UNIVERSITY OF AMSTERDAM, FACULTY OF SCIENCE
Comparisons between In-Game Behavior, Self-Reported Player
Types, and Game Element Preference
Thesis Master Information Studies, track Human Centered Multimedia
Jerom Fernig (11343702)
jeromfernig@gmail.com
Final version: 26 August 2018
First assessor:Prof. Juho Hamari, Tampere University of Technology
Comparisons between In-Game Behavior, Self-reported Player
Types, and Game Element Preference
Jerom Fernig
Tampere University of Technology & University of Amsterdam
Tampere, Finland jeromfernig@gmail.com
ABSTRACT
The field of gamification lacks literature that considers both player type and game elements. This research therefore makes a com-parison between player types and game element preference by combining self-reported data with in-game behavior from a game specifically developed for this research. All variables are measured in terms of the player types achievement, social. Results show no significant difference between all variables, with large margins. This supports the prediction that players prefer game elements that match their player type. Moreover, the results validate the empirical use of the developed game. Therefore, future studies that experi-ment with or utilize games for data collection are recommended.
1
INTRODUCTION
The new and fast-growing field of gamification is concerned with adding game elements, also known as motivational affordances, to non-game contexts to promote motivation, such as levels, com-petition, progress, cooperation, and avatars. As such, a common research objective is to determine how the positive effects from these game elements can be increased. One way to maximize the motivational impact of game elements is by learning what kind of user prefers what kind of game elements, as Busch et al. [2] point out. A popular approach to differentiate between individuals in both game and gamification literature is with player types.
Hamari and Keronen [3] provide an overview of recent research into motivational affordances in their meta-analysis. However, no clear consensus exists about the exact definition of motivational affordances as the different listed studies vary in the motivational affordances they consider. Additionally, a list of most common mo-tivational affordances according to Jia et al. [5] includes different motivational affordances still. Also, the term motivational affor-danceseems to be a subset of game element. For these reasons, this research focuses on the broader term game elements.
Interestingly, despite the pivotal role of both player type and game elements in game and gamification research, it remains un-clear how these two constructs relate to each other. In a contribu-tion to closing this knowledge gap, this research compares player types with game elements by regarding the difference in preference for game elements of different player types. Thus, the proposed research question is: Which player types prefer what kind of game elements?
1.1
Related Work
In a similar vein as the proposed research, Jia et al. [5] and Karanam et al. [6] have both examined the relationships between person-ality types and motivational affordances. However, both studies differ from this research in terms of the method for differentiat-ing between individuals. Both studies have used the five-factor model to measure personality traits, where this research focuses on the concept of player types because of its prevalence in game and gamification literature.
Another study with a similar subject is that of Tondello et al. [9]. They explore the predictive power of Marczewski’s [8] User Type Hexad, which is a framework used to categorize players that is partly based on player types. In their study, the User Type Hexad is compared with, among others, the five-factor model and various design elements used in games.
Since its origin in 1996 [1], critique on Bartle’s original taxonomy has resulted in a wide variety of player type taxonomies. Hamari and Tuunanen [4] list ten of those taxonomies and conclude that they can ultimately be classified as different shades of the original achiever, socializer, killer, and explorer, except that the player type of immersion was added. Similarly, in his factor analysis, Yee [12] has consolidated the player types into achievement, social, and im-mersion. Subsequent studies [11][10] have validated this approach to player types. Over time, these studies have also yielded increas-ingly accurate measurement instruments for determining player type. Its most recent instance consists of a 12-item survey, where each of the 3 player types is represented by 4 unique items.
As to the game elements, a list of game elements was compiled by Koivisto [7] where each game element is classified in terms of Yee’s player types: achievement, social and immersion. These game elements are considered to fit well into a comparison with Yee’s 12-item instrument, considering the identical player type classification.
The final construct relevant to this study then remains the pref-erenceof the game elements. In their study, Jia et al. [5] use the measures enjoyable, reliable, helpful, and useful to quantify the perception of motivational affordances. Some of these perception measures, namely helpful and useful, overlap. Also, reliable relates more to perception than it does to preference. Therefore, and to preserve the scope of this research, it is proposed to condense these measures into the two variables enjoyment and usefulness. Thus, both the enjoyment and usefulness of game elements will serve as variables for the preference of game elements. Additionally, it might yield interesting results when the two variables enjoyment and usefulness are compared to each other.
1.2
Game
Both methods used by Yee and Koivisto are survey based. While the survey instrument is a staple of psychometric analysis, there is some debate about its ecological validity. Considering game research, the difference between filling out a survey and playing a game becomes instantly apparent. To account for the potential resulting bias, this research proposes a second instrument with indisputable ecological validity by having its respondents play a game, specifically developed for this research. This game captures player behavior and classifies it in terms of Yee’s player types. Moreover, the addition of the game behavior variable prompts the question in what way does the self-reported player type data differ from the game behavior data?
1.3
Research Model
This research aims to resolve which player types prefer which game elements. For the player type construct, two variables were defined: (1) self-reported player type and (2) game behavior player type. For the construct of game element preference, the two defined variables are: (1) enjoyment of game elements and (2) usefulness of game elements, both self-reported. Figure 1a provides an overview of the resulting research model. Furthermore, variables of the same construct will also be analyzed in comparison to each other, as illustrated in figure 1b. Finally, each variable will be measured in terms of three player types achievement, social and immersion. Figure 1c represents the proposed research model in its complete form.
The sum of the comparisons between all variables for each player type adds up to 18 comparisons. Consequently, this results in 18 unique hypotheses. However, since the extension of these hy-potheses is the same, they can be abbreviated into the aggregate expectation that the matching player types of any two variables will not be significantly different from each other.
The next section discusses the methodology that was used to test the hypotheses. Subsequent sections present the results of this method and their implications, respectively.
2
METHODOLOGY
To determine which player types prefer what kind of game elements, the two constructs of player type and game elements have to be considered. In this research, both constructs have been split into two measurable variables. The construct player type was measured using (1) Yee’s [11] 12-item survey (2012) and (2) by measuring in-game behavior. The construct game elements has been split into the self-reported (1) enjoyment and (2) usefulness of game elements, both measured by survey. How these variables were measured is explained in section 2.4 Measurement. All four of these variables have been split into Yee’s player types of achievement, social, and immersion. The comparison of these 4 variables was made for every player type pair. In other words, for every two variables, their achievementdata distribution was compared to each other, as was their social data distribution, and their immersion data distribution. It was hypothesized that the compared distributions would not have significant differences from each other.
