The effect of competitive gamification on employees’ intrinsic motivation : the role of employees' perceived, relative skill level and their need for social status

(1)

MSc Thesis

The effect of competitive gamification on employees’ intrinsic motivation.

The role of employees’ perceived, relative skill level and their need for social status.

Julia Voskuilen

10448993

University of Amsterdam

Master Research Thesis Business Administration Digital Business Track

Supervisor: Nick van der Meulen Academic year: 2016-2017 Semester 2

(2)

2 Statement of originality

This document is written by student Julia Voskuilen, who declares to take full responsibility for the contents of this document.

I declare that the text and work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3 Abstract

This study is designed to address a deficit in knowledge about competitive gamification (i.e., leaderboards) in the workplace and its effect on employees’ intrinsic job motivation. Based on previous studies, a positive effect of employees’ leaderboard ranking on their intrinsic job motivation is expected. However, the current study fills a gap in the literature by investigating two other variables that might affect the gradient of that relationship; employees’ perceived skill level in comparison to their colleagues and their overall need for social status. More specifically, it is hypothesized that their need for social status influences the moderating effect of perceived skill level on the relationship between leaderboard ranking and intrinsic job motivation. The hypotheses, including the three-way interaction, are tested by surveying employees who work within companies that apply competitive gamification and who are represented on a leaderboard. Data was collected from 132 employees covering seven organizations. Results showed that employees’ leaderboard ranking indeed positively influences their intrinsic job motivation and that this relationship is negatively moderated by their perceived skill level. However, the relationship is not found to be curvilinear for high values of perceived skill level, as was expected. Furthermore, employees’ need for social status has a significant moderating effect on the leaderboard ranking – intrinsic motivation relationship, but surprisingly, the effect is negative. The direct effect of perceived skill level on intrinsic motivation and the three-way interaction effect were not found. Implications of these findings are discussed and suggestions for future research are made. Keywords: competitive gamification, leaderboard ranking, intrinsic job motivation, perceived

(4)

4

conducting research on the topic. Although the road towards completing my Master’s thesis was all but easy, I have experienced this period as very interesting and instructive. I have been able to achieve a result I am very satisfied with. My effort put into this thesis would not have been be enough or pleasant without the guidance of my supervisor. I would like to thank him for his excellent support during this process. I also wish to thank the participating companies and respondents, without whose cooperation I would not have been able to conduct this study. And last but not least, I am thanking my family and friends who did not hesitate to help me with any difficulties and who made sure I had the necessary distraction during the process. Thereby I would like to close one chapter of my life and open up a different one, hopefully as interesting and rewarding as my time at the University of Amsterdam.

I hope you enjoy your reading.

Julia Voskuilen

(7)

7 1. Introduction

“Games are the new normal”

– Al Gore, 2011

Today, many children still consider eating broccoli to be a real problem. At no surprise, as almost 70 percent of them have a gene that makes it taste bitter. However, eating broccoli has a lot a health benefits and for that reason, parents try many things to change their kid’s mind about eating broccoli, including the famous ‘airplane landing’. Here, parents are turning the experience into a game that changes children’s’ perception about eating broccoli where bitterness is no longer perceived bad (Zichermann, & Cunningham, 2011). This phenomenon is called gamification: “The use of game-elements to drive behavior in non-game

environments” (Deterding, Dixon, Khaled, & Nacke, 2011). The broccoli example is a simple but pure form of gamification, but it can be applied to many other contexts as well.

The overall application of gamification is growing rapidly (Webb, 2013). It has become one of the major human-computer interface trends of the 21st_{century (Rauch, 2013).} The total gamification industry is expected to jump from $2.88 billion in 2016 to $5.5 billion in 2018. Within this industry, enterprise gamification is now the largest segment, consisting of approximately 25 percent (Rauch, 2013). This heightened interest is the result of three

developments; the growth of the computer game industry, the unattainable amount of data available within firms and the continuous search for new ways to influence the behaviors of both customers and employees. Especially the use of gamification among employees receives much attention lately. According to a study by Blessing White (2013), 69 percent of European workers are not engaged or even actively disengaged in their work. Firms therefore try to understand the fun and engaging elements of a game and transfer this to the business context. They increasingly use game design to make the affective experience of work more positive

(8)

8 for employees and believe it will drive positive organizational outcomes such as

organizational commitment and job performance (Herzig, Ameling, & Schill, 2012).

To create a positive work experience, most companies simply add game elements such as leaderboards to the workplace (Sutter, 2010; Nicholson, 2015). A leaderboard is not only easy and affordable to implement, but can also drive employees’ motivation as humans are by nature a competitive race (Duncan, Humphreys, & Ward, 1997). Any game or game design contains some sort of competition and thus explains why people love to play games.

Competitive gamification can thus be seen as a powerful tool to motivate employees at work (Deterding, et al., 2011).

Competitive gamification, however, can be a double-edged sword regarding

motivation as it has both a controlling and an informational aspect (Deci, 1972). When people compete with the ultimate goal to win the competition, the controlling aspect becomes more salient and will (like other extrinsic rewards) decrease people’s intrinsic motivation. The informational aspect, on the other hand, increases intrinsic motivation. Here, competition is viewed as a challenge that provides people opportunities for competence feedback (Deci, Betley, Kahle, Abrams, & Porac, 1981). Such intrinsic motivators are more effective in the long-term. Only when intrinsically motivated, employees are performing tasks because they see them as inherently enjoyable, interesting and satisfying. They will realize the value of their new learned behavior and continue using it (Pink, 2011). Firms increasingly recognize this and try to intrinsically motivate the workforce through the use of leaderboards.

There are, however, opposing views about the use of a leaderboard within an organization. Some scholars contest its effectiveness on employees’ intrinsic motivation (Song, Kim Tenzek, & Lee, 2013; Nicholson, 2013; Hanus, & Fox, 2015; Dominguez, et al., 2013). It can encourage some type of people to do better, while others are discouraged within the competitive environment (Nicholson, 2015). A leaderboard can, for instance, inspire those

(9)

9 at the top as they want to maintain their position while those at the bottom are discouraged (Nicholson, 2013). Scheepers (2009) found that the level of motivation depends on the

stability of such inter-group ranking. Members with a low position on the leaderboard seem to be threatened by a ranking that is stable but this threat turns into a challenge when the ranking is unstable. A ranking may thus be unstable when people have the ability to climb up (or down) the leaderboard. This only seems possible when one’s skills or capabilities improve (or deteriorate) relative to others. Motivation, therefore, likely depends on one’s skill perceptions. A leaderboard is seen as a status item. However, not all people within a social milieu are jockeying for status (Zichermann, & Cunningham, 2011). It is perceived important only to those individuals that seek recognition within a social setting (e.g., their workplace). They see competition not inherent in the availability of positive outcomes, but by the evaluation that is attached to the outcomes of others (Turner, 1975). If they fail to maintain a positive

distinctiveness, their motivation likely declines.

