The effect of age and gender on the impact of
gamification on employees’ work performance in a
call center
Student name: A.F. (Anne) Alsem
Student number: 11411023
First supervisor: Prof. dr. P.J. (Peter) van Baalen
Study: MSc. Business Administration – Digital Business Institution: Amsterdam Business School, University of Amsterdam
Course: Master’s Thesis
Statement of originality
This document is written by Anne Alsem who declares to take full responsibility for the contents of this document. I declare that the text and the 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.
Acknowledgements
I would like to thank Prof. dr. P.J. van Baalen for the valuable feedback and supervision during the thesis process. Furthermore, I would like to thank Ashantaly Bitorina from Kalff for helping me to get to know the organization and assisting me with the collection of data. Lastly, I would like to show gratitude to Meilon van Abs and Chris van den Berg from EngageIT, for introducing me to Kalff and providing feedback during the process.
Table of contents
Abstract ... 8
1.Introduction ... 9
1.1 An introduction to gamification and performance ... 10
1.2 Problem definition ... 12 1.3 Managerial contribution ... 12 1.4 Theoretical contribution ... 13 1.5 Structure ... 14 2.Literature review ... 15 2.1 Gamification ... 15
2.1.1 The aim of gamification ... 16
2.1.2 Gamification elements and mechanisms ... 16
2.1.3 Levels of game design elements ... 17
2.1.4 Gamification as motivational behavior ... 18
2.1.5 Gamification in the workplace ... 20
2.1.6 Gamification leaderboard... 21
2.2. Performance ... 22
2.2.1 Measuring performance ... 23
2.2.2 Performance in call centers ... 24
2.2.3 Gamification and performance ... 25
2.3 Individual differences ... 26
2.3.1 Age ... 26
2.3.2 Gender ... 27
3. Hypotheses and conceptual framework ... 29
3.1 Hypotheses ... 29 3.2 Conceptual framework ... 30 4. Methodology ... 32 4.1 Research design ... 32 4.2 Statistical procedure ... 33 4.2.1 Pre-test ... 33 4.2.2 Power analysis ... 34 4.3 Sample ... 34 4.4 Measures ... 35 4.4.1 Independent variable ... 35
4.4.2 Dependent variable ... 35 4.4.3 Moderators ... 36 4.4.4 Control variables ... 36 4.5 Interviews ... 36 5. Results ... 38 5.1 Descriptives ... 38
5.2 Testing the hypotheses ... 39
5.3 Conclusion of the analyses ... 43
6. Discussion ... 44
6.1 Interpreting the results and hypotheses tests ... 44
6.2 Interview results ... 46
6.3 Propositions ... 49
6.4 Theoretical contributions ... 52
6.5 Managerial contributions ... 53
6.6 Limitations ... 53
6.7 Future research suggestions ... 55
7. Conclusion ... 57
References ... 59
Appendix ... 65
A. Dashboard without gamification ... 65
List of Tables
Table 1. Levels of game design elements (Deterding et al., 2011) ... 18 Table 2. Means, standard deviations and correlations of the variables ... 39 Table 3. Hierarchical regression analysis for variables predicting work performance (n=46). .. 42
List of Figures
Figure 1. Conceptual framework ... 31
Figure 2. Plot of the estimated marginal means of work performance ... 41
Figure 3. The dashboard for the control group (Amsterdam) ... 65
Abstract
Recent research suggests that gamification positively influences organization’s profitability. However, little is known about the influence of gamification on work performance and individual differences, especially in a real working environment. Therefore, this study examined if gamification improves work performance of employees and whether this effect becomes stronger for men and younger employees. Data was gathered from two locations of a Dutch call center, specialized in telephone recruitment of donors for nonprofit organizations. One of these locations uses a gamification dashboard since March 2017, which shows, amongst others, the performance of an employee and its colleagues. Method triangulation was used by combining quantitative and qualitative data. Quantitative analyses have been used to compare performance data of two time periods of seven months. Moreover, interviews with employees have been conducted to explore new propositions. The main effect was examined by conducting a repeated measures ANOVA, whereas a hierarchical regression tested the moderator effects. Results showed no quantitative support for the influence of gamification on work performance. Furthermore, no support was found for the hypotheses that this effect is stronger for men and younger employees. Therefore, interviews were conducted with employees in order to explore possible reasons for the absence of the hypothesized effects. This resulted in the propositions that gamification should use collaboration mechanisms instead of competition mechanisms, gamification should focus more on long-term effects, and a gamified dashboard should be dynamic and interactive. Overall, this study contributes to the unexplored field of the influence of gamification on performance by gaining insights in the effects of gamification and proposing hypotheses for maximizing this effect.
1. Introduction
Gamification has been a trending topic in the last couple of years, in both a managerial and academic context (Hamari, Koivisto & Sarsa, 2014; Buckly & Doyle, 2017). For managers, gamification is seen as a new way to improve marketing and customer engagement. In addition, there is an increasing number of successful start-ups fully dedicated to designing and implementing games for other organizations. In the academic world there is increasing attention for gamification as well. Specifically, the number of published papers about this topic is growing rapidly. The number of papers with the term “gamification” in the title has increased significantly (Hamari, Koivisto & Sarsa, 2014). However, despite the growing popularity, much is unknown about the methods of gamification, effectiveness, results, and under which circumstances gamification works.
There are two widely known organizations that use gamification mechanisms towards customers, Nike and Starbucks. Nike launched the app Nike +, where users receive points, accumulated based on the distance travelled. This is displayed on an openly available leader board and shows individuals who trained more and earned a highly developed physique. This way, they were stimulated to perform better than their peers. This gamification mechanism is a leaderboard. Starbucks wanted to enhance the “Starbucks experience” and boost their sales as well. Every time a customer buys a Starbucks product, they accumulate stars displayed in the application as filled coffee mugs and is therefore motivated to buy more. These are two examples of a larger phenomenon. Specifically, the principles of gamification are already being applied in more contexts than people are fully aware of and have already been used in marketing for decades (Blohm & Leimeister, 2013). For example, loyalty cards for groceries, bonus payments, energy saving schemes and soccer ranking (Dale, 2014) are situations where gamification is used as well.