Since the functional description of the game is both a vital part of this research’s methodology and somewhat wordy, the description
(a) Initial Research Model
(b) Intermediate Research Model
(c) Complete Research Model
Figure 1: Startup screens
of the game will be discussed first. This is followed by the overviews of the participants, procedure, measurements, and statistical analy-sis of the research.
2.1
Game Description
The goal of the game used for this research is to measure the player type of its players. The game achieves this by offering the players interactive choices that correspond to different player types.
The game can be classified as a short action-adventure game. This genre was chosen as it leaves the most room to implement choices for different player types. During the game, the player
moves an avatar from a 3rd person perspective on a 3D plane that consists of several unique areas. In the left-bottom corner of the screen, a timer runs, indicating the play time. The right-bottom corner displays instructions for playing the game. The right-top displays a minimap for the game space, where areas that are not yet visited by the player are covered up. Throughout the game, the left side of the screen will show a series of icons on which the player can track their progress on various quests. These inferface elements are shown in figure 4e.
At the start, the player is notified of three ways to win the game, as is depicted in figure 5b. These win conditions correspond with the 3 player types. The player can win by (1. achievement) killing sheep; (2. social) helping non playable characters (NPCs); or (3. immersion) exploring the different areas.
Figure 2: Conversational Flowchart for NPC
To complete one or more of these win conditions, the player can have extensive dialogues with NPCs, pick up weapons and swing them, pick up orbs that increase walking speed or jump height, cut down trees, kill sheep or NPCs, and influence the opinions the NPCs have of each other. Figure 2 illustrates the conversational flowchart for a single NPC.
The different areas of the game are:
· a field with some trees, a herd of nine sheep, a hammer to pick up, the four NPCs, and the start location of the player, as shown in figure 3a;
· a forest with a high density of trees, a hidden golden tree and an ax to pick up, as shown in figure 3b;
· a lake with around it a golden tree, a hoe to pick up and a herd of nine sheep, as shown in figure 3c;
· a desert with a sword to pick up and at the end of the desert, both another herd of nine sheep and a golden tree, as shown in figure 3d;
· a mountain range with, hidden behind its tall peaks, a valley with a golden tree, as shown in figure 3e;
Between all areas, small pieces of field can be found, similar to the starting area. Spread over all areas, colored orbs can be found that hover over the ground. Red orbs increase movement speed while blue orbs increase jumping ability.
2.1.1 Player Experience. The player starts in the field area of the game with one of the NPCs right in front of them, as is shown in figure 4b The interface of the game shows when and how the player can talk to this or one of the other NPCs. Players navigate through conversations by clicking any key when an NPC speaks. After the
(a) Field Area (b) Forest Area
(c) Lake Area (d) Desert Area
(e) Mountain Area
Figure 3: Various Game Areas
NPC has finished its sentence, the player can choose their response by clicking on the phrase of choice from a menu and then, the NPC will respond accordingly. An example of such a dialogue menu is shown in figure 5a. During a player’s first conversation, the NPC will explain the goals of the game. After that, the player can ask an NPC about these goals, as is shown in figure 5b, or a number of other subjects including other game elements, their opinions of other NPCs and how the player can help the various NPCs.
The player moves around fairly slow at first. When the player picks up a red orb, a dialogue box informs the player of their in-creased speed. Similarly, picking up blue orbs results in a dialogue box about increased jumping ability.
When the player walks by a weapon, as shown in figure 4a, a dialogue box explains how to pick-up, switch and use the weapon. The game provides visual and audio feedback for when the player hits a tree, sheep or NPC with their weapon. Also, when a sheep is killed, a dialogue box explains that the used weapon has grown in size and strength. The various quests that can be completed during the game each have an icon on the screen that indicates the progress, as is shown in figure 4e. The last feedback system is the minimap. It shows the visited areas from above and is also shown in figure 4e
As explained to the player in their first conversation, there are three ways to win the game. When the player kills ten sheep, a giant sheep will appear, that is shown in figure 4c. If their weapon is strong enough, the player can kill the giant sheep and thereby
(a) Weapon to Pick Up
(b) Starting Point of the Game
(c) Giant Sheep (d) Golden Tree
(e) Drained Lake and Minimap
with Connected Checkmarks (f) NPC Request
Figure 4: Various Game States
(a) Several options during a conversation
(b) Introduction to the Goals of the Game
Figure 5: Dialogue Menu Options
win the game. Secondly, when the player finds and cuts down a golden tree, as shown in figure 4d, a green checkmark will appear over its location on the minimap. When all four golden trees have been found and are cut down, two green lines will have appeared on the minimap, each connecting a pair of checkmarks. At the intersection of these lines is the lake. When the player enters the lake after the lines have appeared, the water will drain from it and a treasure chest will become visible. Both the drained lake and the minimap are shown in figure 4e. Touching this chest also wins the player the game. Lastly, the player can ask all four NPCs if they need help. One lost his hammer, one wants the player to kill a sheep, one wants the player to find and cut down a golden tree, and the last one wants you to change the opinions of the other NPCs so they will like him. The player wins if all four requests have been completed. When the player wins, a dialogue box will congratulate them. When all three win conditions have been met, a special dialogue box appears.
2.2
Participants
Respondents both filled out the survey and played the game for a minimum of 15 minutes. In total, 139 respondents have been regarded in the analysis. Figure 6a and 6b show the age distribution and the male and female portions of the sample respectively. As can be seen, approximately two-thirds of the sample is male and the age distribution roughly follows a normal distribution around a mean of 26.3 with a standard deviation of 6.8. The distribution of the nationality shown in figure 6c. Since a large majority of the sample is Dutch, a similar distribution where Dutch respondents have been excluded is shown in figure 6d. Figure 6e depicts a histogram of the completed education levels among respondents. A plurality of respondents, 41.0%, have a bachelor’s degree and two groups of both 27.3% have master’s degrees and high school diplomas as their highest education. 2.2% of respondents have doctorates and 1.4% has no high school diploma. It is expected that a large sum of respondents with a high school diploma is enrolled in a bachelor’s programme at the current time. The same expectation is made for bachelor’s degrees and enrollment in master’s programmes. In the interest of gamification research, the question How much time do you spend playing video games (including mobile games) on average per week?”was added to the survey. Figure 6f shows the distribution of the answers. Curiously, rounded numbers have been filled in more frequently, relatively. When this is taken into account, the graph resembles a half-normal distribution folded around a point between 0 and 1.