It is thus expected that a certain position on the leaderboard affects employees’ intrinsic motivation and that this relation is moderated by their perceived skill level. The gradient of that relationship might in turn be determined by their need for social status. No study has yet looked into this three-way interaction effect on intrinsic motivation. More specifically, no study has yet investigated the moderating effect of employees’ perceived skill level on the leaderboard ranking – intrinsic motivation relationship and, in turn, if their need for social status influences this effect. To fill this gap in the literature, the present study tries to answer the following question: “How does employees’ perceived skill level and their need for social status affect the relationship between leaderboard ranking and their intrinsic motivation?” The answer to this question contributes to the overall understanding of the gamification effect that benefits both the organization and its employees as it provides a step towards a more positive and engaged workplace.

(10)

10 2. Literature Review

In this chapter, the topic will be defined more precisely using previous research. First, the literature concerning gamification will be discussed with the focus on competition.

Competitive gamification is then further specified to the use of leaderboards in the workplace. Here, both positive and negative views on this topic are given. The second part of this chapter explores motivation or more specific, employees’ intrinsic motivation and how this is affected by competitive gamification. The third section will discuss the perceived skill level of

employees in comparison to their colleagues, their need for social status and explains how these two constructs fit within the conceptual model of this study. Throughout this chapter, the hypotheses are formulated as well. Finally, a summary is provided that emphasizes the need for this research.

2.1 The emergence of gamification

According to Newman (2004), games are inherently fun and not serious. Combining fun and seriousness, however, is increasingly evident in both academia and industry. Games and game technologies have transcended the traditional frontiers of their medium and are now applied to non-game contexts (Deterding, Dixon, Khaled, & Nacke, 2011). This phenomenon is called gamification and Deterding, et al. (2011) propose the following definition: “The use of game design elements in non-game contexts”. While they emphasize that the properties or qualities implemented in gamification have to be similar to those used in games, Huotari and Hamari (2012) state that the goal of gamification is more important. It is about the psychological experiences rather than the method. They define gamification as “a process of enhancing a service with affordances for game full experiences in order to support user’s overall value creation”. It entails the same structure, look and feel of a game in order to create the same experience when playing the game. Although there is no single widely accepted definition of gamification, the logic behind it is clear and simple. People like to play at least some sort of

(11)

11 game and introducing game mechanisms into activities generally disliked (e.g., exercising), will make those activities more fun to perform (Kapp, 2012). Therefore, an increasing number of companies are using gamification to engage with the customer and stir them in favorable directions.

One well-known example is Foursquare (www.foursquare.com). Foursquare is a location-based social network where people check-in through the application and share their location with others. As opposed to Dodgeball, its less successful predecessor, Foursquare uses game elements such as points, badges, achievements and status to engage the crowd. People can, for instance, earn a Gym Rat badge when they go to the gym at least ten times a month. Its application is seen as the purest form of gamification and has proven to be successful. It now has a community of approximately 60 million people while Dodgeball failed to keep its users engaged and was forced to shut down in 2009.

This is one example of the many organizations that use game elements to better interact with their customers to improve customer loyalty (Frith, 2013). However,

gamification is not restricted to one specific area. It can be applied to several contexts such as education (Fitz-Walter, Tjondronegoro, & Wyeth, 2011), government (Coronado Escobar, & Vasquez Urriago, 2014), healthcare (Hamari, & Koivisto, 2013), sustainability (Gustafsson, & Bang, 2009), transportation (Vieira, et al., 2012) and work (Oprescu, Jonges, & Katsikitis, 2014).

Whatever context used, a game design can be classified into three game elements: dynamics, mechanics and components (Werbach, & Hunter, 2012). According to the authors, dynamics are the themes around which the game revolves and show what the underlying forces are. These are the big-picture aspects that never directly enter into a game. Mechanics are more specified as they imply specific action. Those elements can stir players’ behavior into certain directions. Components are even more concrete and are used in the interface of

(12)

12 the game. It is what the player sees. The elements are organized in a decreasing order of abstraction with dynamics at the highest level of thinking, followed by mechanics and

components. Designers, either of games or gamification designs, should start with deciding on its core dynamics and based on this decision, the mechanics and components are chosen. Figure 1 shows the three elements within the framework and the existing forms of each.

Figure (1). Pyramid of game elements (Werbach, & Hunter, 2012).

2.1.1 Gamification in an organizational context

Games have long been played by employees to alleviate the deleterious aspects of work and thus emerged spontaneously from employees themselves. Gamification, however, has transformed games from employee-generated to managerially-imposed play (Mollick, & Rothbard, 2014). Although Burawoy (1979) did not use the exact word ‘gamification’ in his study, he showed that games were no longer seen as a respite to workers and a time-wasting resistance to management. In contrast, games could provide benefits to management as it allows them to manipulate workers into consenting to their role in the workplace. According to Burawoy, employees believe to engage in the rule of the game but actually they engage in the rules of the production environment itself. Workplace gamification can therefore be defined as “an employer-imposed game in a work environment where the goals of the game

(13)

13

are designed to reinforce the goals and purpose of the employers” (Mollick, & Rothbard, 2014).

Almost 70 percent of the European workforce is not engaged or even disengaged in their work (Blessing White, 2013). Firms are therefore continually looking for innovating ways to influence and improve the behaviors of employees and they increasingly recognize the importance of fun in organizational activities to keep them engaged (Liu, et al., 2013). Several studies point out that gamification is an innovative solution to address organizational issues such as high levels of stress (Perryer, et al., 2012), reduced loyalty (Dorling and McCaffery, 2012) and reduced social capital (Zhy, et al., 2013). Studies also found that gamification has the power to enhance employees’ engagement and productive collaboration (Deterding, et al., 2011), their motivation (Koutropoulos, 2012; Dominguez et al., 2013) and their personal control (Rednic, et al., 2013). Xerox, for example, uses gamification to train its managers to collaboratively complete online quests and Salesforce is now employing

challenges and leaderboards to increase sales (Oprescu, Jonges, & Katsikitis, 2014).

Companies thus increasingly apply selected principles of game design and game interaction to transform its organizational processes into a game-like experience for its employees (Oprescu, et al., 2014).