1.1 An introduction to gamification and performance
Performance is a relevant subject for managers as well as lower level employees in the call center industry. In general, high performing individuals in call centers tend to make more money, because mostly their salary is dependent on their performance. High performing individuals can also account for higher collective performance, because better-performing staff may also be used to transfer knowledge to other members who are not performing as well (Anton & Gustin, 2000). Besides individual performance, high importance is given to collective performance as well. Sometimes, collective performance of a call center is worse than their competitor’s. This is also commonly known as a ‘performance gap’. These gaps are very costly, and should therefore be closed soon (Anton & Gustin, 2000). In addition, Anton and Gustin (2000) found that more motivated call center agents were better able to manage stressful situations than their non-motivated peers. Subsequently, because motivated call center agents experience less stress, this results in better performance. Consequently, the question is, how can call center organizations motivate their employees to perform better and thus prevent a performance gap? There are different ways to enhance employee performance. This paper investigated if a gamified application has potential to accomplish this goal.
This potential is expected, because playing (video) games is currently very popular. Specifically, world-wide people game three billion hours per week, according to Jane McGonigal, director of game research and development at the Institute for the Future in California (Zimbardo & Coulombe, 2015). Moreover, teenagers game approximately 10.000 hours during their high school period. As such, concluding from the above-mentioned numbers of gaming world-wide, people seem to like gaming. As a result, there might be huge potential in using game elements for other goals. It became a significant global trend in recent years and is currently an established practice and industry segment (Markets and Markets, 2016). Furthermore, gamification is applied in numerous industries, such as finance, health,
education, sustainability, marketing and news and entertainment media (Deterding, Sicart, Nacke, O’Hara & Dixon, 2011; Cardador, Northcraft & Whicker, 2017). Specifically, the gamification market is estimated to grow from $1.65 billion in 2015 to $11.1 billion by 2020 (Statista, 2017). Gamification can help users in achieving their goals, as shown with the example of the Nike+ application. However, users can be employees as well. If employees’ goals are aligned with the organizational goals, the organization will achieve its goals as a result of players achieving their goals (Dale, 2014). An example of a gamification mechanism is the use of badges. Previous researchers have investigated the effect of badges. Hamari (2017) found that the use of badges positively influences the performance of a user. Nonetheless, researchers do not seem to agree on the impact of leaderboards, which is another gamification mechanism. Specifically, Farzan et al. (2008) found that leaderboards were effective in the short-term, but did not show a long-term effect. In contrast, Eickhoff et al. (2012) concluded that leaderboards were only a moderate success in encouraging competition. However, researchers found a positive impact of combined gamification mechanisms, with leaderboards as one of them. The above findings suggest that there is potential in the use of leaderboards, but more research is needed.
As such, researchers are still questioning the results of these gamification initiatives. Besides, individual characteristics have to be taken into account as well when suggesting recommendations about gamification effectiveness. This topic was researched in a call center specialized in telephone recruitment of funds and donors for nonprofit organizations and focuses specifically on the gamified leaderboard. This question will be answered by quantitative data analysis, combined with interviews with employees.
1.2 Problem definition
This paper addressed the issue of the impact of gamification on performance. Within this topic, focus specifically lied on measuring the possibility of using gamification to gain higher work performance in a call center. The gamification mechanism that is being used is a gamified dashboard with multiple competitive elements in it. Moreover, the influence of age and gender on the relation between gamification and work performance was investigated. The current body of literature suggests there are differences within age and gender with regards to technology adoption (Koivisto & Hamari, 2014). Therefore, these moderators have been chosen. Specifically, younger people generally have more experience with gaming and have a higher perceived ease of use. Consequently, age might impact the relation of gamification and performance. Furthermore, differences between men and women can affect this relation as well. Men are considered to being more competitive and pragmatic whereas the technology adoption of women is more influenced by social pressure (Koivisto & Hamari, 2014; Gefen & Straub, 1997). An elaborate explanation of this theory can be found in chapter 2.3. In conclusion, the main research question in this paper is:
What is the effect of employees’ age and gender on the relation between the use of a gamification dashboard and employees’ work performance?
1.3 Managerial contribution
Considering the managerial contribution of this paper, the results of this study will be relevant. It has been projected that 80% of current gamification applications will fail to meet their business objectives (Dale, 2014). This paper researches the effectiveness of gamification in an organization. Managers can use the results when considering making use of gamification mechanisms, because it has implications for the choice and offering of using
gamification mechanisms. Especially, for managers in a result-based environment, or specifically within a call center, this research is very relevant.
Moreover, managers can use the outcomes of this study to better reach their target group. In order to understand what improvements have to be made, we must take a deeper look into what kind of individuals are more influenced by gamification mechanisms. This study researches whether this gamification mechanism is only useful for users within a specific age range or for male or female users. As such, this research serves to provide guidance as how to better utilize gamification in a result-based industry for maximum effect.
1.4 Theoretical contribution
Previous researchers do not agree on the effectiveness of gamification, because most results are based on anecdotal and intuitive presumptions (Hamari et al., 2014). Hamari et al. (2014) performed a literature search within 24 empirical studies on the effects of gamification. They mainly focused on quantitative studies, but also included qualitative and mixed method studies. They conclude that the findings consisted both positive and negative perceptions regarding the impact of gamification. Therefore, empirical results on the effectiveness of gamification are needed.
Moreover, researchers have paid little attention to the relation between individual characteristics and the influence of gamification on their work performance. This research is academically relevant by adding individual differences, namely gender and age, as moderators on the relation between gamification mechanism and user’s work performance. Buckley and Doyle (2017) found that the experience of gamification varies depending on individual attributes. The experiment was executed amongst undergraduate students within a study project. They stated that future research is needed in different contexts. The fundraising industry serves as a different context. This study contributes to the theory in a way that there
is a lot of disagreement on the effects of gamification on user’s work performance and individual differences influencing this effect. Therefore, this research will help understanding this relation. Specifically, there is a lack of research focused on the influence of age and gender, while it is suggested that these elements have significant effects on the impact of gamification (Buckley & Doyle, 2017). Investigating these effects can provide a better understanding of the context of gamification.
Furthermore, most research about the influence of leaderboards was conducted in an experiment setting, such as a lab approach. However, this research is applied in a real-life setting where the respondents did not know they are being studied. Therefore, this research can be seen as a field experiment. Lastly, most research is focused on either measuring the effects of multiple gamification elements or single gamification elements. Nonetheless, little research has been conducted where a leaderboard was experimentally isolated (Landers, Bauer & Callan, 2017). Therefore, it is difficult to conclude the effects of leaderboards.
