conceptual model of persuasive visual design for web design……. ... 13

Figure 4 Technology Acceptance Model ... 15

Figure 5 Elaboration Likelihood model ... 16

Figure 6 Proposed Research Framework ... 20

Figure 7 Descriptive statistics: mean scores for scales for both groups ... 31

Figure 8 Moderated mediation model: persuasive design factors ... 37

Figure 9 Moderated mediation model: scarcity ... 39

Figure 10 Moderated mediation model: reciprocity ... 40

Figure 11 Moderated mediation model: social proof ... 41

Tables

Table 2 Descriptive statistics and independent t-tests ... 32

Table 3 Correlation matrix for experimental group: persuasive design factors .... 33

Table 4 Correlation matrix for experimental group: scarcity, reciprocity, social proof ... 34

Table 5 Collinearity diagnostics ... 35

Table 6 Results of the moderated mediation model regression for persuasive design factors ... 37

Table 7 Results of the moderated mediation model regression for scarcity ... 38

Table 8 Results of the moderated mediation model regression for reciprocity .... 40

Table 9 Results of the moderated mediation model regression for social proof.. . 41

1.

Introduction

Technology is an important part of people’s daily lives. Access to internet through smartphones, tablets and computers allow people to easily manage their tasks, shop, work and much more. Such devices turned out to be one of the things people cannot imagine their lives without and browsing on websites to find anything is typical nowadays.

Moreover, e-services have expanded together with the growth of internet. Most of the information can be accessed online, making it convenient to look up anything online at any moment. Hence, “a wide range of services, from government to private agencies, businesses or public welfares, sports or entertainment; all in various possible ways, are reachable within a mouse click” (Ibrahim & Shiratuddin, 2013). This grants users with the choice of what to access, whether to stay or to leave, to remember or to ignore certain web locations, which makes them the ones in control (Ibrahim & Shiratuddin, 2013). However, the great amount of online information brings in the paradox of having a wide array of choices, constantly questioning and stressing over what to choose (Schwartz, 2004). Hence, online users struggle to find the needed information online and frequently are likely to make impulsive decisions (Ibrahim & Shiratuddin, 2013; Kim & Fesenmaier, 2008). In addition, it has become a challenge for businesses and web designers to come up with strategic ways to motivate users to remain on their site and to influence them to make certain decisions that are in line with business goals (Ibrahim & Shiratuddin, 2013).

However, the domain of Human Computer Interaction (HCI) advanced - which leverages techniques and models from computer science and several other disciplines like human behaviour, psychology and cognitive science (Harris, Qadir, Khan, & Islam, 2017). Examples of such techniques are the persuasion principles by Cialdini (Cialdini, 2009), the elaboration likelihood model (ELM) (Petty & Cacioppo, 1986), or Fogg’s behavior model (Fogg, 2009). They are put in use to design technology so that HCI becomes as intuitive and as user friendly as possible (Harris, Qadir, Khan, & Islam, 2017).

According to Tam & Ho (2005), due to advances in tracking and database technologies companies are more and more able to figure out their customer’s needs and translate these

needs into products/services that attract them. Furthermore, companies are able to tailor their offerings on an individual level by using different technologies such as “collaborative filtering, data mining, and click-stream analysis”. Different stimuli included in a website such as audio, video or text are implemented to facilitate the decision-making process (Tam & Ho, 2005). Besides these, the design, content, colour and timing of those stimuli represent diverse attempts to influence a user (Tam & Ho, 2005).

With interactive technology, the possibility to influence behaviour has extremely increased. Nowadays, technology can adjust to the users’ actions, which creates an opportunity to show the most appropriate persuasive principle. Advancing technologies available to designers means tailoring the user experience by weaving persuasive cues into it (Interaction Design Foundation, 2019). These factors “rely on the verbal and non-verbal symbols and allow people voluntary participation in the persuasion process” (Kim & Fesenmaier, 2008; Oinas-Kukkonen & Harjumaa, 2008).

Social media platforms such as Facebook and Twitter, are examples of the capability of technology to persuade people to stay, buy or engage in other behaviours. According to Harris et al. (2017), “the use of persuasive technology is on the rise in multiple domains where technology can be used as an effective medium to influence human behaviour” (Harris, Qadir, Khan, & Islam, 2017). Furthermore, persuasive technology can be applied across many different areas with different purposes such as:

- Commerce - to sell products

- Education and learning - to engage in activities that promote learning - Safety - e.g. to drive more safely

- Environmental conservation - to reuse resources - Preventative health care - e.g. to quit smoking

- Occupational productivity - to set and attain goals at work (Fogg, 2003; Harris, Qadir, Khan, & Islam, 2017; Ferraro, Stepanivic, & Ferraris, 2017).

This thesis will focus on online persuasion, specifically in the context of e-commerce websites.

When it comes to interacting with a website, persuasive techniques are applied to “influence a user by diverting attention, reallocating cognitive resources, and provoking effective responses and behaviours” (Tam & Ho, 2005). According to Alhammad & Gulliver (2014), for “e-commerce websites, the persuasive factors’ aim is to influence users’ attitudes towards achieving their final goal, which is buying a product from the website”.

Investigating persuasive factors on e-commerce websites is a relevant issue as online shopping has become more and more popular. Ebay.com is one of the most popular e-commerce websites reaching “174 million active buyers worldwide” for the first quarter of 2020 (Statistica.com, 2020). As well, ebay.com is an example of an e-commerce website that highly utilizes persuasive factors to bring more users and convert visitors into shoppers (Alhammad & Gulliver, 2014). Based on this, ebay.com will be used as an example for this thesis.

1.1 Problem

The amount of information to which a person has access online is overwhelming. However, this comes with the power users have which is to decide in favour of one product over another. This pushes e-service providers (i.e. websites offering online services) to come up with different strategies to influence users’ attitude and/or behaviour. Most businesses offer information and their products/services online. The number of studies linked to web user interface (UI) is also growing, many researchers and writers discuss and propose guidelines and approaches to web functionality and usability (Ibrahim, Shiratuddin, & Wong, 2014). Additionally, forms of interaction where physical elements are implemented online, highlight the close connection to the studies and the importance of persuasive technology (Redström, 2006). However, despite the rising interest in persuasive design, little is known about what effect persuasion factors have in the online context, specifically in the context of human-computer communication (Ibrahim, Shiratuddin, & Wong, 2015).