(a) Age Distribution of the Sample
(b) Gender Distribution of the Sample
(c) Nationality Distribution of the sample
(d) Nationality Distribution of the sample excluding Dutch
(e) Educational Background Distribution of the Sample
(f) Average Game Play Time Distribution of Respondents, Including Mobile Games
Figure 6: Various Demographics of the Sample
Sampling has primarily been based on convenience resulting in skewed distributions for age, gender, and nationality. The survey and game were distributed mostly by personal contact and, result-ing from that, indirect word-to-mouth. The survey and game were also distributed using social media communities, such as Facebook1 and Reddit2. Though it is expected that the sum of its respondents accounts for only a small minority of the overall amount of respon-dents. A number of original responses have been excluded because they violated at least one of the following exclusion criteria:
· a self-reported non-sufficient level of English reading com-prehension;
· at least part of the survey uncompleted; and
· insufficient game data, caused by playing less than 15 min-utes while also not completing one of the win conditions in the game.
The survey and game were intentionally not sent to potential respondents that were considered too young, too old or otherwise unable to use the required software of the survey or game.
Before starting the survey or game, respondents are shown a title screen stating:
· the estimated time of the survey;
· minimum required play time for the game; · data is handled anonymously;
· responses are not individually identifiable; · participation is completely voluntary; and · respondents can withdraw at any time.
Respondents have not received compensation of any kind for participating. The title screen of the survey is shown in figure 7
Figure 7: Title Screen of the Survey
2.3
Procedure
Every potential respondent that was contacted for this research has been notified from the start that the survey and game serve an academic purpose. The respondent was then asked to follow a link to Dropbox3where a zip folder of the game can be downloaded either for Windows or Mac. After downloading and unzipping the file, the respondent would then double-click a file named Game for Player Type Researchand a new window would pop up. After choosing graphics settings in this window, the survey starts with the screen shown in figure 7. The respondent progresses through
1https://facebook.com/ 2https://reddit.com/
3The game can be found and downloaded at https://www.dropbox.com/sh/
i0elez6hz09jv60/AAA2HhxRcph64v2UAwoocX-ua?dl=0
the survey by clicking the next button in the bottom right corner. In the survey, subjects are bundled into seven categories. The current category is shown in the top-left corner of the screen and a progress bar on the top of every screen provides an indication of the respondent’s progress through the survey. A full overview of the survey can be found in Appendix A.
On the last screen, the respondent is again asked to play the game for a minimum of 15 minutes or until they complete one of the win conditions. Additionally, on this last screen, it is explained that game controls can be found in the bottom-right corner of the game and that this is purposely the only information provided about the game. Upon completing the survey, the game is loaded. Section 2.1.1 Player Experience describes how the player experiences the game.
A few unexpected events have occurred during data collection. On the first day of data collection, the SMTP4server used to export the data from the game had authorization issues. A new version of the game was quickly released later that day that fixed the issue and added an option to test the connection to the server before filling out the survey. Aside from this, no major problems arose. Since the game is an executable file, some computers, typically work-issued, did not allow opening the file. Also, some respondents experienced difficulties downloading the game from Dropbox, but these issues were typically quickly fixed with some extra instructions.
2.4
Measurement
2.4.1 Survey. The survey has measured the player types and both the enjoyment and usefulness of the game elements. For the player type items, a 12-item survey from Yee et al. [11] was used. The game elements are based on Koivisto [7] including their clas-sification of game elements as achievement, social, and immersion. All items are recorded using a 5-point Likert scale. For player types, the question asked is How important are these gameplay elements when you play online games?and the scale ranged from completely unimportant(1) to very important (5). The 12 items are:
· Becoming powerful (Achievement) · Acquiring rare items (Achievement)
· Optimizing your character as much as possible (Achieve-ment)
· Competing with other players (Achievement) · Chatting with other players (Social)
· Being part of a guild (Social) · Grouping with other players (Social) · Keeping in touch with your friends (Social)
· Learning about stories and lore of the world (Immersion) · Feeling immersed in the world (Immersion)
· Exploring the world just for the sake of exploring it (Immer-sion)
· Creating a background story and history for your character (Immersion)
The bracketed player type behind each item was not featured in the survey.
The question for the enjoyment of game elements was Please rate howenjoyable you consider these game elements to be in a game. and
4Simple Mail Transfer Protocol 5
the scale ranged from completely unenjoyable (1) to very enjoyable (5). Its items were:
· Points, score, experience points (Achievement)
· Challenges, quests, missions, tasks, clear goals (Achieve-ment)
· Badges, achievements, medals, trophies (Achievement) · Competition (Achievement)
· Leaderboards, rankings (Achievement) · Levels (Achievement)
· Performance stats, performance feedback (Achievement) · Progress, status bar, skill trees (Achievement)
· Quizzes, questions (Achievement) · Timer, speed (Achievement) · Increasing difficulty (Achievement) · Retries, health, health points (Achievement) · Social networking features (Social) · Cooperation, teams (Social) · Peer-rating (Social)
· Customization, personalization (Social) · Multiplayer (Social)
· Collective voting (Social)
· Assistance, virtual helpers (Social)
· Avatar, character, virtual identity (Immersion) · Narrative, storytelling, dialogues (Immersion) · Virtual world (Immersion)
· In-game rewards (Immersion) · Role play (Immersion)
Again, the bracketed player types were not part of the survey. The question for the usefulness of game elements and its items are identical to the enjoyment except enjoyment is changed into useful and its scale ranged from completely useless (1) to very useful (5).