1.5 Structure
This thesis is structured as follows. Chapter 2.1 reviews existing literature on gamification, the characteristics and the effects of gamification. Chapter 2.2 consists of a literature review on work performance and the relation of gamification and work performance. The individual differences are explained in chapter 2.3. The conceptual model and hypotheses are presented in chapter 3. Chapter 4 explains the data and methodology of this study. Results of the main analysis and moderator effects are presented in chapter 5. Chapter 6 includes a discussion of the results, followed by a conclusion in chapter 7.
2. Literature review
This chapter provides an extensive review on existing literature about the research topic. First of all, previous research about gamification is elaborated on. This covers the aim of gamification, different mechanisms and gamification elements and applications of gamification. Paragraph 2.2 describes how to measure performance, takes a look at performance in call centers and ends with the relation between gamification and performance. This chapter ends with a paragraph explaining the relation between age, gender and gamification.
2.1 Gamification
Gamification can be defined as “the concept of embedding game features into activities that are not themselves games” (Werbach & Hunter, 2012: 27). In other words, gamification is “the use of game design elements in non-game contexts” (Deterding et al., 2011: 2426). Building further on that, gamification does not mean that users have to play a game, because it only uses elements of games. Moreover, these definitions distinct gamification from video game-playing and apps, because these have no business context. This makes sense, since video games or games in apps are designed primarily to be entertaining. They have been demonstrated to motivate users to engage with them intensely and enduring. Therefore, users see gamification as gameful. Consequently, game elements should be able to engage users in non-game contexts as well (Deterding et al., 2011). In other words, because games are fun, so any product or service using the same mechanism should also prove to be more valuable and engaging (Hamari, 2013). As such, it is believed that gamification is effective.
2.1.1 The aim of gamification
Koivisto and Hamari (2014) state that the aim of gamification is to support and motivate the users to perform tasks promoted by the gamification services. It is commonly used to advance goals outside the game itself. In other words, gamification mechanisms are being used to direct people’s motivations, rather than their attitude and/or behavior directly (Hamari & Koivisto, 2013). For example, the services in the gamification application used by Nike aim to increase user’s exercise by motivating the users to work out more. Nevertheless, the ‘users’ can be lots of different stakeholder groups. Therefore, gamification is applied in achieving different goals. Organizations use gamification to increase the customer value and encourage value-creating behaviors, such as increased consumption, greater loyalty, engagement or product advocacy (Blohm & Leimeister, 2013; Zichermann & Cunningham, 2011).
2.1.2 Gamification elements and mechanisms
There are several gamification elements that organizations use in order to reach the goals stated above. Game elements are a “set of building blocks or features shared by games” (Deterding et al., 2011: 4). For gamification, these elements are characteristic to games, so they are found in most games, are associated with games and play a significant role in games (Deterding et al., 2011). Three of the most commonly employed game elements are points, leaderboards and levels (Mekler, Bruhlmann, Tuch & Opwis, 2017). Other meaningful gamification elements are (Galli, Fraternali & Bozzon, 2014):
Points: number that represents a certain skill of the user, or their progress and serve as continuous and immediate feedback and as a reward;
Leaderboards: rank of players based on the results of a certain game or system, shows who performs best in a certain activity;
Badges: visually track the progress in a system and can be earned in the gamification environment. They also provide the user with feedback on their achievements and accomplishment of levels or goals;
Avatars: visual representation of the players in the game or gamification environment. They can be chosen or created by the player and identify the player.
Other mechanisms are time constraints, promotions, lotteries, clear goals or virtual rewards (Lucassen & Jansen, 2014). In total, Lucassen and Jansen (2014) identified 19 gamification elements. These can be divided into three categories attributed to the primary benefit of gamification. These benefits are engagement, loyalty and awareness, and are based on 12 interviews with marketing executives. The majority (14 out of 19) of the elements contribute to an increase in ‘engagement’ as the primary benefit. The other five are categorized to ‘loyalty’ (access restrictions, virtual rewards and badges) and ‘awareness’ (promotions and lotteries). These elements trigger different mechanisms in order to reach the goal, such as rewards, feedback, cooperation, challenges, transactions, turns, competition, resource acquisition, win states and turns. The game elements do not define the game, but the mechanisms they trigger do. Besides, a gamified application uses several design element(s) from games, but is often not a full ‘game proper’.
2.1.3 Levels of game design elements
Deterding et al. (2011: 12) have found five layers of abstraction for all game design elements. These can be found in Table 1. The levels are presented from concrete to abstract. The first and most concrete level is game interface design patterns (Chao, 2001), followed by game design patterns (Björk & Holopainen, 2005) or game mechanics (Taylor, 2009). More abstract is level three, game design principles, heuristics or ‘lenses’ (Schaffer, 2008). The fourth level is conceptual models or game design units (Brathwaite & Schreiber, 2008;
Calvillo-Gámez, Cairn, & Cox, 2010). The most abstract level is game design methods and design processes (Belman & Flanagan, 2010).
Level Description Example
Game interface design patterns
Common, successful interaction design components and design solutions for a known problem in a context, including prototypical implementations
Badge, leaderboard, level
Game design patterns and mechanics
Commonly reoccurring parts of the design of a game that concern gameplay
Time constraint, limited resources, turns
Game design principles and heuristics
Evaluative guidelines to approach a design problem or analyze a given design solution
Enduring play, clear goals, variety of game styles Game models Conceptual models of the components of
games or game experience
MDA; challenge, fantasy, curiosity; game design atoms; CEGE
Game design methods
Game design-specific practices and processes
Playtesting, play centric design, value conscious game design
Table 1. Levels of game design elements (Deterding et al., 2011)
2.1.4 Gamification as motivational behavior
As stated before, gamification can serve as a motivational trigger to perform a certain task. Ryan and Deci (2000) identified two forms of motivation according to the self-determination theory (SDT). Intrinsic motivation can be defined as “doing something, because it is inherently interesting or enjoyable” (Deci & Ryan, 1985: 55). External motivation refers to “doing something due to a separable outcome, such as pressure or extrinsic rewards” in the form of money or verbal feedback (Deci & Ryan, 1985: 5). Gamification combines both forms of motivation, because it uses extrinsic rewards such as points, levels and badges to improve engagement. However, it also strives to raise feelings of achieving mastery, autonomy and sense of belonging (Muntean, 2011). These are examples of intrinsic motivation. However, it is demonstrated that the effect of rewards decreases intrinsic motivation over time. Nevskaya and Albuquerque (2015) argue that rewards must be
continually offered to maintain interest. Taking these rewards away might lead to consumer abandonment. In other words, for many consumers, these extrinsic incentives are never fully internalized, which also affects the usage of game-specific rewards. Specifically, the long-term effect of rewards on motivation is not clear yet, when stimulated by gamification mechanisms.