As persuasive technology is largely adopted by the public, “it becomes increasingly relevant to understand how to use such technologies and how to analyse them” (Torning & Oinas-Kukkonen, 2009). Moreover, according to Simpson, Slattery, and Utesheva

(2013), “these techniques have rarely been tested in order to determine if they are as effective in online contexts as offline” (Slattery, Simpson, & Utesheva, 2013).

In the field of information architecture, Resmini & Rosati, (2007) outlined a unified model for retail that can be applied to both "digital and physical spaces, providing a conceptual framework for the design of cognitive and informational continuity between environments that allows users not to shift constantly between diverse, often colliding patterns". The model, later generalized, integrated, and expanded (Resmini & Lacerda, 2016; Benyon & Resmini, 2017), provides design foundations for approaching the user experience across physical and digital spaces as one single whole, but does not specifically considers persuasive design factors (PDF) and their impact on online purchase behaviour: this research gap is what this thesis investigates in relation to e-commerce websites.

In e-commerce website context, persuasive factors are implemented to support product assessment and influence a user’s purchase decision (Gwee & Chang, 2013). Hence, an indicator of successfully implemented persuasive factors is users’ likeliness to purchase from the website (Chakravorti, 2004). Important to highlight is that the way an individual reacts to a piece of persuasive information can be further influenced by his/her personal traits (Gwee & Chang, 2013), such as one’s need for cognition (i.e. one’s tendency to engage and analyse information). The need for cognition theorizes that there are two types of people. The first type extensively analyses information before deciding and the second type does not need much information to decide, relying mostly on past experiences or other associations (Petty & Cacioppo, 1986). Therefore, when analysing the user behaviour, in this case purchase intention, it is important to include one’s personality trait, in case of this thesis being one’s need for cognition.

1.2 Purpose

Building on previous theories and frameworks, this thesis aims to examine the effect of persuasive design factors, on e-commerce websites, on the users’ purchase intention. Additionally, this thesis aims to develop insights into the role of users’ need for cognition as a personality variable, on the relationship between the PDF and purchase intention.

1.3 Research questions

Based on the problem and purpose of this thesis, the following research questions will be addressed:

RQ1: How do persuasive design factors influence purchase intention on e-commerce websites?

RQ2: How does a user’s need for cognition influence the perception of persuasive design factors on e-commerce websites?

1.4 Delimitations

The attention of this study will be limited to e-commerce websites and a number of persuasive factors (reciprocity, social proof and scarcity) that can increase the likelihood of online purchasing. Furthermore, the analysis will take place in the stage when the user would have to decide to buy or not. Hence, this thesis neither concerns the actual purchase of a product online nor on other motivating elements present on a website.

1.5 Definitions

Persuasive design is an area of design practice based on psychological and social theories

that focuses on influencing human behaviour through the characteristics of a product or service (Interaction Design Foundation, 2019).

Persuasive technology is any interactive computing system designed to change people’s

attitudes or behaviours (Fogg, 2003).

Persuasive design factors – verbal and non-verbal elements deliberately presented with a message containing information intended to engage and/or change one’s attitude (Briñol & Petty, 2005). In this thesis, persuasive factors are represented by a number of Cialdini’s persuasion principles.

2.

Theoretical Framework

2.1 Approach to literature review

An essential part of any academic project is to review existing relevant literature (Webster & Watson, 2002). A concept centric methodology was applied for this thesis’ literature review, with the purpose to synthesize relevant literature on the topic (Webster & Watson, 2002). Further, to gather the first insights on persuasion on websites, Google scholar and Jönköping University’s Primo search engine were accessed. Besides these search engines, different journal databases such as Research Gate, Science Direct and Scopus were used to access peer-reviewed articles and conference proceedings.

To get relevant literature for the topic, different keywords were used such as “persuasive design”, “persuasive design in e-commerce” and “persuasive techniques” (for a detailed list of keywords see Appendix 1). An analysis of the articles helped to scope down the focus of research which led to performing several more searches for an extensive analysis.

Moving on, several theories and models were identified while reading articles on the topic. Therefore, the search continued focusing on the models and using keywords such as “ELM”, “Cialdini principles”, “Fogg behavioural model”, and others. The same platforms were accessed.

Each keyword would give a certain amount of results, sometimes repeated articles. However, for each of the keywords the first 30-40 articles were analysed. When an article was useful it was downloaded for further analysis. In cases when articles were based on researches of other authors, those articles were included in the analysis as well. In total 156 articles were downloaded and revised.

2.2 Overview of user experience on websites

Before understanding persuasive factors and how they contribute to the e-commerce website’s experience, it is substancial to see what the basic elements are that support a positive user experience (UX) on a website.

User experience is described as a “concept that defines the user’s reactions and feelings, which arise from interplay with a system in certain situations or use contexts” (Forlizzi

& Battarbee, 2004). Hence, one can conclude that UX is the result of the interaction between people and a product/service. UX is important to keep in mind when designing a website because it provides features/qualities and tools that aim to support both pragmatic (goal-oriented) and hedonic (pleasure-oriented) experiences (Segerståhl, Kotro, & Väänänen-Vainio-Mattila).

Kim and Fesenmaier (2008) identify “informativeness” and “usability” as being two most important hygiene features that a website should have in order to provide a good experience (Kim & Fesenmaier, 2008). However, while analysing the model of Kim and Fesenmaier (2008), Ibrahim et al. (2014) included one more hygiene feature for creating a positive UX - "aesthetics” (Ibrahim, Shiratuddin, & Wong, 2014).

Following the three features will be explained:

• Informativeness: one of the main reasons for users to visit a website is information. (Ibrahim, Shiratuddin, & Wong, 2014). This factor relates to the quality and accuracy of the information presented to users. The way information is delivered online, can facilitate or hinder its utilization by the user (Rosen & Purinton, 2004; Ibrahim, Shiratuddin, & Wong, 2014).