Other self-reported data included general demographic questions at the start of the survey, that include gender, age, highest completed education, occupation, nationality, and English reading proficiency. Also, they inquire after the average time spent playing video games per week and the most played genre video games of the respondent. 2.4.2 Game. The game collects data using a set of 21 triggers. These triggers are activated throughout the game without the player knowing about it. For example, every time the player starts a conversation, a specific trigger is activated. When a trigger is activated, the action that set it off is recorded along with a time stamp. These triggers have all been classified as one of the three player types. The 21 triggers are:
· pick-up a speed orb (Achievement) · pick-up a jump orb (Achievement) · kill a sheep (Achievement)
· spawn the giant sheep (Achievement) · kill the giant sheep (Achievement)
· choose to talk about killing sheep (Achievement) · kill an NPC (Achievement)
· start a conversation (Social) · ask if an NPC needs help (Social)
· influence the opinion of an NPC about another NPC (Social) · choose to talk about helping (Social)
· complete an NPC request (Social) · completed all four NPC requests (Social)
· uncover a new area on the minimap (Immersion) · pick-up a weapon (Immersion)
· find and cut down a golden tree down (Immersion) · choose to talk about finding treasure (Immersion) · drain the water from the lake (Immersion) · find the treasure (Immersion)
· complete all three win conditions (Immersion)
Every 5 minutes, all data from the current play session is exported through an SMTP server. This also happens the moment that the player completes one of the win conditions.
2.5
Statistical Analysis
The statistical analysis consisted of four steps: (1) raw data pro-cessing; (2) normalizing data per player type; (3) measuring normal distribution; and (4) comparing variables by their player type com-ponents.
The raw data was imported from the SMTP server. For every response, the 21 game triggers were tallied. These triggers are henceforth considered as items. Then, for every item, the mean and standard deviation were calculated. Once those were known, the data was normalized. Since the mean and standard deviation of the general population are unknown, the Student’s t-statistic was used to normalize the data:
¯ X − x
s
where x is the data point, ¯X is the mean of the item and s is the standard deviation of the item.
This normalized data was then averaged for every player type of every variable. For example, the variable self-reported player typehas 4 items that are classified as achievement. The normalized data points of these 4 items were averaged for every response. This was also done for this variable’s 4 social items and its 4 immersion items. The same happened for the enjoyment and usefulness of game elements and the tallied up triggers from the game. The full tables of averaged data can be found in Appendix B.
This resulted in 12 distributions of data: 3 player types each for 4 variables. Histograms of these data distributions are shown in figure 8. The data points in these distributions are indicators of the player type per respondent measured with one of the 4 variables. To compare these data distributions, paired t-tests would normally be used.
However, the paired t-test has 4 assumptions: (1) the data must be numerical and continuous; (2) observations have to be independent of each other; (3) the data needs to be normally distributed; and (4) the data must have no outliers.
The data follows both the first and second assumption, but at this point, it is unclear whether the data is in violation of the third assumption. To check whether the data distributions are normally distributed, a Shapiro Wilk Normality test was performed. Lastly, data distributions that were normally distributed were checked on outliers because of the fourth assumption.
If two data distributions of the same player type meet all 4 as-sumptions, a comparison between them was made using the paired t-test. If these 4 assumptions are not all met, a comparison was
(a) Data Distribution of the Self-Reported Player Type Data
(b) Data Distribution of the Game Behavior Data
(c) Data Distribution of the Enjoyment of Game Elements Data
(d) Data Distribution of the Usefulness of Game Elements Data
Figure 8: Histograms of the Data Distributions of All 4 Vari-ables
made using the non-parametric version of the paired t-test, the Wilcoxon signed rank test.
2.6
Software
The game has been developed in Unity5. Its scripts have been written in Microsoft Visual Studio6using C#7. The 3D models in the game have been made in Blender8. The survey and game have been packed together in an executable Unity program. For the distribution of the game, Dropbox9was used. For the SMTP server, SendGrid10and Gmail11have been used.
Microsoft Excel12was used for data gathering, cleaning, tally-ing, normaliztally-ing, and averaging the data into the 12 distributions presented in figure 8. SPSS13has been used for the Shapiro Wilk Normality test, the paired t-test, the Wilcoxon signed rank test and to check the data for outliers.
3
RESULTS
Table 1 presents the results of the Shapiro Wilk Normality test for each player type of the variables self-reported player type, game behavior, enjoyment of game elements, and usefulness of game ele-ments.
These results show that the social components of the variables enjoyment of game elements and usefulness of game elements both have a significance level greater than 0.05 and can thus be consid-ered normally distributed, with 95% confidence. It would seem that
5https://unity3d.com/
6https://visualstudio.microsoft.com/
7Developed by .NET https://www.microsoft.com/net 8https://www.blender.org/ 9https://www.dropbox.com/ 10https://sendgrid.com/ 11https://www.google.com/gmail/about/ 12https://products.office.com/excel 13https://www.ibm.com/analytics/spss-statistics-software
Variable Significance (p-value) Self-reported player type (Achievement) 0
Self-reported player type (Social) 0 Self-reported player type (Immersion) 0 Game behavior (Achievement) 0.001 Game behavior (Social) 0 Game behavior (Immersion) 0 Enjoyment of game elements (Achievement) 0.37 Enjoyment of game elements (Social) 0.018 Enjoyment of game elements (Immersion) 0.001 Usefulness of game elements (Achievement) 0.162 Usefulness of game elements (Social) 0.003 Usefulness of game elements (Immersion) 0.008
Table 1: Shapiro-Wilk Normality test
a paired t-test could be used to compare these variables. However, the social component of the usefulness of game elements contains outliers and is thereby in violation of the fourth assumption of the paired t-test. Therefore, all comparisons were calculated using the non-parametric Wilcoxon signed rank test.