2.1.2 Competitive gamification through the use of leaderboards.

Nowadays, most companies simply add game elements in the form of points, levels, badges and leaderboards to their organizational processes and they believe this will improve

productivity and motivation (Nicholson, 2015). Here, points are converted into levels or badges which may be used in a leaderboard to encourage competition among employees. As described in the previous section and showed in Table 1, each game component can be linked to a game mechanism. A leaderboard is associated with competition as it is used to display the results of a competition. This taps into people’s natural competitiveness and encourage them

(14)

14 to do better. It might motivate them by the desire to improve their position on the leaderboard (Natvig, et al., 2004).

Game mechanics Game components

Rewards Points

Status Levels

Achievement Challenges

Self-expression Virtual goods and spaces Competition Leaderboards

Altruism Gifts and charity

Table (1). The link between game mechanics and components (Werbach, & Hunter, 2012).

It is not weird to think of gamification as competition or leaderboards. According to Natvig and his colleagues (2004), competition, either against one’s self or others, is the driving force behind the success of gamification. A key part of games is having dynamic challenges and competition among players is an important source of such challenges (Liu, & Li, 2013). One early proponent of adding competition to the workplace is Lenin (Nelson, 2012). His theory of ‘socialist competition’ has goals that are similar to the gamification-of-work movement we see today. Here, workers, groups of workers or factories competed against each other to motivate greater production. A more recent study of Mekler, Brühlmann, Opwis and Tuch (2013) also found the positive effect of points, levels and leaderboards on individuals. According to their study, in which participants had to perform an image annotation task, competitive elements boost performance. Image annotation tasks, however, are usually performed by people for fun and are, according to the authors, to a lesser extent associated with non-game contexts such as a workplace. Moreover, the study of Sepehr and Head (2013) found that competitive gamification can help students engage in learning through the

immediacy of feedback. It shows that competition can increase intrinsic motivation due to the informational aspect of competition (Deci, et al., 1981).

(15)

15 studies that describe these, however, claim that the outcome of competitive gamification is dependent on the context or one’s individual characteristics. The study of Hanus and Fox (2015), for instance, suggests that competition components do not improve educational outcomes and at worst even harm students’ results. Students who are initially interested in the study material are harmed by those gamification efforts, while those who are bored in class are instead motivated by competition as it provides them a form of distraction. Song, Kim, Tenzek and Lee (2013) conducted a study to see whether individual differences play a role in people’s motivation when a competitive game design is applied. They distinguish between competitive and non-competitive individuals and found that highly competitive individuals indeed have more favorable experiences in the competitive setting compared to a non-competitive setting and that competition factors do not benefit less non-competitive individuals. Furthermore, Sepehr and Head’s study (2013) as well found that students who did not win the competition were less motivated and report less positive affect to the system. It might erode one’s intrinsic motivation due to the controlling aspect of competition (Deci, et al., 1981). Related to the findings of Sepehr and Head (2013) is the study of Nicholson (2013). He found that competition in the form of leaderboards can inspire those who acquired a high rank to stay at the top, while that same leaderboard can be demotivating to those at the bottom. He conducted a research in a classroom that showed that most students who ranked low had given up on doing certain class assignments as they felt they could not ‘beat’ the top performers. Based on the findings of these two studies, it is expected that:

H1: Employees’ leaderboard ranking positively affects their intrinsic motivation. Thus, the outcome of competitive gamification is not always positive and can differ from one individual to another. Some argue that it enhances intrinsic motivation, while others found that it can also discourage people. To better understand its effect, it is important to fully understand the concept motivation and to know the difference between intrinsic and extrinsic

(16)

16 motivators. Only then, assumptions about the effect of competitive gamification on

motivation can be made.

2.2 Employees’ motivation to work 2.2.1 Extrinsic versus intrinsic motivators

Maslow’s hierarchy of needs (1943) is a theory that describes what people need in life and how much they need it. The hierarchy consists of five levels. The first four levels are considered physiological needs and the fifth level is considered growth needs. The lower-order needs must be met before an individual desires higher-lower-order needs. Within this hierarchy, tangible and intangible needs are combined simultaneously. However, once a certain threshold is met, meeting intangible desires can lead to higher levels of satisfaction and performance than tangible goods do (Maslow, 1943). His hierarchy relates to the

framework SAPS: Status, Access, Power and Stuff of Zichermann and Cunningham (2011). Their framework is a system of rewards. Status is seen as the highest form, which is harder to implement than the other motivators but results in higher levels of stickiness (engagement) and reduced costs to the organization. Each form is explained below.

Status: the relative position of a person to others (e.g., leaderboards).

Access: the number of privileges (e.g., early access on sale for loyal customers). Power: the amount of control over others (e.g., moderator on platform).

Stuff: the tangible rewards (e.g., money). Stuff is least effective as the incentive to contribute erodes once the reward is given. People thus value intangible goods (e.g., status, access or power) over tangible stuff. A similar distinction is made by Ryan and Deci (2000). They differentiate between extrinsic and intrinsic motivation. Extrinsic motivation refers to doing something because it leads to a separable outcome. People are motivated because they get a reward in the form of goods (e.g., money). Intrinsic motivation, on the other hand, refers

(17)

17 to doing something because it is inherently enjoyable, interesting and satisfying. This

motivation comes from inside an individual as they actually want to do something without receiving a tangible reward (Pink, 2011). Table 2 shows some examples of both extrinsic and intrinsic motivators.

Table (2). Extrinsic versus Intrinsic rewards (Dale, 2014).

Over three decades of research showed that performance and the quality of experiences can differ when one is behaving for extrinsic versus intrinsic reasons. It is found that extrinsic forms of rewards can complement intrinsic motivators (Hennessey, Moran, Altringer, & Amabile, 2005; Amabile, 1993). An extrinsic reward might be an effective tool to motivate individuals until intrinsic motivation evolves (Rimm, 1986). However, the predominant psychological view suggests that extrinsic rewards work in opposition to intrinsic rewards (Deci 1972; Deci & Ryan 1985; Lepper & Greene 1978). Generally, these theorists propose that when extrinsic motivators are (too) strong, they will crowd out intrinsic motivation as it shifts people’s attention away from the task and cause the behavior to be perceived as

externally controlled (Deci, & Ryan, 1985). Extrinsic motivators are also less effective in the long-term as those rewards will erode after a certain threshold is reached (Pink, 2011). Intrinsic motivation has been linked to higher performance, longer-lasting learning, and perseverance (Hennessey, Moran, Altringer, & Amabile, 2005). People who are intrinsically motivated are more engaged and retain information better (Hanus, & Fox, 2015; Ryan, &

(18)