• Aesthetics: the importance of visual aesthetics has been long emphasized by the researchers in Information Systems field, highlighting its importance for user satisfaction and visual appeal. Aesthetics has a range of definitions, such as beauty in appearance (Lavie, 2004), visual appeal (Lindgaard & Dudek, 2003), an experience (Ramachandran & Blakeslee, 1998), an attitude (Cupchik, 1993) and others. All these definitions testify to the complexity of this feature (Lindgaard G. , 2007). In addition, with regards to the first impression, aesthetics might be the factor that engages and influences users’ decision to stay on a website longer (Reineck, et al., 2013).

• Usability: usability can be defined as “the ability of an interactive system to assist users and tasks” (Bock, Kuan, & Vathanophas, 2003). Usability is related to such issues as readability, consistency, functionality, error handling, complexity (Ibrahim, Shiratuddin, & Wong, 2014). Usability might play an important role to

form a positive attitude, stimulate users to come back to the website and increase the probability for a user to buy (Becker & Mottay, 2001).

These are identified as the basic requirements a website must include to avoid user dissatisfaction. Other factors, such as the persuasive ones, are added to enhance user satisfaction, also referred to as motivating factors (Kim & Fesenmaier, 2008).

2.3 Website persuasion

While UX has an important role in developing good interaction between a website and the user, certain techniques are applied especially on the e-commerce websites with the purpose of enhancing user’s motivation to achieve a certain goal i.e. purchase.

2.3.1 Defining persuasion

The concept of persuasion exists since a long period of time. Aristotle, within the 4th century BC, discussed the idea of “rhetoric by means of ethos (appeals to credibility), logos (appeals to logic), and pathos (appeals to emotion)” (Ibrahim, Shiratuddin, & Wong, 2014). At that time, persuasion or rhetoric was explained as specific oral communication, referring to human interaction (Ibrahim et al., 2014). However, the meaning of persuasion has not changed a lot, it is still concerned with how communication can be done in order to change attitudes. According to Gass & Seiter (2007), persuasion is a process that takes into consideration attitudes, beliefs, motivations, intentions or behaviour (Gass & Seiter, 2007).

In the context of this research three types of persuasion were identified: interpersonal persuasion, computer-mediated persuasion and human computer persuasion (Harjumaa & Oinas-Kukkonen, 2007).

The difference between the three is:

• Interpersonal persuasion can happen between two or more people when it includes e.g. verbal and non-verbal behaviours, personal feedback, and coherence of behaviour and when at least one of the people involved contributes to changing attitude/behaviour (Oinas-Kukkonen & Harjumaa, 2008).

• Computer-mediated persuasion means “people persuading others through computer-mediated communication, e.g. e-mail, instant messages, or blogs” (Oinas-Kukkonen & Harjumaa, 2008).

• Human-computer persuasion is “the study of how people are persuaded when interacting with computer technology” (Fogg, 2003; Harjumaa & Oinas-Kukkonen, 2007; Oinas-Kukkonen & Harjumaa, 2008). This type of persuasion is different from the other two because it is not always clear who is the persuader in this case. “As computers do not have intentions of their own, those who create, distribute, or adopt the technology have the intention to affect someone’s attitudes or behaviour” (Harjumaa & Oinas-Kukkonen, 2007). This thesis analyses PDF from the human computer persuasion perspective.

Figure 1: Three types of persuasion. Source: Harjumaa & Oinas-Kukkonen (2007)

Important to highlight is that “persuasion implies a voluntary change of behaviour or attitude or both, and if force (coercion) or misinformation (deception) are used, these would fall outside of the realm of persuasive technology” (Redström, 2006).

2.3.2 Design vs persuasive design

Before understanding what persuasive design is, it is important to distinguish it from design itself. Certainly, “the notion of persuasion in design touches fundamental aspects of what design is about” (Redström, 2006).

Resmini and Rosati (2011), define design as an activity that is very “concrete, directed, and geared toward building both an artifact and a better comprehension of the problem space … design produces ideas and objects” (Resmini & Rosati, 2011).

Ralph and Wand (2009) propose their definition of design as: “design activity is a process, executed by an agent, for the purpose of generating a specification of an object based on: the environment in which the object will exist, the goals ascribed to the object, the desired structural and behavioural properties of the object (requirements), a given set of component types (primitives), and constraints that limit the acceptable solutions” (Ralph & Wand, 2009).

Design has also been defined as a point of view that acts as an argument in favour of adopting that particular point of view (Redström, 2006). This is based on Buchanan’s definition of design, which is: “an art of communication on two levels: it attempts to persuade audiences not only that a given design is useful, but also that the designer's premises or attitudes and values regarding practical life or the proper role of technology are important, as well” (Buchanan R. , 1985).

Buchanan’s definition suggests that the goal of a design is to persuade audiences, therefore the question is still in place: how persuasive design differs from design, if there is any difference at all.

Attempting to identify this difference, several articles were examined. With statements like: “compare the effectiveness of persuasive visual design against a plain website designed“ (Ibrahim, Shiratuddin, & Wong, 2015), or “the next wave in web site design is persuasive design” (Schaffer, 2009), it was noticed that authors make a distinction between design that excludes and design that includes persuasive factors. Throughout these articles design is considered less persuasive or more persuasive when other factors with a persuasive intent are included in the design (referring to web design).

Consequently, one can conclude that design is classified as persuasive when its objective is to influence the user by including other elements or messages in it. By contrast, Redström (2006), states that “all design is persuasive”, even if a designer’s aim is far from influencing. An example to this being the design of a simple artefact such as an umbrella. “Designers of the umbrella probably did not think of their design in terms of persuasion, however the resulting change in user behaviour, and possibly also attitude towards being outdoors when it is raining, is certainly intentional” (Redström, 2006).

Summarizing, no clear differentiation was found between persuasive design and design, yet there is a difference in the approach towards design which depends on the intent behind it. Next, as the focus of this research is persuasive design, the concept will be analysed as such.

2.3.3 Persuasive design

The Interaction Design Foundation platform (2019) defines persuasive design (PD) as an area of design practice that focuses on influencing human behaviour through a product’s or service’s characteristics (Interaction Design Foundation, 2019). Additionally, Nash et al. (2016), describes PD as the act of designing messages and features by analysing and evaluating the content, visual content and the way they are perceived and interpreted by users, using established psychological, sociopsychological and behavioural theories and models (Nash, Thomson, & Maeder, 2016; Fogg, 2009; Oinas-Kukkonen & Harjumaa, 2009). PD is often used in e-commerce, organizational management, and public health (Interaction Design Foundation, 2019).