First variable Second variable Z-value p-value Self-reported player type (Achievement) Game behavior (Achievement) -0.242b 0.809 Self-reported player type (Social) Game behavior (Social) -0.347b 0.729 Self-reported player type (Immersion) Game behavior (Immersion) -0.153b 0.878 Self-reported player type (Achievement) Enjoyment of game elements (Achievement) -0.029c 0.977 Self-reported player type (Social) Enjoyment of game elements (Social) -0.023b 0.982 Self-reported player type (Immersion) Enjoyment of game elements (Immersion) -0.080b 0.936 Self-reported player type (Achievement) Usefulness of game elements (Achievement) -0.210b 0.833 Self-reported player type (Social) Usefulness of game elements (Social) -0.141c 0.888 Self-reported player type (Immersion) Usefulness of game elements (Immersion) -0.139c 0.890 Game behavior (Achievement) Enjoyment of game elements (Achievement) -0.402b 0.688 Game behavior (Social) Enjoyment of game elements (Social) -0.114b 0.910 Game behavior (Immersion) Enjoyment of game elements (Immersion) -0.248b 0.804 Game behavior (Achievement) Usefulness of game elements (Achievement) -0.175b 0.861 Game behavior (Social) Usefulness of game elements (Social) -0.166b 0.868 Game behavior (Immersion) Usefulness of game elements (Immersion) -0.223b 0.824 Enjoyment of game elements (Achievement) Usefulness of game elements (Achievement) -0.189c 0.850 Enjoyment of game elements (Social) Usefulness of game elements (Social) -0.692c 0.489 Enjoyment of game elements (Immersion) Usefulness of game elements (Immersion) -0.584b 0.559
Table 2: Wilcoxon signed rank test
Table 2 displays the calculated comparisons. No variable pairs are significant at a 95% confidence level. Indeed, these results seem especially far removed from the threshold required to be significantly different. This is further illustrated in figure 9 where the p-values are visually represented for all 3 player types of all 6 variable pairs.
Figure 9: Visually Illustrated p-values for all Variable Pairs
4
DISCUSSION
In this research, the 4 variables (1) self-reported player type, (2) game behavior, (3) enjoyment of, and (4) usefulness of game ele-ments have been measured of 139 respondents with both a survey and a game. Every variable was split into the three player type components achievement, social, and immersion and was compared to each other for those three player types. The statistical tests proved that, for all 18 comparisons, the two variables do not differ significantly, as is presented in table 2. Most variable pairs are even far removed from any statistical evidence that they differ from each other. Thus, it can be stated, with more certainty, that one variable predicts the other variable and vice versa, for all 18 variable pairs of the same player type. With regard to the hypotheses, there is no evidence that they should be rejected.
4.1
Interpretation
All variable pairs are not significantly different from each other. Interestingly, while this was expected, the p-values in most vari-able pairs is relatively high, meaning that the difference between the variable pairs is not even close to a significant level. Only 4 pairs out of 18 yielded a p-value under 0.8. Out of those 4 pairs, only one p-value is lower than 0.5 and only barely at 0.489. This consideration strengthens the hypotheses even more. Interestingly, 3 of those 4 variable pairs are not pairs between player type and game elements, but rather one among player type variables (i.e. the socialcomponents of self-reported data and game behavior data) and two among game element variables (i.e. both the social and immersioncomponents of the enjoyment and usefulness of game elements). In conclusion, applying the described methodology, no evidence was found that player type does not predict game element preference. Or, in terms of the research questions, no evidence was found that suggest that players do not prefer game elements that are classified as their own player type. Thus, these results strengthen the expectation that players do prefer the game elements that are classified as their own player type.
Out of the 18 variable pairs, 12 were comparisons between a player type variable and a game element variable. Table 2 shows that the p-values of these 12 comparisons are remarkably higher than those of the 3 comparisons between the two player type vari-ables or the 3 comparisons between the two game element varivari-ables. The highest p-values were found between the variables self-reported player typeand enjoyment of game elements of 0.977 for achievement, 0.982 for social, and 0.936 for immersion. Variables self-reported player typeand usefulness of game elements have slightly lower p-values at 0.833, 0.888, and 0.890 for achievement, social and im-mersionrespectively. Also high are most p-values between the game behavior variable and both the enjoyment and usefulness of game elements. All but one of those p-values are higher than 0.8. Only the pair between game behavior and enjoyment of game elements for their achievement component is marginally lower at 0.688, which is still far removed from the significance level of 0.05. These high p-values implicate that player type and game element preference are relatively reliable indicators for each other.
High p-values were also found between the two player type vari-ables with values between 0.7 and 0.9. Interestingly, the variable
pairs with the lowest p-values are the social and immersion compo-nents of the enjoyment and the usefulness of game elements. This forces the conclusion that social- and immersion-oriented players perceive a higher variance between the enjoyment and usefulness of game elements, as compared to achievement-oriented players.
4.2
Implications
By not finding significant evidence to reject the hypotheses, this research has not yielded groundbreaking new views. Rather, it validates the existing views that have been the foundation or this research. Firstly, the ontological use of Yee’s [12] player types is validated. The same is true for the player type classification of game elements by Koivisto [7].
Additionally, the use of the game itself and the data triggers in it are validated. If game data would have been significantly different from game element data or self-reported player type data, then it could be argued that players behave differently from what they report in a survey. But, in such a hypothetical scenario, a case could also be made for poor measurement validity of the game. However, since this is not the case, by a large margin, it can be concluded that the game measured what it intended to. Or, at a bare minimum, the game is affected by the same biases as the survey is.
4.2.1 Theoretical Implications. Henceforth, the constructs of player type and game elements can now be used more reliably as indicators or even proxies for each other. This can also be done retroactively when considering past research. Additionally, game elements can be grouped more easily and with more certainty in future research. Lastly, this research offers considerable indicators that self-reported player types match the player’s behavior in-game for the action-adventure genre, and possibly other genres. This implicates high ecological validity for this kind of survey data. Alternatively, these conclusions can be used to promote the use of games as a means of data collection. While the investments required for game development are substantially larger than survey design, having respondents play games does have its advantages over survey-like instruments. For one, respondents’ motivation may be higher to participate and complete the experiment, as this is the core premise of gamification.
4.2.2 Practical Implications. The practical application of gamifi-cation, in most cases, is to promote motivation. As such, research in this field should ultimately serve that purpose, be it directly or in-directly. On account of this research, applications can be developed that better match their game elements to their target audience. Both the comparison between player types and corresponding game ele-ments as well as the classification of the game eleele-ments can prove useful for this purpose since players that prefer certain game el-ements are more likely to also prefer game elel-ements of the same player type class. Considering game development, this research can be used to build player profiles more easily or more accurately. For example, by basing player types on interaction with specific game elements, in addition to other variables. Since the predictive power between the variables goes both ways, player type analysis can also be used more confidently to determine what game elements are preferred by target audiences.