18 Deci, 2000).

Game elements are often perceived as extrinsic motivators. However, they must be seen as a tool to promote intrinsic motivation. Intrinsic motivation requires an activity to be engaging and fun. If a certain activity is not, gamification will acknowledge that and then seeks to hide the non-engaging nature of the activity by integrating a game. Game design can thus be referred to as ‘chocolate covered broccoli’. An example is the study of Burawoy (1979), in which workers who produced over quota received a financial reward. It was, however, not the money incentive but the game itself that became the focus of workers. Moreover, the self-determination theory of Ryan and Deci (2000) states that three

psychological human needs – autonomy, competence and relatedness - need to be satisfied in order to increase the intrinsic motivation for a task. Competitive gamification stimulates the feeling of being autonomous. In a competitive environment, people try to influence the outcome of the competition and know that their own behavior and actions might have an impact. When people are able to produce a desired outcome (e.g., improve their position), it results in the perception of competence. People thereby experience mastery and effectiveness. Furthermore, the authors describe relatedness as the feeling of being connected to others. Competitive gamification connects people through competition. Competition causes people to be continuously challenged by others. Thereby, they continuously engage and interact with their colleagues. Game elements thus have the potential to increase intrinsic motivation for a certain task. The following section will further describe intrinsic motivation and its

importance to the workplace.

2.2.2 Employees intrinsic job motivation

An organization depends on human resources. Therefore, motivating the workforce is a major concern for organizations. Especially with the emergence of the knowledge economy,

(19)

19 fostering intrinsic motivation became important (Hennessey, Moran, Altringer, & Amabile, 2005). It means building a community of employees who have positive emotions about and a favorable view of the company, their job and their colleagues. Intrinsic motivation is seen as an important instrument in inducing employees to produce efficiently and effectively and to create a positive work environment where they commit to their duties and do their jobs seriously and joyfully (Azar, & Shafighi, 2013). According to Warr, Cook and Wall (1979), intrinsic job motivation can be defined as “the degree to which a job holder is motivated to perform well because of some subjective rewards or feelings that he expects to receive or experience as a result of performing well”.

There are two dominant behavioral theories about intrinsic motivation. The operant theory of Skinner (1963) states that behaviors are motivated by rewards. Intrinsic motivation is the outcome for those that perceive the activity itself as a reward. An activity is perceived as a reward when it has characteristics that make it interesting. The learning theory of Hull (1943) asserted that behaviors are motivated by physiological drives. Intrinsic motivation derives from activities that satisfy those physiological needs. People have a need to learn and explore, but they must be challenged within certain limits. If a task is too challenging, it might harm people’s competence perception and thereby their motivation (Ryan, & Deci, 1985). An interesting task, but also a match between the task difficulty and employees’ competence, are thus determinants of the level of their intrinsic job motivation. The relevance of the match is discussed in further detail in the next section to emphasize the importance of employees’ perceived skill level.

2.3 Employees’ perceived skill level relative to their colleagues

The match between task difficulty and skill level is the outcome of Csikszentmihalyi’s flow theory (1990). This theory describes the flow as follow: “A state of deep absorption in an

(20)

20

activity that is intrinsically enjoyable”. This optimal flow is reached when there is a

symbiotic relationship between the challenge and the skills needed to meet such a challenge. A positive experience occurs when one’s skills are neither overmatched nor underutilized. A negative experience occurs when the balance is disrupted. If so, feelings of relaxation (low challenge, high skills), anxiety (high challenge, low skills) or apathy (low challenge, low skills) are likely to be experienced (Csikszentmihalyi, 1997). The different states are shown in Figure 2. Within the optimal flow zone, performance is perceived as pleasant and successful and the job is worth doing it for its own sake, even if no further goal or extrinsic reward is reached (Deci, 1972). The experience itself becomes the reward. Csikszentmihalyi (1997) suggests that issuing appropriate challenges or enhancing one’s skills is the ideal way to stir employees into the ‘flow’ to engage them with their jobs.

Figure (2). Csikszentmihalyi’s Flow (1990).

The study of Csikzentmihalyi and LeFevre (1989) found that employees who perceive their skill level to be equivalent to task difficulty experience a positive mood. The perception of competence has been clearly identified as a determinant of a person’s level of intrinsic motivation (Deci, & Ryan, 1985). This is also supported by a study of Harackiewicz (1979). Within the study, she gave participants normative data such as ‘you did better than the

(21)

21 average person on these puzzles’. It gave individuals the perception of competence and thus enhanced their intrinsic motivation to continue. Individuals who perceived themselves as competent said to have enjoyed solving the puzzle, while those who did not receive the message (or the opposite message) felt less competent and expressed less enjoyment in solving the puzzle (Harackiewicz, 1979). One’s mood and motivation thus depends on the perception of skill level. Therefore, the second hypothesis is:

H2: Employees perceived skill level has a direct positive effect on their intrinsic

motivation.

Competition can be seen as the challenge and a leaderboard is the visual representation of its outcome (Liu, & Li, 2013). To meet the challenge and obtain a high rank, skills are needed. A distinction can be made between perceived and actual skill level. An employee may think his skill level is insufficient, but actually has an adequate skill set (in comparison to his

colleagues). This person believes he is unable to acquire a high position on the leaderboard, but accomplish to do so. The rise in the ranking was thereby not expected, which results in higher levels of intrinsic motivation compared to someone who has high skill perceptions and thus expected to acquire a high position. This is explained by a study of Barankay (2011). The response to ranking feedback may be different depending on the expectation. Whether that feedback conveys ‘good’ or ‘bad’ news is dependent on the expected and the realized ranking. An employee with a low skill perception does not expect a high rank. If he acquires a high position on the leaderboard, this is regarded as ‘good’ news and as a response, his motivation level is high. If that person, however, has high skill perceptions, a high position on the

leaderboard was expected and he will thereby be less motivated by it. The news he receives is ‘less’ good than for someone with low perceptions. Taken together, this study expects that: H3a: The relationship between leaderboard ranking and employees’ intrinsic

(22)

22

stronger for lower rather than higher values of perceived skill level.

When higher values of perceived skill level are considered, this effect can also be linked to the flow theory of Csikzentmihalyi (1990). Employees with high perceptions about their skill level are more likely to end up in a stage of relaxation. They believe their skill level is higher than that of their colleagues and a higher ranking on the leaderboard confirms their

confidence. Employees who are in a stage of relaxation are not challenged at work. This will eventually result in boredom and thereby in less motivated employees (Csikzentmihalyi, 1990). It is therefore expected that:

H3b: The relationship between leaderboard ranking and employees’ intrinsic

motivation is moderated by their perceived skill level and this relationship is curvilinear (i.e., an inverted u-shape) for high values of perceived skill

level.

2.4 The need for social status

The strive for status has long been recognized. Although status is no longer critical to human survival, people still place a high priority on social status (Tran, & Zeckhauser, 2012).