More general, Redström (2006) analyses PD from three perspectives. The first perspective is user centred PD, which is focused on the user satisfaction and participation. In this case, PD is viewed as a process and not outcome. The second perspective is to view PD as a dialog between the designer and the user. This perspective takes into consideration not only the communication process, but also the artefacts resulted from combination of physical and digital space. The third perspective is to view the technical aspect of PD, which encompasses all the analytic tools, design methods, evaluation techniques that enable designers to predict their outcomes (Redström, 2006).

The difference between persuasive design technology and other technology is that the former is inherently more transformative, because it can “adjust to user’s inputs, needs and situations”, making for example different suggestions depending on the context (Fogg, 2003). Additionally, the aim of any persuasive technology is computer-mediated change of either attitudes or behaviours, including a change by boosting or reinforcing existing attitudes or behaviours (Torning & Oinas-Kukkonen, 2009; Oinas-Kukkonen & Harjumaa, 2009).

A contribution in this sense was done by Fogg (2009), who brought the concept of “captology”, later addressed as persuasive technology or persuasive design (Ibrahim, Shiratuddin, & Wong, 2014). Fogg (2003) describes this term as follows: “captology focuses on the design, research and analysis of interactive computing products created for the purpose of changing people’s attitudes or behaviour”. According to Fogg’s behavior model (FBM), behaviour is a product of three factors, namely motivation, ability and triggers. Motivation and ability are necessary to perform the behaviour, and triggers are different stimuli that often serve as a call to action. In brief, the model argues that “for a target behaviour to happen, a person must have sufficient motivation, sufficient ability, and an effective trigger” (Fogg, 2009). In the FBM model, Figure 2, can be seen that other combinations of the three factors are possible. For instance, when there is low ability, but high motivation and the trigger is present, the target behaviour can happen as well.

Figure 2: The Fogg Behavior Model of persuasion: motivation, ability and triggers. Source: Fogg (2008).

Ibrahim et al. (2015) analysed the effect of persuasive visual design on user motivation and behavioural intention, with a focus on the first impression. Their framework includes hygiene factors, motivating factors, and persuasive visual triggers which are based on Cialdini’s persuasive principles. To measure whether there is a difference in the

participant’s first impression, they used two websites. The first website including only the hygiene factors and the second website, with the same hygiene factors, as well including motivating and persuasive visual triggers. In their finding they concluded that visual persuasion plays an important role in influencing behaviour. Their result suggests that the users who viewed the website where persuasive factors were included provided much more positive insights compared to the users who viewed the non-persuasive website. However, there is no evidence how persuasive design affects perceived informativeness, usability and visual engagement (Ibrahim, et. al, 2015).

Figure 3: The revised conceptual model of persuasive visual design for web design. Source: Ibrahim et al. (2015)

Additionally, a study done by Kaptein et al. (2009), examines persuasive design in a quantitative study. Their research focuses on whether persuasive cues increase compliance to a request and how compliance depends on participant’s susceptibility. The results show that persuasive claims increase behavioural compliance while participants susceptibility is lower, which was in line with their proposed hypotheses (Kaptein, Markopoulos, de Ruyter, & Aarts, 2009).

2.3.4 Persuasive design in the context of e-commerce websites

Consumers nowadays have the choice of buying online or offline. The difference between the two spaces is that online consumers are “generally more demanding, powerful and practical in their shopping journeys” (Marios, 2002). Unlike other websites, e-commerce websites include attributes of an information and a selling environment, thus it is essential

to ensure that an e-commerce website provides all the required information, and that all the aspects are examined from user perspective (Ibrahim, Shiratuddin, & Wong, 2014). While browsing on an e-commerce website, the experience of a user is influenced by a range of factors among which, the availability of the good/service, relevant content and an enjoyable context (Eighmey, 1997). Moreover, certain site features, such as navigation, informativeness and relevancy, have an important impact on the overall experience while trying to achieve their goals (Marios, 2002).

Including persuasive factors on a website can influence UX as well. Within the context of e-commerce, the main focus is to help a user complete a specific task by providing an effective and efficient environment, which should lead to simplifying the decision-making process to buy from internet (Benbasat, 2010). The persuasive factors applied in e-commerce websites are quite like the strategies used by face to face resellers. For example, a typical persuasive factor adopted by e-commerce websites is showing the reviews of other customers or specialists. This persuasive factor, amid others, is used as crucial element on a website with the aim to influence a user’s attitude and behaviour (Oinas-Kukkonen H. , 2010; Alhammad & Gulliver, 2014). Crilly (2011) explains that this “happens because persuasion is achieved through the selective emphasis and de-emphasis of information; persuasion requires information, and providing or withholding information can be persuasive” (McCoy, 2000; Crilly, 2011).

2.4 Theoretical model background

In the next section, several models and concepts, which influenced the conceptual model of this paper will be presented.

2.4.1 Technology acceptance model

The technology acceptance model (TAM) is an information system theory widely used by researchers to understand the acceptance of new technology (Durodolu, 2016). The user acceptance is important because the technology that is to be designed is intended for use but also to change people’s attitude and behaviour towards some recognized goal (Harris & Qadlr, 2017).

TAM is based on the theory of reasoned action (TRA) that has the purpose to describe computer usage behaviour (Pantano & Di Pietro, 2012; Fishbein & Ajzen).

TAM theorises that “the attitude towards using a specific technology is determined by the perceived usefulness (PU) as well as the perceived ease of use (PEU) of that technology and finally leads to the intention to use as well as the actual use of the technology” (Davis F. , 1989). Consequently, based on the TAM, it can be concluded that users’ perception about a website is based on beliefs, attitudes, and intentions (Ibrahim, Shiratuddin, & Wong, 2014).

Figure 4: Technology Acceptance Model (TAM)

From the Figure 4, two main constructs can be observed – PU and PEU, considered to influence the attitude of a person towards new technology. PU is the extent at which a person believes that using a technology will improve their performance of routine responsibility (Davis F. , 1989). PEU is the degree at which one perceives a technology as being effortless and hassle free (Davis F. , 1989). Correspondingly, the two elements would change the attitude of a person, having an effect on their behaviour, in this context to use or not to use the technology.