A potential side effect of gamification is the crowding out effect. This is defined as “a situation in which motivation with an internal locus of causality decreases, for instance in response to an intervention such as PBF” (Lohmann, Houlfort & De Allegri, 2016: 3). PBF is the abbreviation for performance-based financing. This internal locus of causality means that employees are intrinsically motivated to perform a task or certain behavior. In contrast, an external locus of causality means that an individual perceives external or internal pressure and hereby feels motivated. In the crowding out effect, motivation with an internal locus of causality is decreased, whereas the external locus of control is increased. Subsequently, this replaces the internal locus of causality. The disadvantage is the fact that behavior with an internal locus of control tends to account for better performance (Lohmann, Houlfort & De Allegri, 2016). PBF is a form of external rewards and is used within Kalff as well: the better an employee performs, the more they get paid. This means that employees at Kalff might be subject to the crowding out effect as well.
Since intrinsic motivation predicts behavior change, it is suggested that the more rewards one gets, the less it will behave accordingly (Hofacker, De Ruyter, Lurie, Manchanda & Donaldson, 2016). External stimuli often fail to increase long-term motivation, because adaptation effects undermine the effectiveness (Blohm & Leimeister, 2013). This is in accordance with the research of Mekler et al. (2013), who found that game elements did not increase intrinsic motivation. Therefore, you cannot rely upon them to sustain long-term user engagement. However, intrinsic motives and flow can be systemically activated by setting
extrinsic incentives (Ryan & Deci, 2000: 63-65). This can be achieved by offering game-specific symbolic awards, because those rewards provide visual evidence of one’s performance. For example, these rewards help to document progress toward personal goals, facilitate social interaction in a community of peers and competition and they function as an instrument of social recognition within games. So, gamification services are able to increase intrinsic motivation of users (Blohm & Leimeister, 2013), Therefore, game-specific symbolic rewards (such as points and badges) can serve as both intrinsic and extrinsic motivators.
There is another framework that distinguishes different types of motivational behavior (Kim, 2012). The first type is expression, where people are motivated by gaining more and better abilities to show their creativity and express who they are. The second type is exploration, for example content, people, tools and worlds. They are motivated by information, access and knowledge. Besides, collaboration behavior consists of a purposeful, non-zero-sum way of socializing. People who are motivated by collaboration want to ‘win together’ with others. The last type, and in this research the most relevant one, is competition. Competition drives social gameplay and self-improvement; the motivation to compete against yourself. (Kim, 2012). This research mainly focused on the competitive element, because Kalff uses a dashboard with people’s own results and results of their colleagues, with the aim of stimulating competition.
2.1.5 Gamification in the workplace
In paragraph 2.1 the aim of gamification was described. The aim depends on the desired benefits of the company. Potential benefits are increased motivation and productivity of employees; alignment of goals expectations of employees, stakeholders and customers with the company’s goals; full engagement of employees with new company initiatives; and employees converted into advocates of the company (Dale, 2014). Many companies are
struggling with the question on how to improve on these benefits. However, there is mounting evidence that shows that people can change their behavior when given the right incentives (Dale, 2014). These incentives not necessarily have to be financial. Rewards should be more associated with ‘fun’, whereas financial incentives are mainly associated with ‘normal work’. One important aspect is that the gamified application must be able to measure the required activities. In this research, the gamified dashboard constantly measures the activities and accomplishments of the employees and is therefore appropriate as a gamification application.
2.1.6 Gamification leaderboard
The gamification mechanism in this study is a leaderboard, which is a commonly used mechanism to improve work performance. Specifically, the waiting screen is a dashboard which includes a leaderboard. A leaderboard “introduces a sense of competition by letting people know where they stand relative to their peers” (Dale, 2014: 85). An example is the app Strava, which calls itself “the social network for athletes”. The app allows athletes to track & analyze their activities (for example cycling or running), share & connect with friends and explore & compete against each other. The competition part is reflected in the fact that routes are divided into segments, which are popular stretches of road or trail. A leaderboard of times is created set by every Strava athlete who has been there before. This competitive element is shown visually to the user and is highly popular among users. However, what is the (potential) impact of a leaderboard?
Several researchers have investigated the effectiveness of a leaderboard. Farzan et al. (2008) introduced a points system in order to increase user contribution in an enterprise social networking. They concluded that the users who saw the points they and their peers earned, contributed much more content to the website. They found that the users started comparing themselves with their peers and accordingly user contributions increased four weeks after
deployment. Consequently, they assume that a leaderboard can increase user activity. However, they did not show a long-term effect. In contrast, other researchers concluded that leaderboards were a moderate success in encouraging competition (Eickhoff, et al., 2012). They used leaderboards as part of a larger game to improve flow states. The effect of leaderboards has also been tested among students in an e-learning platform. Dominguez et al. (2013) found that students viewing the leaderboard scored higher on the practical assignments and in overall score, but lower on written assignments. All these researchers suggested that leaderboards are effective. However none of them included leaderboards as the only gamifying mechanism. Nonetheless, Landers et al. (2015) have researched the impact of leaderboards on task performance in an experiment setting. Respondents used a leaderboard where they were able to see prior scores. This research showed that the presence of a leaderboard was successful in motivating participants. However, the prior scores on the leaderboard were only random initials and did not represent their peers. Therefore, they suggested that the results might have been different if these scores represented actual people, in order to draw conclusions on the social component of leaderboards. In this study, the leaderboard represents all colleagues who are making phone calls for the same campaign. In addition, they suggest that gamification by using a leaderboard might be more effective for relatively simple tasks, such as sales performance. In this research, it was expected that the leaderboard is effective, since acquiring new consumers is a task where it is easy to measure performance.
2.2. Performance
To date, there are a lot of different definitions of performance in the academic world. Moreover, organizations are struggling with finding the right measures to assess performance. The most cited definition of performance measurement is from Neely, Gregory and Platts
(2002: 1229): “the process of quantifying the efficiency and effectiveness of past actions”. As such, it is stated that performance is always a quantitative measure. However, for many managers it is unclear what they should quantify and how (Moullin, 2007). Therefore, Moullin (2007: 188) suggested another definition: “evaluating how well organizations are managed and the value they deliver for customers and other stakeholders”. Hence, researchers do not seem to agree on the definition. Moullin (2007) suggests that evaluating performance is more than just measuring, because it includes interpretation and analysis as well.