Research in sociology and biology showed that status seeking and competitive preferences are ‘evolutionarily’ rooted in human behavior (Spence, 1974). Status is used as a signal to others and a way to obtain resources (Lin, 1994) and power (Thye, 2000). Individuals who possess status and send the signal are, generally, of higher quality than those who lack it (Spence, 1974). A college degree is an example of such a signal. It is easier for higher-quality individuals to obtain a college degree than it is for lower-quality individuals. It is an

indication of future productivity in the workplace, even if the content of the college degree is unrelated to the local skill requirements of the job. Status can thus be seen as a sign of competence. But status may not only be a means to an end. In addition to the pursuit of resources and power, it became an end in itself. Today, people increasingly strive for social

(23)

23 status and see it as an intangible reward. They prefer it above access, power and stuff

(Zichermann, & Cunningham, 2011). This is evident from an experiment performed by Van den Bos and his colleagues (2013). During an auction, people knowingly incurred financial losses to win the auctioned good. They were willing to trade of material benefits for social gains. It is an affective response to social comparisons and shows that people value status independently of extrinsic factors. Their behavior is influenced by the utility ascribed to being either the winner or a loser. People intensely care about what others think of them and spend much time and energy tracking their relative status in a group (Kaufman,2011). These status seekers undertake actions necessary to maintain or improve their position within a social setting (De Carvalho, & Neumann, 2014). Status seeking is a hierarchized activity and often visible in a competitive environment where people are informed about their relative

performance. Within such an environment, people put forth additional effort to accomplish great things (Kaufman,2011). However, it might also cause people to behave disreputable in order to improve their position (Charness, Masclet, & Villeval, 2013). Especially ranking incentives at work, through the use of leaderboards, should be applied with caution. People associate a high ranking on the leaderboard with high status and peer recognition. A

leaderboard can thus be seen as a status item. However, not all employees can acquire a high position on a single leaderboard. As a result, the risk of in-group rivalry and sabotage among employees increases. They will try to maximize their payoffs relative to others. However, the degree of rivalry and the level of unethical behavior depends on the motive behind the

competitive behavior. If people try to increase their position solely because they enjoy outperforming others, in-group rivalry will exist (Dohmen, et al., 2011; Chowdhury, & Sheremeta, 2012). It results in rank-ordered relationships among employees (Ridgeway, & Walker, 1995). But status seeking can also be related to self-image (Benabou, & Tirole, 2006; Köszegi, 2006). People might increase their effort in order to rise in the ranking as a way to

(24)

24 improve their self-image (Barankay, 2011). Not everyone within a competitive environment has the same motive to compete.

Moreover, not everyone has the same desire for social status. The pure taste for having the best position in the performance distribution differs among individuals (Zichermann, & Cunningham, 2011). While some individuals perceive their outcome as either ‘good or bad’, others view it as ‘better or worse’. The latter group likes to compare the outcomes to others (Turner, 1975). Competition here is not inherent in the availability of positive outcomes (good or bad) but is created by the evaluation that is attached to outcomes of others (better or

worse). People who are jockeying for social status but fail to maintain a positive

distinctiveness are likely demotivated by this. Those without or with less desire for social status, will not or less likely be influenced by a shift in their leaderboard position. Therefore, this study expects that:

H4: The relationship between employees’ leaderboard ranking and their intrinsic

motivation is moderated by their need for social status, so that the relationship is stronger for higher rather than for lower values of need for social status.

So far, it is expected that employees’ perceived skill level and their need for social status both have a moderating effect on the relationship between leaderboard ranking and intrinsic

motivation. The effect of the former variable is relatively moderate. Someone with low skill perceptions is motivated by the fact that his position on the leaderboard increases. However, only if this person has a (high) need for social status and thus values a high ranking on the leaderboard, he will put forth additional effort to maintain that position. It will further increase his motivation. So, when employees have a need for social status, the moderating effect of perceived skill level on the main relationship between leaderboard ranking and intrinsic motivation will be significantly stronger. The last hypothesis is therefore as follow:

(25)

25

relationship between leaderboard ranking and intrinsic motivation is, in turn,

positively affected by the their need for social status, so that a) the relationship

between leaderboard ranking and intrinsic motivation is stronger for employees with a

low perceived skill level and a high need for social status than for employees with a

low perceived skill level and a low need for social status; b) the relationship between

leaderboard ranking and intrinsic motivation is stronger for employees with a high

perceived skill level and a high need for social status rather than for employees with a high perceived skill level and a low need for social status.

2.5 The conceptual model

To make the relationships between the variables clearer, the hypotheses are visualized in the following conceptual model (see Figure 3).

Figure (3). The Conceptual Model.

2.6 An overview of hypotheses

H1: Employees’ leaderboard ranking positively affects their intrinsic motivation. H2: Employees perceived skill level has a direct positive effect on their intrinsic

motivation

(26)

26

motivation is moderated by their perceived skill level, so that the relationship is

stronger for lower rather than higher values of perceived skill level.

H3b: The relationship between leaderboard ranking and employees’ intrinsic

motivation is moderated by their perceived skill level and this relationship is curvilinear (i.e., an inverted u-shape) for high values of perceived skill

level.

H4: The relationship between employees’ leaderboard ranking and their intrinsic

motivation is moderated by their need for social status, so that the relationship is

stronger for higher rather than for lower values of need for social status.

H5: The negative moderating effect of employees’ perceived skill level on the

relationship between leaderboard ranking and intrinsic motivation is, in turn,

positively affected by the their need for social status, so that a) the relationship

between leaderboard ranking and intrinsic motivation is stronger for employees with a

low perceived skill level and a high need for social status than for employees with a

low perceived skill level and a low need for social status; b) the relationship between

leaderboard ranking and intrinsic motivation is stronger for employees with a high

perceived skill level and a high need for social status rather than for employees with a high perceived skill level and a low need for social status.

2.7 Conclusion

Competitive gamification is increasingly used by organizations to engage the workforce. They add points, levels, badges and leaderboards to their organizational processes and believe this will benefit their employees and thus the organization itself. However, opposing views exist on the effectiveness of competitive gamification. Previous research shows that it can both enhance and reduce employees’ intrinsic motivation. It might be demotivating for those who acquire a low position on the leaderboard while those at the top are motivated to stay there.