When it comes to persuasive design factors, TAM plays a role in understanding what is the attitude of a person towards PDF (Redström, 2006). For example, Alhammad & Gulliver (2014) reviewed the TAM model towards persuasive technology which can help to evaluate users’ willingness to accept or reject online persuasion, assessing perceived

Perceived Usefulness Perceived ease of use Attitude towards using Behavioural intention

persuasiveness. However, their model still needs to be supported by empirical findings (Alhammad & Gulliver, 2014).

2.4.2 Elaboration likelihood model

The Elaboration likelihood model (ELM), developed by Petty & Cacioppo (1986), is a theory that proposes a global view of “how attitudes are formed and changed” (Fagerstrøm & Ghinea, 2009). With this model, Petty & Cacioppo (1986) proposed “a framework for organizing, categorizing, and understanding the basic processes underlying the effectiveness of persuasive communications”.

The term elaboration in this context means “the amount of effort an individual has to use in order to process and evaluate a message, remember it, and then accept or reject it” (Interaction Foundation Design (b), 2020). Drawing on this, the ELM theorizes that “the likelihood of elaboration will be determined by a person’s motivation and ability to evaluate the communication presented” (Petty & Cacioppo, 1986; Oduor, Alahäivälä, & Oinas-Kukkonen, 2014).

According to ELM a person can be persuaded through two routes: central (direct) route or peripherical (indirect) route. In short, the basic idea is that the receivers (in this case online users) are different in how they “engage in elaboration of information”, on which different persuasive messages are based. For instance, when information is personally relevant, users are willing to engage more and spend more time analysing that information. More attention will be paid on certain website features, examining the overall information, and remembering other past situations relevant to the presented information. By contrast, in cases when the information is not personally relevant, users will not spend much time analysing it, displaying rather little effort in elaboration (Ibrahim, Shiratuddin, & Wong, 2014).

Cyr et al. (2015) investigate online persuasion, using the ELM theory to determine the effects of argument quality, image appeal and navigation design on attitude change. Their findings conclude that there is a difference of how people are influenced and that people can be influenced depending on their level of cognition. A bigger influence being from the image appeal and argument quality - important in influencing change in attitude. In addition, they emphasize the prior engagement or knowledge which can determine through which route the influence can occur (Cyr, Lim, Head, & Stibe, 2015).

The ELM theory is based on a personal characteristic called in psychology “need for cognition” (Petty & Cacioppo, 1986). Need for cognition (NFC) is “a personality trait indicative of cognitive motivation and reflects one’s inclination to extensively seek and process information” (Petty & Cacioppo, 1986; Curşeu & de Jong, 2017).

Based on this theory, there are two types of people, with higher and lower need for cognition. Similar to the two routes presented in ELM, people with higher need for cognition tend to be more engaged with information than people with lower need for cognition.

Relating it to the topic of this research, Petty & Wegener (1998), state that a user’s need for cognition can be implemented in the persuasion strategy. People who have a high need for cognition tend to follow the central route to persuasion and those with low need for cognition tend to follow peripherical route (Petty & Wegener, 1998; Oinas-Kukkonen & Harjumaa, 2009). Therefore, to further research this, the need for cognition is employed as a moderator in the proposed theoretical framework for this thesis (Figure 6).

2.4.3 Cialdini’s principles of persuasion

The principles of persuasion or principles of social influence were proposed by Robert Cialdini in 1984 in his book called “Influence: The psychology of Persuasion”. The six principles described in his book were based on observations, covering different approaches. Cialdini’s principles are widely used in many fields (Ibrahim, Shiratuddin, & Wong, 2014). The six principles of persuasion are the following: reciprocation, consistency, liking, social proof, scarcity and authority (Cialdini, 2009).

According to Cialdini (2007) “consumers are often not aware of the techniques they are experiencing when looking at some information because people cannot always make elaborate decisions. Due to, for example, a lack of time or knowledge, one often bases decisions on heuristic or mental short cuts” (Cialdini, 2007). Based on this theory, Cialdini’s principles are extensively applied on e-commerce websites to change users’ attitude.

The principles have been also linked with the ELM (Oinas-Kukkonen & Harjumaa, 2008). This, because Cialdini’s persuasive principles support both direct and indirect ways of persuasion, depending on the context and aim (Oinas-Kukkonen & Harjumaa, 2008). An overview of the principles is shown in Table 1.

Table 1: Cialdini’s six principles of persuasion and their explanation, Cialdini (2009)

In a research conducted by Guadagno and Cialdini (2005), it was concluded that the principles are found to be effective in some contexts but not others (Guadano, Muscanell, & Greenlee, 2013; Guadagno & Cialdini, 2005). For example, an individual will not be influenced by a displayed authority (figure or brand), unless the presented authority is personally relevant (Guadano, Muscanell, & Greenlee, 2013). Thus, the authority principle would need a personalized approach, based on the history and actions of the

online user. The same approach is needed when it comes to commitment, consistency, and liking principles as they require information on previous actions of the user.

Therefore, due to time limitations and not having the resources to track users’ previous actions, only three of the six principles will be further included in the thesis analysis: reciprocity, social proof and scarcity.

Following, the three principles will be described:

• Reciprocity principle is based on the “human need to respond to positive behaviour received with positive behaviour in return” (Groves, Cialdini, & Couper, 1992). The reciprocity strategy aims to change people’s behaviour by including a service, gift or diverse favours before a person needs to take an action (Seethaler & Rose, 2005). Examples of such favours can be free trials, free gifts, codes with discounts, in brief anything that can be in people’s benefit.

• Social proof principle states that by looking at what others are doing and how they do it serves as a validation that it is worth to imitate that behaviour, thus influenced people generate the same actions. Examples of social proof principle strategies on a website are the ratings or reviews on the service/good (Ibrahim, Shiratuddin, & Wong, 2014)

• Scarcity principle is linked to people trying to avoid loss (Cialdini, 2009). By making a service/good appear scarce, users assign more value to it (Lynn, 1989). Examples of the scarcity principles on a website is when information about the limited time and amount or how many people are watching the same item is displayed.