4.3
Limitations
Low hanging fruit in the tree of limitations for myriad studies are sample conditions. As discussed in section 2.2 Participants, this research has been based on 139 responses, after applying exclusion criteria. The size of this sample can be viewed in two different lights. A sample size of 200 is a generally accepted standard for psychometric analyses. However, calculating the required sample size for the performed Wilcoxon signed rank test with the final mean values gives a minimum of 44 respondents for a power higher than 0.80. Perhaps a more clear limitation of the sample is its con-venience nature. In all probability, this has caused the distribution of age of the respondents to bundle around a mean of 26.3 with a relatively small standard deviation of 6.80. The potential impact of this limitation hinges on the question does player type vary with age?A similar question can be asked for gender and educational background, though both demographics are more proportionally represented in the sample. The last general demographic, national-ity, is also disproportionally represented in the sample, expectedly also as a result of the convenience sampling. All in all, since the results of this research are far removed from the statistical thresh-old for rejecting the hypotheses, these limitations are regarded as important, though insignificant with regard to the conclusions.
At first sight, it might seem inappropriate to base game related conclusions on a body of respondents where a plurality reported to play video games for zero or one hour on average weekly. It should, however, be considered that, within a rounding error, any respondent is sufficiently familiar with games to be aware of their own preferences, as games in various forms are all around us. Ad-ditionally, implications of this research are primarily directed at gamification, where non-gamers are as important as gamers.
As explained in section 2.4 Measurement, the game behavior data was gathered by triggers in the game that were set off by certain actions of the player. The three player types were all represented by 7 triggers each. While the statistical results of the research validate the classification of these triggers as a whole, it is difficult to conclude whether individual triggers were classified optimally. Lastly, as mentioned in section 2.3 Procedure, there were three minor unexpected events during data collection. (1) For the first few hours were the game was online, the SMTP server did not accept survey data, (2) the game could not be opened on some, typically work-issued, computers, and (3) some respondents had trouble downloading the files from Dropbox. All three instances were quickly dealt with. Thus, it is not believed that they affected the data in any significant way.
4.4
Future Research
In the analysis of this research, some interesting connections be-tween data series have been left out of scope. For instance, future research could compare the demographic data with both the player type data and the game element preference data. In this regard, combining data with the survey items about weekly play time and game genre preference might also yield interesting insights.
Another possibility with the current dataset is comparing differ-ent player type compondiffer-ents of the 4 variables. For example, a com-parison can be made between the social component of self-reported
player type and the immersion component of the enjoyment of game elements. For such an analysis, low p-values would be expected.
The survey also featured three more questions, each of about 30 items. These questions inquire after another player type taxonomy and pleasurability of certain game aspects. Due to time constrains, these questions have not been analyzed in this research and are recommended targets for subsequent research. Data from these questions can be analyzed on its own as well as be combined with the variables from the current analysis. These questions have been included in the complete survey in Appendix A.
A few directions for future research can be derived from the limitations of this research. Namely to increase sample size and improve population representation while copying the methodology of this research. This would possibly also cause the new sum of data to be normally distributed, allowing for the parametric paired t-test. As to the game behavior triggers, subsequent studies are encouraged to experiment with different triggers and with different classifications of the triggers to find comparisons with even higher p-values.
As discussed in section 1 Introduction, there is a lack of research that compares player types with game elements. This research can be considered a starting point. It is recommended that future research focuses on the same two constructs while using other variables to represent them, such as other player type taxonomies or game elements that fall into the stricter definition of motivational affordance. Research that uses different player type taxonomies could additionally create insight into how the level of specificity of the player types impacts the conclusions. Concerning game elements, it might prove insightful to experiment with different classifications of the game elements, like adding classes or even treating every game element as its own class. The same can be done for the game behavior triggers.
A final suggestion for future research is to experiment with different game genres. For this game, the action-adventure genre was chosen because it is believed to optimally fit Yee’s player types. Future research could then use this research as a baseline to find out how the genre impacts the player’s behavior, from the player type point of view.
4.5
Conclusions and Recommendations
In summary, player type, both self-reported and in-game, does not differ significantly from game element preference, measured in enjoyment and usefulness. Thus, with regard to the research question ”Which player types prefer what kind of game elements?”, it can be stated with more confidence that player types prefer their equivalent game elements. Because of this conclusion, future research can use either construct as a more reliable indicator for the other. Still, it is highly recommended that more research is carried out on these two constructs to both validate these findings and add nuance to the classifications of the variables used in this research. Additionally, insignificant differences between the variables also validate the game as a research instrument. This has implications that range from survey instruments with higher ecological validity to more reliable player type profiles as a result of player behavior. To solidify these implications, future research into player behavior
is recommended that experiments with the game genre in relation to game element preference.
REFERENCES
[1] Richard (MUSE Ltd) Bartle. 1996. Hearts, Clubs, Diamonds, Spades: Players who suit MUDs. Journal of MUD research 6, 1 (1996), 39. DOI:http://dx.doi.org/10. 1007/s00256-004-0875-6
[2] Marc Busch, Elke Mattheiss, Rita Orji, Andrzej Marczewski, Wolfgang Hochleit-ner, Michael Lankes, Lennart E. Nacke, and Manfred Tscheligi. 2015. Personal-ization in Serious and Persuasive Games and Gamified Interactions. Proceedings of the 2015 Annual Symposium on Computer-Human Interaction in Play(2015), 811 – 816.
[3] Juho Hamari and Lauri Keronen. 2017. Why do people play games? A meta-analysis. International Journal of Information Management 37, 3 (2017), 125–141. DOI:http://dx.doi.org/10.1016/j.ijinfomgt.2017.01.006
[4] Juho Hamari and Janne Tuunanen. 2014. Player Types : A Meta-synthesis. Transactions of the Digital Games Research Association1, 2 (2014), 29. [5] Yuan Jia, Bin Xu, Yamini Karanam, and Stephen Voida. 2016. Personality-targeted
Gamification: A Survey Study on Personality Traits and Motivational Affor-dances. Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems - CHI ’16(2016), 2001–2013. DOI:http://dx.doi.org/10.1145/2858036. 2858515
[6] Yamini Karanam, Hanan Alotaibi, Leslie Filko, Elham Makhsoom, Lindsay Kaser, and Stephen Voida. 2014. Motivational affordances and personality types in personal informatics. Proceedings of the 2014 ACM International Joint Conference on Pervasive and Ubiquitous Computing Adjunct Publication - UbiComp ’14 Adjunct (2014), 79–82. DOI:http://dx.doi.org/10.1145/2638728.2638800
[7] Jonna Koivisto. 2017. Gamification : A study on users, benefits and literature. Ph.D. Dissertation. Tampere University of Technology.