(27)

27 The stability of such a ranking is dependent on the skill distribution of the workforce. It determines if people are able to climb up or down a leaderboard. Their intrinsic motivation therefore depends on how people perceive their own skills in relation to others. If an employee believes he has an adequate skill set and thus believes he is able to improve his position on the leaderboard, his motivation increases. However, even those individuals who think their skill level is insufficient, might have the right skills to increase their position on a leaderboard. As they did not expect the rise, their motivation will increase even further. Although striving towards the highest position is ‘evolutionary’ rooted in human behavior, the degree to which someone is seeking status differs. Only those individuals with a desire for recognition and social status will be motivated by an increased ranking. The relationship between leaderboard ranking and intrinsic motivation is thus moderated by both employees’ need for social status and their perceived skill level. The gradient of the latter effect is, in turn, also affected by the need for social status. An employee might not expect a rise in the ranking due to low skill perceptions and will be motivated when it happens. But only if he finds his status important, a rise in the ranking will further strengthen his intrinsic motivation.

However, no study has yet investigated this three-way interaction. More specifically, no study looked into the moderating effects of both employees’ perceived skill level and their need for social status on the relationship between leaderboard ranking and intrinsic motivation and how the need for social status moderates the moderating effect of perceived skill level on the main relationship. This study will therefore further investigate these effects, by which it contributes to the overall understanding of the gamification effect that benefits both the organization and its employees as it provides a step towards a more positive and engaged workplace.

(28)

28 3. Methodology

The previous chapter discussed the existing literature, identified the research gap and

formulated the associated research question and hypotheses. This chapter provides a detailed description of the research methodology through which the hypotheses will be tested. First, the research philosophy, approach and design will be discussed. This is followed by an explanation of the sample and how the data was collected. Next, the measurements used for each variable and the final structure of the survey are explained. The statistical procedure is considered in the last section.

3.1 Research philosophy and approach

The chosen research philosophy determines the used research strategy and the method associated with that strategy (Saunders, Lewis, & Thornhill, 2012). This research reflects the principles of positivism as it works with observable and measurable variables (e.g., the experiences faced by employees that work for organizations that apply competitive

gamification). It proposes a theory that can be tested and refined until it accurately predicts reality (Saunders, et al., 2012). Furthermore, a deductive approach is applied. The hypotheses are derived from the existing literature and through the implementation of the right

methodology, the study tests if these hypotheses are supported or not (Saunders, et al., 2012).

3.2 Research design

The positivism philosophy is often associated with a structured methodology in order to facilitate replication (Gill, & Johnson, 2002). For this research, a questionnaire-based survey is used to collect the data needed to test the propositions. It is a fast and affordable way to collect data from a sizeable population (Saunders, 2009). Quantitative values are attached to the used variables. This allows the variables to be subject to statistical tests (e.g., regression) in order to determine the moderation as depicted by the hypotheses. Moreover, the survey

(29)

29 includes standardized questions that increase the reliability when comparing the answers of the respondents. Given the presence of the time frame to conduct this research, data is only collected at one point in time (cross-sectional) rather than at multiple time points

(longitudinal) (Saunders, et al., 2012). Furthermore, it was intended to use a mixed-method study. However, not all companies were willing to directly send information about their employees’ leaderboard position to the researcher. Therefore, only one method – a self-administered questionnaire - is used to measure all variables. Self-reporting saves time for both the researcher and the respondents and, more importantly, guarantees anonymity. Anonymity will reduce subject or participants bias and therefore improves the reliability of the data collected (Saunders, et al., 2012).

There are, of course, some limitations considering this methodology. Due to the cross-sectional design, the internal validity is lower in comparison and no statements can be made about causality. Also, measures might be affected by common method bias due to the use of a self-reporting questionnaire. Standardization of the questions increases the risk of

misinterpretation. When respondents interpret questions differently, it might result in invalid outcomes (Saunders, et al., 2012). Last, using a questionnaire provides less detailed

information and makes it more difficult to understand why certain relationships were found. Designing the questionnaire was therefore a conscious and well-thought process.

3.3 Sample and data collection

For this research, firms that apply competitive gamification are targeted to survey employees who are presented on a leaderboard. As this population is large and the sampling frame unknown, the research is conducted using a non-probability sample (Saunders, & Lewis, 2012). The aim was purposive sampling, but convenience sampling predominates as most companies were difficult to access. The survey is thus mainly distributed via the author’s personal network. In addition, the snowball technique is used. If a company, either

(30)

30 participating in the study or not, knew another company that uses competitive gamification, contact information was exchanged. The link to the online survey was then sent to the participants through e-mail, either directly by the researcher or via the contact person of the company. To increase the number of respondents, the researcher visited some of the targeted companies as well. Employees used the researcher’s personally owned tablet to fill out the survey.

To obtain answers from employees who rank low as well as high on the leaderboard, multiple employees within a single organization or team were surveyed. Furthermore, it is important to collect data from employees who perceive their skills as either above and below group average and from those who have a low but also a high need for social status. However, it is difficult to predict people’s perceptions. To assure this distinctiveness, an attempt is made by surveying people with different job levels and different work experiences (e.g.,

organizational tenure). Moreover, different types of companies are approached to cover multiple markets and industries and thus different types of employees. A total of eighteen companies were targeted. Table 3 shows some basic information about the seven companies that eventually cooperated. The organizations are operational for more than 12 years, which indicate that no start-ups are included. Respondents who participated in this study mainly work within departments related to sales or customer services as competitive gamification (i.e., a leaderboard) is mainly used in these sections of the organization.

(31)

31 The total number of employees who started the survey was 150. However, five respondents did not completely finish the questionnaire. These cases were excluded from the sample. Subsequently, another 13 responses were eliminated as some had indicated the same leaderboard ranking as a colleague. In that situation, both (or all three) respondents were removed from the sample. The final sample therefore covers 132 respondents. 58% of the respondents are male and the average age is 28 years old. Moreover, 29.8% of the respondents reported an organizational tenure less than a year, 55.2% an organizational tenure between 1 and 5 years and 15% of the respondents work at their current employer longer than 5 years. The minimum and maximum organizational tenure in months is 3 and 144 respectively. The average tenure is 33,5 months. On average, respondents thus work relatively short for their current employer. Furthermore, only 9 out of the 132 respondents requested the English version of the questionnaire, but no further assumptions can be made regarding their nationality.

3.4 Measures

Variables that are central to this research include: employees’ leaderboard ranking, their intrinsic motivation, perceived skill level and need for social status. The applied measures and scales are adopted from original scales found in peer-reviewed journals, with the exception of leaderboard ranking. Only scales with a Cronbach’s Alpha exceeding 0.70 are used as it proves internal consistency and thus reliability (Cronbach, 1951). The variables, including their operational definitions, are discussed below.