Cialdini’s principles are independent of each other, each of them tackles different ways of changing one’s attitude and/or behaviour. The principles can be applied together, or separately in different contexts depending on their focus. As mentioned before, three of the principles – commitment, liking and authority, will not be examined in this research. So, using three out of the six principles does not have an impact on the overall objective of the research: to examine persuasive factors on purchase intention. Nevertheless, it can represent a limitation which will be discussed in chapter 6.5.

Based on the previously described models existing in the persuasive technology area, a conceptual framework will be proposed.

2.5 Research framework and hypothesis development

As more and more e-commerce websites adopt persuasive design techniques as an important part of their UX design strategy and their success depends on a thorough understanding of how these techniques can help their short- and long-term objectives there is an indispensable need to understand the way persuasive design techniques influence customers’ purchase intentions.

Therefore, to examine how persuasive factors influence the purchase intention in the e-commerce websites, a framework of the study based on the theory and models presented above was created for this thesis. The framework is presented in Figure 6, followed by a detailed discussion on each of the model’s constructs and the hypotheses formulation.

Figure 6: Proposed Research Framework

User condition Need for cognition

Mediating mechanism Attitude towards using

the website Persuasive design factors Reciprocity Social proof Scarcity Behavioural intention Purchase intention H1 H2 H2 H3

The thesis focuses on the persuasive factors implemented in web design on an e-commerce website. To analyse the persuasion context, three of the Cialdini principles are used as persuasive design factors in the proposed framework: reciprocity, social proof and scarcity.

As mentioned before, the role of the persuasive factors on a website is to change or reinforce the attitude of a user to perform a certain behaviour (Fogg, 2009), which, is case of an e-commerce website without doubt is purchase intention. The best way to test persuasive factors’ capacity to influence users’ purchase intention is by comparing users’ purchase intention on a website that capitalizes on PDF with one that does not use any, thus, the following hypothesis is proposed:

H1a: The presence of persuasive design factors on a website leads to higher purchase intention compared to websites with no persuasive design factors

2.5.1 Behavioural intentions: purchase

According to Li et al. (2002), a useful tool to measure customer behaviour is purchase intention (Li, Daugherty, & Biocca, 2002). Purchase intention is considered an immediate antecedent and the strongest predictor of people’s attitude towards the buying behaviour (Wang & Krikke, 2013). Therefore, the dependent variable used for the proposed model is purchase intention. The construct measures “a combination of consumers’ interest in, and possibility of buying a product” (Kim & Ko, 2012) and has a strong relationship with customer’s future purchasing actions (Hung, et al., 2011). Hence, the following hypothesis is formed concerning the relationship between persuasive factors and purchase intention.

H1b: The presence of persuasive design factors will have a positive effect on purchase intention

2.5.2 Attitude towards using the website

Attitude refers to a physical tendency that is expressed by evaluating a particular entity with some degree of favour or disfavour (Davis F. , 1989) and has a direct effect on the behavioural intention (Suwanno & Jarernvongrayab, 2014). Attitude is an important

factor for both TAM and persuasive design. Attitude has been described as one of the fundamental concepts of persuasion (Oinas-Kukkonen & Harjumaa, 2008) and it predisposes people to act in one way rather than another (Simons, Morreale, & Gronbeck, 2001). In the online context, Chen and Wells (1999) propose that attitude toward a website indicates the online shoppers' “predispositions to respond favourably or unfavourably” (Chen, Clifford, & Wells, 1999).

The attitude towards the website can be influenced by the information, usability and aesthetics combined (Hausman & Siekpe, 2009). Additionally, attitude has been linked with how people perceive a website based on personal experiences (Oinas-Kukkonen & Harjumaa, 2009). Based on the TAM the users’ perception about a website is based on their beliefs, attitudes, and intentions (Ibrahim, Shiratuddin, & Wong, 2014), whereas attitude has a mediating role between the user’s perception of a website and their actual purchase intentions. The same mediating role of attitude is adopted for this thesis; hence, the following is hypothesised:

H2: The positive effect of persuasive design factors on purchase intentions will be mediated by the attitude towards using the website

2.5.3 Need for cognition

As discussed before, need for cognition is a personality trait on which ELM theory is based. Need for cognition represents the tendency for individuals to engage in and enjoy thinking (Petty & Cacioppo, 1986). According to the need for cognition theory, there are people with high and low need for cognition (Petty & Cacioppo, 1986). People with higher need for cognition enjoy thinking, whilst people with lower need for cognition tend to avoid effortful thinking. The latter type of individuals is not described as unable to differentiate convincing from incorrect arguments, but rather they usually “prefer to avoid the effortful cognitive work deriving their attitudes based on the merits of arguments presented” (Haugtvedt, Petty, & Cacioppo, 1992).

The concept has been used in consumer research in examining the effects of persuasive factors. Among these applications, “the concept has been shown to be useful in understanding how argument strength and endorser attractiveness may influence

consumer attitudes” (Bearden & Netemeyer, 1999). The current study will also analyse the degree to which a user’s need for cognition influences their attitudes when exposed to persuasive factors, thus, the following hypothesis is proposed:

H3: The positive effect of persuasive design factors on attitude towards using the website is moderated by the need for cognition, such as it will be stronger for users with low need for cognition and weaker for users with high need for cognition.

3.

Methodology

3.1 Research philosophy

According to Saunders et al. (2009), the research philosophy one adopts depends on the research question meant to be answered. Consequently, the research philosophy dictates the research methods and strategy (Saunders, Lewis, & Thornhill, 2009).

There are three ways of thinking about research philosophy: ontology, epistemology, and axiology. Ontology is concerned with what reality is and how things exist. Epistemology is about the nature of knowledge and what knowledge is. Axiology is a branch that studies value, or what is valuable in the context of research (Babin & Zikmund, 2016).

The four known philosophies are: realism, interpretivism, pragmatism and positivism (Saunders et al., 2009). Realism considers reality to be objective, the reality exists independently from the human mind (Saunders et al., 2009). Interpretivism advocates that it is necessary for a researcher to understand the differences when studying humans or objects, hence it is advisable to apply different methods (Saunders et al., 2009). Pragmatism argues that the determinant of the study is the research question, hence it is not necessary to choose one philosophy, on the contrary it supports a variation among them (Saunders et al., 2009). Positivism assumes that “the researcher is independent of and neither affects nor is affected by the subject of the research” (Welman, Kruger, & Mitchell, 2005; Saunders et al., 2009).