[8] Andrzej Marczewski. 2013. The Intrinsic Motivaion RAMP. (2013).
[9] Gustavo F. Tondello, Rina R. Wehbe, Lisa Diamond, Marc Busch, Andrzej Mar-czewski, and Lennart E. Nacke. 2016. The Gamification User Types Hexad Scale. Proceedings of the 2016 Annual Symposium on Computer-Human Interaction in Play - CHI PLAY ’16(2016), 229–243. DOI:http://dx.doi.org/10.1145/2967934.2968082 [10] Nick Yee. 2016. The gamer motivation profile: what we learned from 250,000
gamers. Proceedings of the 2016 Annual Symposium on Computer-Human Interac-tion in Play(2016), 2–2.
[11] Nick Yee, Nicolas Ducheneaut, and Les Nelson. 2012. Online gaming motivations scale. Proceedings of the 2012 ACM annual conference on Human Factors in Computing Systems - CHI ’12(2012), 2803. DOI:http://dx.doi.org/10.1145/2207676. 2208681
[12] Nicholas K Yee. 2005. Motivations of Play in MMORPGs. DiGRA 2005 Conference: Changing Views–Worlds in Play(2005). http://www.digra.org/dl/db/06276.26370. pdf
5
APPENDIX
5.1
Survey Items
Figure 10 depict the survey items in the context that the respondents have experienced them.
(a) Initial Survey Screen
(b) General Questions 1
(c) General Questions 2
(d) Yee’s 12-Item Survey
(e) Pleasurability of Game Qualities 1(not used in this re-search)
(f) Pleasurability of Game Qualities 2(not used in this re-search)
(g) Pleasurability of Game Qualities 3(not used in this re-search)
(h) Motivation for Playing Video Games 1(not used in this research)
(i) Motivation for Playing Video Games 1(not used in this research)
(j) Importance of Game Characteristics 1(not used in this re-search)
(k) Importance of Game Characteristics 2(not used in this research)
(l) Enjoyment of Game Elements 1
(m) Enjoyment of Game Elements 2
(n) Usefulness of Game Elements 1
(o) Usefulness of Game Elements 2
(p) Final Survey Screen
Figure 10: Startup screens
5.2
Data Distributions for all 4 Variables
Tables 3, 4, 5, and 6 present the averages per respondent of the normalized survey and game data.
Self-Reported Player Type
Resp. Achievement Social Immersion 1 1.178173884 -0.7189580934 1.205013907 2 -0.9835186116 -0.1760645652 -0.8318983798 3 1.38016107 0.5383042424 -0.04835841145 4 0.7354079916 1.070163979 -0.7989382444 5 -1.168972404 -1.420421169 -1.155352739 6 -1.168972404 0.5284057897 -0.6302827412 7 0.9974783398 -0.3536944173 1.205013907 8 -0.98827686 -0.7041804216 -0.2372010467 9 -1.168972404 -0.3558118908 -0.6111792529 10 0.1597713298 0.7060356418 -0.2648198705 11 1.161640491 -0.3472943109 0.1068014354 12 -0.02044097141 0.536677836 -0.06402135552 13 -0.98827686 -0.8888069663 -1.003994427 14 -0.5672858523 -1.601549367 -1.734653648 15 1.96055598 0.3621952775 0.8475157687 16 -1.168972404 -0.522022403 -1.192019939 17 -1.168972404 -1.242545783 -1.155352739 18 -0.98827686 -0.1824646716 0.8352933688 19 -1.168972404 1.076564085 0.3176550627 20 -0.98827686 0.01170927321 -0.6387980768 21 1.178173884 0.3638216839 0.4623988064 22 1.560856614 0.3475231248 0.4623988064 23 1.303544514 0.8909077201 0.4743546861 24 1.178173884 0.8909077201 -0.1012391588 25 0.959170062 -0.009608519452 0.285778571 26 -0.232427897 -1.601549367 -1.204792942 27 -0.205894764 1.249420237 0.1211142154 28 0.9639283104 1.069918445 1.205013907 29 1.358869428 0.7079075817 0.831035701 30 1.358869428 0.5430779424 0.6586731334 31 -0.0587492492 0.5267793833 0.3011749949 32 -0.008665826641 0.3493950647 0.1014601237 33 1.178173884 -0.1776909716 -0.08819963518 34 -1.168972404 -1.420421169 -0.2969628823 35 1.358869428 1.249420237 0.1046341477 36 0.1337214397 0.00368276044 0.285778571 37 -0.3865903082 0.7095339881 0.2977344508 38 1.325319399 0.01170927321 0.8432581009 39 -1.168972404 0.7095339881 0.2692985033 40 -1.168972404 -1.601549367 -1.734653648 41 -0.4319154823 -0.1744381588 0.6467172537 42 -0.6173692749 -1.249191423 -0.08901675885 43 -0.9717434668 -0.188864778 -0.2568551384 44 0.7236328468 0.536677836 0.8275951568 45 -0.6268857717 0.536677836 0.4536169506 46 -0.179361631 -0.1843366115 1.028393672 47 -0.02044097141 0.3555496375 0.4700970183 48 -0.98827686 -0.5350681493 0.8275951568 49 -0.2229114002 0.5323952031 0.6544154656 50 -0.9717434668 0.3526478774 -1.37715551 51 0.3957917879 1.249420237 1.028393672 52 0.5264039095 0.3590479838 -0.05976368771 53 -0.7862896743 -0.8885614328 0.293476783 54 0.7236328468 1.249420237 1.028393672 55 -0.5672858523 0.5302777296 0.6631973214 56 0.3692586549 0.7097795217 0.4623988064 57 -0.8075813159 -0.8901878392 -0.2568551384 58 -0.9499685823 -1.601549367 -1.558033413 59 0.203321099 0.5451954158 0.285778571 60 -1.168972404 -0.1741926253 -0.4371824381 61 -0.3865903082 1.069918445 0.4985329658 62 -0.6055941301 1.249420237 0.1168565476 63 -0.2394447934 -1.069935165 0.4623988064 64 -0.2346865449 -1.601549367 -0.6621592328 65 0.1602545727 0.3510214711 -0.9997367594 66 -0.2442030418 0.8871638402 0.1125988798 67 -0.02519921984 -0.005080352983 0.2892191152 68 -0.5838192455 1.070163979 -0.06746189966 69 -0.7910479227 -1.601549367 -1.180614663 70 0.5764873321 0.8906621865 -0.2571216586