3.4.1. Leaderboard ranking

Employees’ leaderboard ranking represents the independent variable. A leaderboard signifies rank among people or, for this study, among employees. As said before, the intention was to collect the ranking data directly from the company. However, not all companies were willing to send the ranking information to the researcher due to privacy considerations. Therefore, a

(32)

32 question about employees’ leaderboard ranking was included in the survey. As the

participating organizations differ in the number of employees presented on their leaderboard, the survey needed to be adjusted for each. This resulted in multiple surveys that are combined afterwards in order to analyze the data. The remaining part of the survey was similar for each organization to assure consistency and thus reliability.

With the help of a slider, employees indicated their position on the leaderboard. It was emphasized that they could pause the survey in order to look up their precise position on the leaderboard to correctly answer this question.

3.4.2 Intrinsic job motivation

The dependent variable - intrinsic job motivation - is measured through a well-validated scale by Warr, Cook and Wall (1979). It consists of six items, including: “I take pride in doing my job as well as I can” and “I feel a great sense of personal satisfaction when I do my job well”. The items are scored on a seven-point Likert-scale ranging from ‘strongly disagree’ (1) to ‘strongly agree’ (7). In their research, the items indicated a good reliability (Cronbach’s α= .84). All six items are used in the current study. In addition, their definition of intrinsic job motivation is used in this study as well: “The degree to which a job holder is motivated to perform well because of some subjective rewards or feelings that he expects to receive or experience as a result of performing well”.

3.4.3 Perceived, relative skill level

It is expected that employees’ perceived skill level in relation to their colleagues directly influences their intrinsic motivation and that this variables moderates the relationship between employees’ leaderboard ranking and intrinsic job motivation. To measure this construct, a distinction is made between communication, problem solving, collaboration, technical and creativity skills. This distinction is based on the study of Carnevale (1990). For each skill, the respondents had to indicate on a scale from ‘much worse’ (1) to ‘much better’ (5) how they

(33)

33 perceive their skills in comparison to their colleagues. Although these five skills are regarded as basic job skills, it is possible that someone does not use a particular skill. Therefore, the option ‘not applicable’ is added.

3.4.4 The need for social status

The second moderator is employees’ need for social status. The need for social status is measured through a scale developed by Flynn, Reagans, Amanatullah and Ames (2006), consisting of eight items. An example is “Being a highly valued member of my social group is important to me”. Two items are counter-indicative. They are reverse scored to eliminate response-set bias (Saunders, et al., 2012). Participants were asked to rate the extent to which they agreed with each item on a seven-point Likert scale, ranging from ‘strongly disagree’ (1) to ‘strongly agree’ (7). The Cronbach’s alpha for this construct in their study is 0.82.

3.4.5 Control variables

There are some variables that are not directly integrated in the hypotheses but that might influence the hypothesized relationships. These control variables are measured and held constant to test for the relative impact of employees’ leaderboard position on their intrinsic motivation (Punch, 2003). For this study, the following control variables are used: gender, age and organizational tenure. The influence of age on job motivation has been found in several studies before (Boumans, et al., 2011; Rhodes, 1983). Differences in job motivation are also found when comparing men and women (Fiogionne, & Peeters, 1982). Furthermore, the study of Coats and Feldman (1996) showed that the need for social status differs among men and women. Besides, women’s self-assessed skills are significantly lower than that of men (Hargittai, & Shaffer, 2006). The third control variable, organizational tenure, measures the tenure (in months) of the employees at their current employer. It is expected that tenure influences people’s perceptions about their skills. The longer an employee works for a

(34)

34 company, the better his skills become and thereby his skill perceptions. All three variables are taken into account when analyzing the data as they might add noise to the results.

3.5 The final construction of the survey

This questionnaire is constructed with Qualtrics’ research software (Qualtrics, 2016). The structure of the questionnaire has been thought through deliberately. The questionnaire is divided into seven sub-sections. First, respondents had the option to choose either the English or Dutch version of the survey. The original scales were written in English. The source questionnaire needed to be translated into Dutch as most employees work for companies that are operating in the Netherlands. Back and parallel translation was used to ensure

comprehensibility and consistency between the two versions. Both the English and Dutch version can be found in Appendix B. After the version is chosen, the purpose of the research is explained including a brief introduction to the topic. This section also guaranteed their anonymity. Subsequently, questions about the four research constructs are asked, beginning with employees’ leaderboard ranking and followed by their intrinsic motivation, perceived skill level and need for social status. Leaderboard ranking and perceived skill level are purposefully separated as respondents might otherwise automatically associate a higher leaderboard ranking with higher skill perceptions. Questions about respondents’

demographics are more personal and people might be intimidated by those as it infiltrates their anonymity. Such questions are therefore asked at the end of the survey. By then, people got familiar with the questionnaire and are more comfortable in answering personal questions.

Prior to issuing the questionnaire, a pretest was done to cancel out any issues. A total of four people were asked to fill out the survey and give their feedback. They made

suggestions regarding the extent to which the instructions were clear, the time allotted was correct and whether it looked like the items measured the variables they supposed to measure

(35)

35 to assure face validity. Small adjustments were made accordingly. On average, it took them 3 minutes to complete the questionnaire.

3.6 Statistical procedure

Data was collected online using Qualtrics’ research software. The survey administration started on the 12th_{of April 2017 and it closed on the 25}th_{of May. The data was then copied} into the Statistical software Package for Social Sciences (SPSS) to statistically analyze it. Multiple surveys were used to reflect the differences in leaderboard ranking per company. Seven companies cooperated, however, one organization participated with two different leaderboards. It thus resulted in a total of eight datasets. In order to merge the cases and thus compute a leaderboard ranking that is similar across all datasets, the following mathematical formula was used: LBR*(99/(LBR-1)) – (99/(LBR-1)-1). LBR represent someone’s position on the leaderboard. Furthermore, the leaderboard value had to be reversed as SPSS perceives a high ranking (e.g., number one on the leaderboard) as a low value. Again, a mathematical formula was used: (LBR-101)*-1, where 100 is now perceived as a low and 1 as a high value. Furthermore, the counter-indicative items rNSS2 and rNSS7 were recoded. They now

represent NSS2 and NSS7. Also scale means have been computed. They now represent IM_TOT, PSL_TOT and NSS_TOT.

(36)

36 4. Results

This chapter provides the results of the study. The first part describes some descriptive statistics regarding the variables used. Thereafter, the internal reliability of the scales are shown with the use of Cronbach’s alpha for the reflective constructs need for social status and intrinsic motivation. The third section discusses the correlations between the variables. In the last part, multiple regressions were performed to test the hypotheses. The first and second hypotheses are tested through a simple linear regression, the third (both a and b) and fourth hypotheses through a multiple regression analysis and the fifth is investigated by using process model 3 by Hayes (2013).