In addition, organizations are struggling with finding the right triggers to improve performance. In the call center business, not much research has focused on addressing this topic. Anton and Gustin (2000) concluded that there is not much clarity amongst call center managers about the proper ways to motivate staff.
This remainder of this chapter elaborates on existing literature about performance. Paragraph 2.2.1 explains how performance is being measured, followed by a paragraph about performance in call centers. Paragraph 2.2.3 ends the chapter by elaborating on existing literature on gamification and performance.
2.2.1 Measuring performance
There are two types of performance assessment: subjective and objective. They differ in the presence of the assessing person. Specifically, objective performance evaluation is independent of the assessing person. In contrast, subjective performance measurements are influenced by the attitudes, perceptions and beliefs of the assessing person (Woods, 2012). The numerical outcome of objective performance is always the same, while “the correctness of a subjective assessment cannot be determined by a third party” (Bol, 2008: 2). This research focused on objective performance measures, because this assures the correctness of the assessment. It is measured directly and automatically and is therefore not subject to the
normal problems of subjective performance ratings. Besides, Kalff mainly uses objective measures to determine an employee’s performance and subsequently their wage. Each month, it is determined in which of the 16 salary levels an employee is, which can differ each month. The level is depending on an employee’s individual performance, for example how much acquisitions he/she does and for which amount of money, how fast he/she works, etc Moreover, it is determined by the performance in comparison to co-workers and targets set by the fundraising company. The leaderboard shows, amongst others, measures that are being used to determine the salary.
2.2.2 Performance in call centers
Until now, little research has been done on the performance of call centers. However, Anton (2000) found many different metrics that can be used to measure performance in a call center. For example, average speed of answer, average talk time, average after call work time, average abandonment rate, average time before abandoning and sales per hour (Anton, 2000). Although these measures have proven to be useful in research to performance, not all measures can be used to measure performance in outbound call centers. For example, abandonment means that incoming calls from customers hang up the phone. Since Kalff only focuses on outbound telephone calls, this measure is not relevant for them. More relevant measures are for example, percentage of calls that result in a sale or average sale value of a call (Anton, 2000).
Kalff’s measurement system has characteristics of Moullin’s definition of performance, because it involves interpretation and analysis to evaluate performance. Within Kalff, the individual and collective effectiveness of employees are constantly measured. However, the management team also tracks to what extent the targets set by the client are achieved and if not, what went wrong and what can be improved. In addition, they also make
use of qualitative data to measure performance. All team leads keep track of the learning development of the employees frequently. Therefore, performance measures are a combination of quantitative and qualitative data.
2.2.3 Gamification and performance
Several researchers have studied the impact of individual game elements on performance. For example, De Marcos, Domínguez, de Saenz-de-Navarrete and Pagés (2014) studied the effects of a gamification plugin in a learning management system. They found that the students presented better performance than with the traditional e-learning approach. This can be explained by the fact that gamification offers the possibility to provide (possibly continuous) real-time information on performance (Cardador et al., 2017). Hamari (2013) conducted a field experiment to assess the impact of badges. He concluded that implementing gamification mechanisms does not automatically lead to higher user activity. However, users who actively monitored their own badges and those of others, did show higher user activity. This suggests that using a leaderboard has a positive effect as well. It shows an individual’s performance and that of their peers, which is an example of active monitoring as well.
The research of Mekler et al. (2017) focused on measuring the effect of points, levels and leaderboards. They found that users who experienced any of the above-mentioned gamification mechanisms, performed better. However, the impact of a gamified dashboard is still lacking. Furthermore, they state that they lacked relatively clearly defined quality metrics, because this would have made performance assessment easier. Besides, Jung, Schneider and Valacich (2010) found that users’ performance increased when the users were given a clear goal (levels and leaderboards), as opposed to users who were asked “to do their best”. This suggests that in the situation of Kalff, higher performance can also be expected, if the employees get clear goals from their team leads.
2.3 Individual differences
Until now, little research has been conducted on investigating individual differences in technology adoption processes (Gefen & Straub, 1997). However, it is suggested that knowledge about how men and women respond to technology adoption process can be essential to compete in an emerging organizational environment. The differences in their perception, motivations and information processing in technology adoption processes can have significant effects (Morris, Venkatesh & Ackerman, 2005). The following paragraphs will elaborate on the potential influence of the moderators age and gender.
2.3.1 Age
Previous researchers studied the effect of age on the benefits of gamification. This showed that ease of use diminishes with the main effect of age. The more mature users are, the less they experience ease of use (Koivisto & Hamari, 2014). This is because teenagers have more gaming experience: they game approximately 10.000 hours during their high school period (McGonigal, 2011). This effect is called digital divide between generations. It is emerged as younger generations are exposed to digital technologies earlier and earlier. In addition, previous researchers have divided gamers into different player types on the basis of several characteristics. Hamari and Tuunanen (2014) included ten studies on player types and listed them on the basis of the concepts they used. It shows that younger workers have been exposed to digital technology at a younger age. This makes computer games an immensely compelling and rewarding experience for them. This results in the fact that 97% of youth play computer and video games, whereas only one out of four gamers is over the age of fifty (McGonigal, 2011).
Similarly, generations differ in terms of educational background, learning styles and approaches to technology. As such, younger workers tend to experience higher self-efficacy
and less computer anxiety than older people. In addition, they perceive their skills to be much higher (Arning & Ziefle, 2007). In conclusion, younger people have more experience with gaming, experience less computer anxiety and higher self-efficacy. Therefore, it is expected that the effect of gamification is stronger for younger users.
2.3.2 Gender
Prior research has shown that men and women differ in their decision-making processes (Venkatesh, Morris & Ackerman, 2000). Men’s decisions about technology usage are more strongly influenced by their perceptions of usefulness. Perceived usefulness is high when a person believes that using a particular technology will increase their job performance. This highlights the instrumentality of the system. Previous studies have indicated that perceived usefulness determines user acceptance, adoption and usage behavior. Moreover, men are generally more task- and achievement-oriented than women and display more instrumental behavior (Koivisto & Hamari, 2014). This suggests that in the context of a call center, men are more focused to accomplish their tasks and achievement, which is acquiring customers. The gamified dashboard can help them to gain insight in the extent to which they have accomplished their goals. Moreover, men show more competitiveness and need for winning than women do (Hartmann & Klimmt, 2006). Therefore, it is expected that men are influenced by the performance results of their colleagues which are visually shown on the gamified dashboard.