A positivist research philosophy was chosen for this thesis. Within the positivism approach the data is produced from observed phenomena/reality. The positivist appoach uses theory in order to develop hypotheses, which need to be tested and cofirmed. This in turn will lead to further theory building.

3.2 Research approach

The main approaches to research are deductive and inductive. The two represent opposite approaches. Inductive approach starts with data, as the researcher introduces hers/his findings into theory (Bryman, 2012). The deductive approach has theory as a basis of the

research. Once the theory has been analysed, hypotheses are developed and tested with the purpose to develop the theory (Saunders et al., 2009).

This thesis uses a structured methodology to facilitate replication, ensure reliable data along with generalisation, which is an important feature of the deductive approach (Malhotra & Birks, 2007).

To facilitate a structured methodology and see the “causal relationship between the variables from the proposed hypothesis” (Saunders et al., 2009), the measurements should be done quantitatively (Johnson & Gill, 2010). Based on this approach and on the existing theories and models, hypotheses are proposed in this thesis.

To test the hypotheses, a quantitative approach is applied to collect the necessary data.

3.3 Research design and strategy

The research design is the plan or framework for answering the research question and can have an exploratory, descriptive, explanatory or evaluative purpose (Saunders et al., 2009). Exploratory approach has the purpose to clarify the nature of a problem; descriptive aproach has the objective to portray an accurate profile of persons, events or situations and explanatory approach has the purpose to study a problem or a situation to explain the relationship among the variables (Robson, 2002; Saunders, Lewis, & Thornhill, 2009).

This thesis aims to see how the persuasive techniques affect purchase intention. Therefore, an explanatory approach is used with the purpose to see the causal relationship between the variables presented in the study framework.

As mentioned before, a quantitative approach is used for this thesis. Additionally, given the research questions and the approach, a survey research strategy was chosen. Surveys are associated with deductive approach (Saunders et al., 2009). The gathered data from a surveys is standardized, and can be used to “suggest possible reasons for particular relationships between variables” (Saunders et al., 2009).

3.4 Data collection method 3.4.1 Survey design

In the context of this thesis, an online self-administered questionnaire, was designed to test the proposed hypotheses (see Appendix 3). Online survey tools promote a user centred model of creation and flow of information (Buchanan & Hvizdak, 2009). In case of an online survey, the participants answer the question themselves, the researcher not being present during the completion process (Babin & Zikmund, 2016).

The survey was created in two version, first - the experimental survey - with the presence of the persuasive factors, and second - the control survey - without the persuasive factors. The persuasive factors - reciprocity, scarcity and social proof were present in the experimental survey, specifically in the form of pictures with products that contained either of the factors. In the first picture the scarcity factor was presented by showing the number of views and the limited quantity of the product. Following, the second picture displayed the reciprocity factor in form of a discount and in the third picture, social proof was displayed by showing the reviews and ratings of a product.

The control survey included the same pictures of the same products but without the persuasive factors mentioned earlier.

The survey was created using the online platform Microsoft Forms, allowing an online anonymous gathering of data.

Before answering the questions, participants were asked to browse through an e-commerce website to familiarize themselves with it. The chosen website for this study was an e-commerce company eBay’s website - ebay.com. The decision to choose this website was taken because it uses the persuasive factors analysed in this thesis and it is one of the world’s largest facilitators of e-commerce (Elfenbein, 2019). Additionally, after investigating several e-commerce websites, it was clear that ebay.com uses the persuasive factors in a way that would be suitable to showcase in the survey.

Before continuing with answering the questions the participants were asked whether they browsed the website or not. In case of a negative response, they were asked to go back to the previous section, or they had the option to submit the response right away. Following,

questions that measured the constructs of the framework (Figure 6) and two demographic questions were asked within the survey.

To measure the construct variables, a 5-point Likert scale was used, ranging from Strongly disagree to Strongly agree, as prescribed in most of the scales that served as basis for the survey (Appendix 2 and Appendix 3).

3.4.2 Reliability and validity

Reliability refers to consistency or stability of measurement (Gidron, 2013). In the survey conducted for this study, the questions to measure the constructs were adapted in conformity with the scales previously developed in the literature. The scales were slightly changed and reformulated in terms of used language to make it easier to understand for the participants. Additionally, the reliability of the scales was checked using Cronbach’s alpha.

A measurement is considered to be reliable if its Cronbach alpha scoring is of at least 0.6 (Babin & Zikmund, 2016). All model variables were checked. A reliability check was also performed for a combined variable, formed of all three persuasive factors: scarcity, reciprocity and social proof. This was considered to be important because the PDF will be used as an independent variable for the analysis. The Cronbach alpha score for persuasive factors was 0.760.

Following, the scarcity scale adapted from Wu et.al (2011) had the Cronbach’s alpha value of 0.7. Reciprocity was measured using a 4-item scale adapted after Fatih (2014) scoring a reliability value of 0.733. Social Proof with a 3-item scale, was adapted from Wasko & Faraj (2005) and Gefen & Straub (2004) and scored 0.64.

The purchase intention scale was developed after Sweeney et al. (1999) and Zeithaml et al. (1996) and had the reliability value of 0.8. Attitude towards using the website, adapted after Moon & Kim (2001) exhibited a reliability score of 0.875. The moderating effect of need for cognition was measured using the 6-item scale elaborated by Ho & Bodoff (2014) without making any changes, therefore the external and internal reliability were consistent with the authors’ findings. A more extensive overview of the scales and the structure of the questionnaire can be seen in the Appendix 2.

Validity refers to the extent to which a research results are generalizable, can be used in different settings (Saunders et al., 2009). As the scales used for the elaboration of this survey were previously used in other studies it can stated that this study is validated.

3.4.3 Population sampling

For this thesis, the non-probability sampling (snowball sampling) approach was used to select participants, it being the least expensive and least time-consuming method according to Aczel & Sounderpandian (2006). The distribution of the surveys took place on social media platforms, ensuring that the participants are internet users.