4.1 Data descriptives

To get a feel of the data and to summarize the features of the sample, some descriptive

statistics regarding employees’ leaderboard ranking, their intrinsic motivation, perceived skill level and need for social status are provided (see Table 4). The overall leaderboard ranking ranges from 100 to 1, with an average ranking of M=51.84. The skewness for leaderboard ranking is -0.049, which indicates a symmetric distribution. The scores for each individual company are indeed equally distributed. However, in the overall leaderboard ranking there is an exception for a very high (i.e., number 1) and a very low (i.e., number 100) score.

Respectively 5,3% and 6,1% of the total sample (N=132) is represented in these two groups. The rest of the scores have a percentage between 0,8 and 2,3. Together, it might explain the negative kurtosis (-1.29, SE = 0.419) found for leaderboard ranking. For intrinsic motivation, a scale from 1 to 7 (1=strongly disagree, 7 = strongly agree) is used. From the mean value (M=4.94), it seems that the intrinsic motivation of the respondents is above average, but that they are not extremely motivated. A slight negative skew (-0.861) is found, which means that there are more scores for higher levels of intrinsic motivation. This probably explains its

(37)

37 ‘higher’ mean value. Furthermore, their perceived skill level in comparison to their

colleagues, ranging from 1 to 5 (1=much worse, 5=much better), is around average with a mean value of M=3.30. The standard deviation, however, is only 0.85, which indicates a lack of variance in the scale. The scale used for need for social status ranges from 1 to 7

((1=strongly disagree, 7 = strongly agree). The descriptive statistics show that employees do not have an outstanding need for social status (M=4.54). So, like employees’ perceived skill level, also their need for social status is medium.

Table 4: Descriptive statistics

(N=132) Mean Value Std. Dev. Min. Max.

Leaderboard ranking 51.84 31.41 1 100

Intrinsic motivation 4.94 1.39 1.33 7.00

Perceived skill level 3.30 0.85 1.40 5.00

Need for social status 4.54 1.34 1.88 6.75

4.2 Reliability of the scales

The independent variable, leaderboard ranking, has no items. A reliability test for this variable is therefore not performed. The Cronbach’s alpha is also not calculated for perceived skill level as it is a formative construct. For formative measures, multiple indicators cause variance in the construct (Bollen, & Lennox, 1991). The indicators for this construct are Likert scale items measuring communication, problem solving, collaboration, technical and creativity skills. As these indicators increase in magnitude, the perceived skill level construct also increases in magnitude. However, an increase in the construct of perceived skill level does not necessarily result in an increase in one of the indicators, as would be the case if the measure was reflective. The typical concepts of reliability are not always meaningful as the internal consistency for this construct is not relevant. The construct is used and validated in previous studies (Carnevale, 1990). It is therefore assumed that the scales are a good predictor of one’s

(38)

38 perceived skill level and that the construct is reliable. Need for social status and intrinsic motivation, on the other hand, are reflective measures. Here, a reverse direction in causality is present in which the constructs cause variance in the indicators. A reliability check is run for both (see Table 5). For both intrinsic motivation and need for social status, the Cronbach’s alpha is high (respectively α=0.946, α=0.951). It indicates a high level of internal consistency among the items that together predict the overall measure. Moreover, the item-total

correlations, for both intrinsic motivation and need for social status, are above 0.3 which indicates that all items have a good correlation (Field, 2009). Also the difference in Cronbach’s alpha if an item is deleted is lower than 0.1 for all items of both intrinsic

motivation and need for social status (Field, 2009). Deleting items is therefore not considered.

Table 5: Means, Standard Deviations, Correlations and Cronbach’s Alpha’s

Variables (N=132) M SD 1 2 3 4 5 6 7 1. Gender 0.42 0.5 - 2. Age 28.36 5.21 -0.086 - 3. Tenure 33.45 30.1 -0.151 0.689** - 4. Leaderboard ranking 51.84 31.41 0.023 0.225** 0.123 - 5. Intrinsic motivation 4.94 1.39 -0.039 0.152 0.129 0.749** -

6. Perceived skill level 3.3 0.85 -0.004 0.260** 0.316** 0.123 0.159 (0.95)

7. Need for social status 4.54 1.34 0.095 0.104 0.055 0.475** 0.626** 0.124 (0.95)

* Correlation is significant at the 0.05 level (2-tailed) ** Correlation is significant at the 0.01 level (2-tailed)

4.3 Correlation

Correlation describes the relationship between two variables and quantifies the intensity and the meaning of that relationship (Field, 2009). As seen from the correlation table (Table 5), the leaderboard value correlates positively (r=0.749) and highly significant (p<0.01) with intrinsic motivation. It indicates that employees with a lower or higher ranking on the

(39)

39 leaderboard tend to score respectively lower or higher for intrinsic motivation and vice versa. Significant correlations are also found for employees’ need for social status, namely with leaderboard ranking (r=0.475) and intrinsic motivation (r=0.626). It suggests that people with a high need for social status tend to acquire a high leaderboard ranking. However, its intensity is moderate. Also, people with a high need for social status are more likely to be intrinsically motivated.

The table also shows the correlations related to gender, age and organizational tenure. Here, age positively relates to tenure (r=0.689, p <0.01), leaderboard ranking (r=0.225, p<0.01) and employees perceived skill level (r=0.260, p<0.01). It suggests that people with a higher age (in years) work longer for their current employer, have a higher ranking on the leaderboard and have higher skill perceptions. Especially for organizational tenure, the

intensity of the relation is high. Also, tenure is positively related to perceived skill level with a Pearson correlation coefficient of r=0.316 and a significance value less than 0.01.

In order to check for the effects, a regression analysis is conducted in the next section.

4.4 Regression analysis: hypotheses testing

In order to test the hypotheses, multiple analyses have been performed.

A simple linear regression analysis is performed to test the first hypothesis.

Hypothesis 1 proposed a positive effect of the independent variable leaderboard ranking on the dependent variable intrinsic motivation. Table 6 shows the results of the analysis. It first shows that none of the demographic variables significantly affect intrinsic motivation (F(3,127)=1.066, p>0.05). Leaderboard ranking, on the other hand, does have a significant and positive effect in explaining employees’ intrinsic motivation (β=0.757, p<0.01).

Leaderboard ranking has much predictive power on intrinsic motivation. After controlling for gender, age and tenure, 54,3% of the total variance in intrinsic motivation is explained by the variance in leaderboard ranking (∆R²=0.543, F(1,126) = 157.98, p<0.01). The first

The effect of competitive gamification on employees’ intrinsic motivation : the role of employees' perceived, relative skill level and their need for social status

MSc Thesis