In contrast, women are more strongly influenced by the perceived ease of use of the particular technology (Venkatish et al., 2000). A high perceived ease of use means that little effort is needed to use the technology. Furthermore, Venkatesh et al. (2000) state that women are more concerned with social relations and are more prone to social influence. During the initial adoption of a technology, women are more concerned with social relations and more
subject to social influence (Koivisto & Hamari, 2014). When applied to an online context, women are more motivated to use a new application when they feel the social pressure to do this. In contrast, men are more pragmatic users. In this study, the gamification dashboard serves a functional need. It is used to gain insight in one’s current performance and the performance of the campaign and their colleagues. Moreover, female players might not perceive as much gratification by the social interactions included in games as men (Koivisto & Hamari, 2014). Consequently, it is expected that men perceive the new dashboard as more useful and are therefore more influenced by it.
Another difference between men and women is the fact that women show higher perceptions of computer anxiety. A reason for this difference might be that IT is generally considered as a male-dominated field (Gefen & Straub, 1997). However, this relation depends on the context of the technology. Specifically, in this study we can assume that female employees already know how to use the dashboard. Nonetheless, the new dashboard shows more visuals and information and might therefore be experienced as a lower perceived ease of use.
3. Hypotheses and conceptual framework
This chapter outlines the proposed hypotheses. Moreover, it includes the conceptual framework in which the relations between the variables are explained.
3.1 Hypotheses
Main effect. To conclude chapter 2, there is a lot of disagreement on the relation between gamification and user’s performance and especially individual differences influencing this relation. With regards to the gamification mechanism, we focused our research on the dashboard, because there is ample evidence that this positively enhances performance. However, research has showed that in other contexts, a gamified dashboard positively influences performance. Besides, Landers et al. (2015) suggested that gamification by using a leaderboard might be more effective for relatively simple tasks, such as sales performance. In this case, the individual performance was being measured and the task can be seen as a relatively simple one. To examine if there is a positive relation between gamification and performance, the following hypothesis is tested:
H1: Employees using the gamification dashboard perform better than employees using the standard dashboard.
Age. As previously stated, the digital divide results in significantly more experienced younger users than older users when it comes to gaming. This results in the fact that their gaming experience is more compelling and rewarding (Hamari & Tuunanen, 2014). Besides, Koivisto and Hamari (2014) showed that ease of use diminishes with age. Therefore, it is expected that older users find it harder to use the gamification tool, which might result in lower effectivity. Lastly, younger people tend to experience less computer anxiety and perceive more computer skills than older people. Therefore, they are expected to be more open towards a gamification
tool, due to a higher self-efficacy. Consequently, it is expected that younger employees will be more influenced by the implementation of a new dashboard which includes gamification elements, hence we posit that:
H2: The effect of gamification on work performance is stronger for younger employees compared to older employees.
Gender. Koivisto and Hamari (2014) found gender differences in the perceived benefits of gamification. Specifically, female users were more positive about the received recognition by the game and are more subject to social influence. This means that they are more likely to use a new application when they feel the social pressure to do this. However, men are in general more competitive and have a higher need for winning than women do. Since the leaderboard in this study contains competitive elements and gives transparency to everyone’s results, it is expected to be more effective for men. Moreover, there is no social pressure to use the new dashboard, since every employee in Almere starts using it at the same time. Furthermore, men attach importance to perceived usefulness and are more task- and achievement-oriented than women (Koivisto & Hamari, 2014). In other words, men are more pragmatic users. The practical usability of the new dashboard is high, because it serves a functional need. In conclusion, the reasons above provide argumentation for the third hypothesis:
H3: The effect of gamification on work performance is stronger for men compared to women.
3.2 Conceptual framework
The three hypotheses are visualized in the conceptual framework. The relations in this framework will be tested in the study. It shows the relations between the independent variable (gamification) and the dependent variable (work performance) with two moderators (age and gender). Furthermore, the control variables (team lead, working time and hours on campaign)
will be taken into account as well. First of all, it is expected that gamification has a positive effect on work performance. In addition, it is hypothesized that the effect of gamification on work performance is stronger for younger workers; therefore a negative relation with the moderator age is expected. Finally, it is expected that the effect of gamification is stronger for male employees than for female employees. This framework will be tested in a call center with a gamified dashboard. Figure 1 visualizes the conceptual framework.
4. Methodology
This chapter explains how the empirical research is conducted. First, a description of the research design is provided. Paragraph 4.2 describes the statistical procedure. The third paragraph contains a description of the sample, followed by a description of the measurement of the variables. The chapter ends with an explanation of the employee interviews.
4.1 Research design
The research was executed by analyzing data from Kalff, an organization in the (telephonic) fundraising industry. Kalff is a Dutch company specialized in telephonic acquiring of foundations and donors for non-profit organizations. The organization has two locations, in Almere and Amsterdam. The location in Almere (+/- 200 employees) has implemented a gamified dashboard since March 2017. This dashboard is showed to the employees during the time in-between phone calls. It shows, their personal development, performance, and performance of their fellow employees in a visual, easy-to-understand way. This situation was suitable for answering the research questions, because the impact of gamification on performance could be measured by comparing the results of both situations.
The aim of this study was to research how gamification influences work performance and how age and gender are affecting this relation. Therefore, this study used an experimental design to test the cause and effect relation. This was done by manipulating the independent variable (gamification) to find the effect on the dependent variable (work performance). In this study, four conditions will be compared: two locations, both before and after gamification. The employees in Amsterdam served as the control group, whereas the employees in Almere were the experimental group. Figures 3 and 4 in the Appendix show the different dashboards. Figure 3 shows the dashboard for the control group, whereas figure 4
represents the gamified dashboard. Employees only work at one location, so all respondents in Almere transferred to a new dashboard.
4.2 Statistical procedure
As stated before, the respondents have been assigned to a certain condition (gamification or no gamification) on the basis of the location where they work. The Amsterdam location was the control group (standard treatment), while the Almere location was the intervention group (experimental treatment), because they make use of the gamified dashboard. This division was done by the management of the organization to use the Almere location as a ‘pilot’ for gamification. Although in the end only data of the ‘Artsen Zonder Grenzen’ (AZG) campaign was used in this study, in the first place data of multiple campaigns was gathered from both locations. A pre-test was conducted to see if it was possible to use performance data of multiple campaigns.