For the 2019 year, most internet users that bought goods or services on e-commerce websites ranged between 16-74 years old (Eurostat, 2020). However, for this thesis’ research the target population consisted of people aged higher than 18 years old, to ensure the accuracy and ethics of the data.

Distributing the surveys on personal social media means that acquaintances of this thesis’ author might have responded. Therefore, it is acknowledged that there might be sampling bias in the data collection. Yet, it is difficult to measure it due to two reasons: first, the survey data was anonymous and second because of the snowball sampling approach.

3.4.4 Pilot testing

Before distributing the survey, a pilot test was made to prevent any misunderstanding or ambiguous wording. Both surveys were distributed to a limited amount of people who were asked to complete the survey and share their thoughts or give questions where necessary.

In general, there were no misunderstandings that would cause a significant effect on the purpose of the questionnaire. However, based on the existing feedback several questions were reformulated to deliver a better clarity of the questions.

3.5 Data analysis

To analyse the quantitative data obtained from the surveys, the SPSS statistical software was used. All the data was imported in SPSS, the missing values were checked, and the

reverse statements were changed. The Cronbach’s alpha was used to check the reliability of the scales (see Appendix 4A).

T-tests were executed to analyse the difference between the control survey and the one with persuasive factors (H1a). T-tests are performed to measure the difference “when there are one control and one experimental conditions and different participants are assigned to each condition” (Field, 2013). To check the relationship between the variables, correlation analysis was performed. Following, to test H1b, H2 and H3, a linear regression with the PROCESS macro (Hayes, 2017) was performed for each independent variable. In total, four regressions were run. “Regression analysis is about predicting the future (the unknown) based on data collected (the known). Performing regression analysis is the process of looking for predictors and determining how well they predict a future outcome” (Griffith, 2010).

3.6 Methodology and ethical considerations

It was ensured that the data provided by the participants will be safe and confidential, by collecting the responses online and not asking any personal details from the participants. Participants were assured that their data is collected solely for the research purpose. The transparency of the questionnaire was ensured by providing an explanation about the research before the participants started completing it.

4.

Empirical Findings and Analysis

4.1 Demographic profile of the sample

For both surveys, experimental (with PDF) and control (without PDF), a total number of 209 respondents was accumulated. However, nine responses were removed as they did not meet the required criteria of browsing the website, therefore submitting their responses without completing the entire survey. Thus, 200 responses were further analysed in this thesis – 100 responses for the experimental group and 100 responses for the control group.

For the experimental group, out of 100 participants, 59% were female, 40% were male and 1% marked “prefer not to say”. The age of the participants ranged from 18 to 65 years old as follows: 51 participants being 18-24 years old; 37 participants being 25-34 years old and 13 participants being above 35 years old.

For the control group, out of 100 participants 59% were female, 39% male and 2% marked “prefer not to say”. The age of the participants ranged from 18-65 years old for the control group as well: 43 participants being 18-24 years old, 36 being 25-34 years old, and 23 participants being older than 35 years old.

Most participants were between 18-34 years old. An explanation of the demographics data is that both surveys were distributed on social media – Facebook, Linkedin, as well as Whatsapp and reshared further by some participants within their communities. The only criteria when it came to the age was that participants should be older than 18 years old, therefore all survey participants were considered to be suitable for the analysis.

4.2 Descriptive statistics

In this section an overview of the descriptive statistics will be presented. All variables were measured on 5-point Likert scale, 5 being “Strongly Agree” and 1 “Strongly Disagree”. In Figure 7 an overview of how people answered in the experimental and control surveys is provided.

Figure 7: Descriptive statistics: mean scores for scales for experimental and control surveys

Comparing the means of each of the constructs, it can be observed that they are higher for the experimental group compared to the control group (Figure 7, Table 2). The PDF construct was measured taking the mean of all the factors that contribute to persuasion together: scarcity, reciprocity and social proof. To have a closer look at the PDF, the mean of each of the constructs is provided as well (Figure 7). Comparing the means of each of the persuasive factors, can be observed that reciprocity and social proof have the highest scores of 3.62 and 3.65, hence agreeing the most with the statements in the experimental survey. Looking at the mean of all the PDF taken together, it can be observed that it is higher in the experimental group – 3.4 compared to control group – 3.1, showing that participants in the experimental survey had a higher perception of the PDF.

A difference between the means is noted for the purchase intention construct as well. The experimental group displayed a mean of 3.2 compared to the control group with a mean of 2.9, meaning that the participants in the experimental group were more willing to buy the displayed products. However, both have a close to neutral response.

0 1 2 3 4 5 Persuasive design factors

Scarcity Reciprocity Social proof Attitude towards using the website Purchase intention Experimental Control

The smallest difference between the means can be noticed for the attitude towards using the website construct, which shows no significant difference between the participants’ attitude for both surveys.

4.3 Hypothesis testing 4.3.1 Performing t-tests

To test the first hypothesis of this thesis, H1a, a series of independent sample t-tests were performed analysing the statistically significant difference between the experimental and control groups. The results of the t-tests are shown in the Table 2.

Table 2: Descriptive statistics and independent t-tests

Experimental survey

Control group t-tests

Mean Std. dev. Mean Std. dev. Sig. (2-tailed) t df Persuasive design factors 3.48 .50917 3.11 .47478 .000* 5.406 198 Scarcity 3.22 .72917 2.94 .74814 .008* 2.680 198 Reciprocity 3.62 .70128 3.25 .67185 .000* 3.836 198 Social proof 3.65 .67412 3.14 .72214 .000* 5.163 198 Attitude 3.57 .76318 3.44 .78278 .254 1.143 198 Purchase intention 3.27 .90558 2.95 .92958 .015** 2.466 198

*significant at the .01 level (2-tailed) **significant at the .05 level (2-tailed)

As seen in the Table 2, there is a significant difference between the means of PDF of the experimental and control groups, p = 0.00. Taking a closer look upon each of the factors displayed in the surveys, it can be noted that all of them have a significant difference between the two groups, with p < 0.05.

Furthermore, the results of the t-test show that there is a significant difference between the means of the purchase intention construct, with a significant difference of p = 0.0015 (p < 0.05), therefore supporting H1a.