4.2.1 Pre-test
A pre-test was conducted in order to explore the opportunities to combine two campaigns and in that way, enlarge the response group. Therefore, performance data from the ‘Wereld Kanker Fonds’ (WKF) campaign was added to the data from the AZG campaign. First, employees were removed that worked less than one hour for the specific campaign. In order to be able to compare the results of the two campaigns, the effective response was divided by the target of the campaign. For example, for the AZG campaign, the target for the effective response is 34%, whereas for the WKF campaign the target is 12%. By dividing the effective response by the target, there is a percentage that shows whether the employees reached their performance goals. After that, an independent samples t-test was conducted in SPSS in order to assess whether the data of the two campaigns could be combined. This showed that the
average performance of AZG was significantly higher M = 90.04 (n = 46) than the performance of the WKF campaign (M = 72.31, n = 36, t (80) = 2.48, p = .015). Therefore, it can be concluded that it is not possible to combine the two campaigns. The campaign that was chosen is AZG, because this campaign has the largest sample size.
4.2.2 Power analysis
As stated above, the data set consists of 46 respondents. A statistical power analysis was performed for sample size estimation. It was used to determine the minimum size of the sample. A moderate effect is expected (Cohens d = 0.3), based on the outcomes from the study of Eickhoff et al. (2012), who found that leaderboards moderately encouraged competition. With an alpha = .05 and power = 0.95, the project sample size needed with this effect size (GPower 3.0.10) is n = 44 for the between/within group comparison. Thus, our proposed sample size of n = 46 will be adequate for the objective of this study.
4.3 Sample
The data consisted of the performance of the employees in Amsterdam and Almere. The two time periods are June 1 2016 until January 1 2017 and the period March 1 2017 until October 1 2017. This time frame was chosen, because the new (gamified) dashboard was implemented and explained to the employees of the Almere location in February 2017. In January and February, the new dashboard was tested extensively. Therefore, these months have been excluded from the analyses. Starting March 1st, all employees in the experimental group knew how to use the gamified dashboard and started using it. This study only includes data from one campaign, because the pre-test showed that combining multiple campaigns was not possible. The campaign was from the organization ‘Artsen Zonder Grenzen’, a non-profit organization providing medical help. This campaign was chosen, because employees in both
locations worked for it, with the largest amount of employees. Respondents who did not spend more than one hour working on the campaign were excluded. Furthermore, missing values were excluded by list wise deletion. The total sample consists of 46 respondents, from which nineteen are working in Amsterdam and 27 in Almere.
4.4 Measures
This paragraph outlines the measurements of the variables that have been used in this study.
4.4.1 Independent variable
Gamification dashboard:
Employees of the Amsterdam location do not make use of the gamification dashboard. Employees of the Almere location are using the gamification dashboard since March 2017 (in the data this is labelled as Location, because the location determines the group).
4.4.2 Dependent variable
Work performance:
Work performance was measured as the effective response as percentage of the target. This percentage shows to which extent an employee reaches the performance goal, as suggested by Anton (2000). The target is agreed upon by the company and the client (in this study ‘Artsen Zonder Grenzen’). The effective response is the ratio of accepts and refusals. Accepts are the number of people who have agreed to donate money to the non-profit organization and refusals are the number of people who have declined to donate money to the non-profit organization. This variable was chosen, because it is the most direct measure of work performance.
4.4.3 Moderators
Age: measured on an interval scale
Gender: male / female, measured on a nominal scale
4.4.4 Control variables
In this study, the control variables team lead, working time, and hours on campaign were taken into account. It was expected that certain team leads might have a direct positive effect on the performance of the employees they manage. Therefore, the team lead served as a control variable. Furthermore, it was expected that the longer the employees work at Kalff, the better their performance, because they are more experienced. Therefore, this was a control variable as well. Lastly, it was expected that the more time an employee spends on a certain campaign, the better they perform. This is because the more experience an employee has, the easier they can tackle questions from (potential) customers. This can directly improve the work performance, but will not be influenced by the gamified dashboard. Therefore the control variables were as follows:
Team lead: the manager of an employee
Working time: the number of days an employee is working at one of the locations
Hours on campaign: the number of hours an employee spent calling for the specific campaign
4.5 Interviews
Next to the statistical analysis, the researcher conducted interviews with employees from both locations. This has been done in order to validate and explain the quantitative results that have been found to be non-significant. Moreover, the interviews were used to come up with propositions about gamification implementation. This study combined two data sources to
collect data on this topic, which is called method triangulation. This was used to develop a comprehensive understanding of a phenomenon (Carter, Bryant-Lukosius, DiCenso, Blythe & Neville, 2014), in this case gamification.
The researcher made use of semi-structured interviews with open-ended questions. This provides the opportunity to identify new ways of understanding the topic. Furthermore, it allows preparing the questions beforehand, so the results of the quantitative analysis can be taken into account. Besides, it is easier to compare the results of semi-structured interviews as well. This is useful in this study, because a comparison was made between the two locations of the organization.
In total, eight interviews have been conducted, four in Almere and four in Amsterdam. Each interview took approximately 15 to 30 minutes. In Almere, interviews were conducted with employees who have worked with the old dashboard as well as the new (gamified) one. In Amsterdam, the selection of respondents was random. However, the researcher tried to talk to both men and women, from different age categories. The questions for the Almere respondents were based on the non-significant results from the quantitative analyses. Based on earlier informal conversations with employees and the quantitative analyses, the researcher made some hypotheses of the reasons why the results were not as expected. For example technical reasons, because employees expressed complaints that the dashboard does not work properly. Furthermore, it was questioned whether there is a need for gamified incentives, when their reward is already dependent on their work performance. After the interviews have been conducted in Almere, the researcher defined some conceptual propositions. These have been tested with the interviews in Amsterdam. The results of the interviews can be found in the Discussion section of this paper.
5. Results
Paragraph 5.1 covers the descriptives of the variables. The subsequent paragraph 5.2 contains the main analyses: the impact of gamification on work performance, followed by a regression analysis to test the moderator effect. The results are concluded in paragraph 5.3.
5.1 Descriptives
The total sample consisted of 46 respondents, from which nineteen are working in Amsterdam and 27 in Almere. The employees in Almere are working with the gamified dashboard, whereas the employees in Amsterdam work with the regular dashboard. Of the total sample, fifteen respondents were male (32%) and 31 are female. The population, consisting of all current employees of both locations, contained similar characteristics; 62% female and 38% male. The respondents were between nineteen and 72 years old and the average age of the sample was 43 years (SD = 14.84). This is slightly higher than the mean age of the population (M = 36.51). This can be explained by the fact that the respondents from the sample were chosen specifically because they have been working in the organization for at least ten months. This selection had to be done, because it was required that they experienced both the old and new dashboard. It is expected that older people work longer at the call center, because they have more work experience and are therefore more capable of their tasks. Moreover, many students work at the call center as a side job, which is mostly temporary.
Table 2 shows the means, standard deviations and correlations of the variables. It shows that some variables were significantly correlated, with correlations ranging from medium to large in size (Cohen, 1988). For gender, a medium correlation was found with age. Furthermore, the work performance before gamification shows a large correlation with age. In addition, the work performance before and after gamification are positively correlated.
This correlation makes sense, because an employee’s work performance is expected to be more or less constant over time. The work performance after gamification is moderately correlated with the hours that employees worked. This correlation is not surprising, since more experience with the campaign will likely result in better performance.
M SD 1 2 3 4 5 6
1. 1. Age 43.83 14.84 -
2. 2. Gender 0.67 0.47 -.38** -
3. 3. Work performance before 86.18 37.39 .50** -.15 -
4. 4. Work performance after 93.07 36.29 .23 -.24 .44** -
5. 5. Hours worked 41.48 57.41 .14 -.22 .19 .43** .18 - 6. 6. Working time 422.15 104.21 .24 -.27 .10 .12 .17 .17
*p < .05 **p < .01
Table 2. Means, standard deviations and correlations of the variables
5.2 Testing the hypotheses
The first hypothesis is as follows:
H1: Employees using the gamification dashboard perform better than employees using the standard dashboard.
This hypothesis tests whether the change in work performance is higher for employees working with the gamification dashboard in Almere than for employees in Amsterdam.
In order to test the first hypothesis, a mixed-design analysis of variance model (ANOVA with repeated measures) was conducted. Specifically, a two-way 2 (location: Almere or Amsterdam) x 2 (time: before and after) mixed ANOVA with repeated measures was performed. A mixed ANOVA can be used to test for differences between two or more independent groups whilst subjecting participants to repeated measures. A mixed model means that the model uses at least one between-subject factor and at least one within-subject
factor. The within-subject factor means that the dependent variable is measured more than once per subject. In this study, the within-subject factor is the work performance of the employees, which is measured both before and after the introduction of gamification. For the between-subjects factor, it is required to use a different group of subjects for each factor level. Specifically, in this study the between-subject factor is the presence or absence of gamification and means both locations, which are the independent groups. For this study, the work performance of the employees is being measured in the specific condition before the introduction of gamification and after the introduction of gamification. Therefore, the study used repeated measures. When testing the main effects of this within-factor, it is tested as an interaction with the factor experiment.
In order to conduct and interpret a repeated-measure analysis, the assumption of sphericity must be met. Sphericity means that the variances of the differences of the dependent variable before and after the experiment are equal for all groups. However, for sphericity to be an issue, there need to be at least three groups. Therefore, in this study we can conclude that sphericity is not an issue, since the repeated-measures variable only has two levels.
The following figure shows the possible effect of gamification on the work performance of employees in both locations, both before and after gamification. It also shows the direction and size of the effect of gamification of work performance, if it is significant.
Figure 2. Plot of the estimated marginal means of work performance
H1. In the graph, it appears that the work performance of the employees in Amsterdam has increased after gamification. However, this is the location where gamification was not introduced. In order to test if this effect is significant, the results of the mixed-model ANOVA are used. There was no significant main effect of time, F(1, 44) = 2.19, p = .146. This tells us that the work performance of the two groups did not significantly change after the introduction of gamification. Furthermore, there was no significant main effect of location, F (1, 44) = 0.40, p = .529. This means there was no overall difference between the two locations over the two time periods. Therefore, it can be concluded that hypothesis 1, which states that gamification positively influences work performance, was not supported by the data in this study.
The hypotheses that test moderation effects are as follows:
H2: The effect of gamification on work performance is stronger for younger employees compared to older employees.
In order to obtain an answer to these hypotheses, a hierarchical regression analysis was executed. This test was chosen, because it shows whether the variables explain a significant amount of variance in the dependent variable. In this study, the hierarchical regression shows if gender, age and the covariates explain a significant amount of variance in employees work performance. The dependent variable is defined as the work performance. Several predictors were added in different steps to the hierarchical model, namely gender and age. Moreover, to control for longer working time, more hours worked and a better coach of an employee, these covariates were added to the model. The results of the hierarchical regression analysis are presented in Table 3. Model B SE of B t p-value Location -3.93 31.57 -0.12 .902 Gender -24.24 39.21 -0.62 .540 Age -0.57 1.37 -0.41 .681 Working Time 0.01 0.10 0.06 .952 Hours Worked 0.00 0.14 0.03 .975 Team lead -0.56 5.27 -0.11 .917 Location * Gender 3.32 25.50 0.13 .897 Location * Age 0.58 0.79 0.73 .471
Table 3. Hierarchical regression analysis for variables predicting work performance (n = 46).
The table shows that the results of the three covariates (working time, hours worked and team lead) are not significant. This implies that these covariates are no predictors of work performance. The R2 shows that 14.46% of the variation in work performance can be explained by the predictors in the model.
H2. Results showed no moderator-effect of gender to the relation between Gamification and Work performance (β = 3.32, t = 0.13, p = .897). This means that this study does not provide
proof to accept the hypothesis that the influence of gamification on work performance of employees is stronger for men than for women.
H3. Moreover, the data do not prove a moderating effect of age on the relation of Gamification and Work performance (β = .58, t = 0.73, p = .471). This implies that the age of an employee does not affect the extent to which their work performance is influenced by gamification. Consequently, the third hypothesis cannot be accepted.
5.3 Conclusion of the analyses
Hence, the analyses of the results show no significant relations between the variables. Since this is in contradiction with the literature on this topic, it is relevant to take a deeper look into the causes of the missing effect. Therefore, interviews were conducted with employees from both locations. These interviews are a way to explore reasons why the proposed hypotheses cannot be accepted. Moreover, the interviews can provide new propositions that are interesting for future research.