Table 3: Averaged Data per Respondent for Self-Reported Player Type
cont. Self-Reported Player Type
71 1.139865606 0.005063633264 0.8352933688 72 -0.98827686 1.070163979 -0.2528639908 73 1.009253485 1.069918445 1.205013907 74 1.144623855 0.3472775912 -0.835338924 75 -1.168972404 -0.8824068599 -1.188579395 76 -0.7527396449 0.7111603945 -0.6215008854 77 0.7519413848 0.7159340945 -0.2770422705 78 0.06795072895 1.249420237 -0.4701425735 79 0.9644115534 0.1826934854 0.6467172537 80 -0.2512199381 -1.601549367 -1.734653648 81 0.1985628505 -0.003208413057 -0.07171956746 82 1.128090461 -0.001336473131 0.4658393505 83 0.7571828762 0.7178060344 0.4700970183 84 1.166398739 0.5331794897 0.1003764799 85 0.3957917879 0.7095339881 0.4777952303 86 -0.3865903082 0.7095339881 0.288952595 87 1.199948769 -0.1891103115 1.011913604 88 1.522548336 0.1701388062 0.4815022946 89 0.7401662401 -0.3550752901 0.2804372594 90 -1.168972404 -0.001582006682 0.4700970183 91 -0.6008358817 -0.1763100987 0.07963874439 92 -0.4031237014 0.8973078265 -0.6146197971 93 0.1985628505 -0.003208413057 -0.08819963518 94 0.1820294573 -1.601549367 0.8432581009 95 -0.5838192455 -0.3507926572 0.0838964122 96 0.7619411246 0.3571760439 0.1014601237 97 1.151640751 -0.001582006682 1.205013907 98 -0.4031237014 0.003437226889 -1.184055207 99 -0.98827686 -1.422047575 -1.204792942 100 0.001333913181 -0.5284225094 0.4700970183 101 0.959170062 0.5385497759 -0.622318009 102 -0.03745760756 0.7116514616 1.028393672 103 0.2416293768 -0.345422371 0.1080746919 104 0.1602545727 0.7178060344 1.011913604 105 0.4175666725 1.249420237 0.4820528981 106 0.203321099 -0.3638384036 0.6544154656 107 -0.8075813159 -0.8888069663 0.6421930657 108 1.166398739 -0.1872383716 0.4908347539 109 -1.168972404 -1.601549367 -1.734653648 110 -0.1841198794 0.8909077201 -0.06746189966 111 -0.8075813159 -1.068063225 0.2727390475 112 0.959170062 -0.3555663573 0.285778571 113 -0.7645147897 -0.8888069663 0.09611881211 114 -0.3648154236 -1.601549367 -1.204792942 115 -1.168972404 -0.88530862 0.3133973949 116 -0.8075813159 -0.8869350264 0.6509749215 117 -0.008665826641 -1.422047575 -1.204792942 118 -1.168972404 -0.3451768375 -1.204792942 119 -0.4131234412 -0.00109093958 -1.558033413 120 0.3740169033 -0.01148045938 -0.4698760534 121 -1.168972404 -1.601549367 -1.558033413 122 -0.03044071123 0.1776742518 0.4448350948 123 -0.008665826641 0.3603233376 0.1003764799 124 0.3622417585 -0.7076787679 -0.2847404824 125 1.358869428 0.5350514296 0.4666564742 126 0.7619411246 0.7111603945 0.2692985033 127 0.7619411246 0.7111603945 0.2692985033 128 -0.3818320597 -0.1810837988 -0.4579201736 129 -1.168972404 0.003191693338 -0.05976368771 130 -0.07052439397 -0.8952070727 -0.8185923361 131 -1.168972404 -1.601549367 -0.06746189966 132 0.5764873321 0.8909077201 1.028393672 133 1.363627677 0.176293379 -0.06402135552 134 0.5429373027 1.070163979 0.4623988064 135 -0.4461902276 -1.601549367 -1.028172707 136 -0.02519921984 0.8887902466 -0.4536625058 137 0.9809449466 0.3569305104 1.205013907 138 -1.168972404 1.069918445 -0.4371824381 139 -0.3865903082 0.005063633264 1.028393672
Table 3: Averaged Data per Respondent for Self-Reported Player Type
Game Data
Resp. Achievement Social Immersion 1 1.448962282 -0.6074478377 0.2816891537 2 0.5491124628 -0.667796112 -0.6270658197 3 0.4569917577 0.6075344806 0.5968733507 4 -0.8152043356 -0.2660116711 -0.395068781 5 -0.205518423 -0.05439399823 0.1874056707 6 -0.8055731678 1.209969599 -0.7578149906 7 0.1330858028 -0.7029320985 -0.4134541726 8 -0.2299046217 -0.3563506606 0.5715290927 9 0.1443638884 -0.7029320985 -0.8880513039 10 -0.4350848092 -0.2816201101 0.7373998215 11 0.4885638272 0.1056774955 -0.3134412507 12 -0.5828901348 -0.6790610333 0.2915439767 13 -0.3767324204 1.173492389 -0.03334374865 14 -0.9348190202 -0.5597057072 -0.6069983623 15 0.5784837554 -0.1580258255 -0.2222015184 16 -0.7915597216 -0.04043256552 -0.5983902707 17 0.2133955093 -0.7029320985 -1.049659162 18 0.5541378639 0.4210475562 0.5221299681 19 1.20485579 -0.6790610333 0.9637229419 20 -0.832675366 0.1273665362 -0.7114668686 21 0.229698996 -0.5279538111 -0.5986466995 22 0.3993576304 1.518732605 1.571699856 23 1.353756327 1.914012588 1.664732856 24 0.113908115 0.1911229977 -0.5434340572 25 0.03356070389 0.04446539945 0.375077161 26 -0.2709093145 -0.5392187324 -0.7948422024 27 0.6480350576 -0.5153476672 0.1780847605 28 0.5231327614 -0.3802217258 0.8080620432 29 -0.3797942615 0.1539441433 -0.3906907196 30 -0.06487111506 0.1250769672 -0.03084046665 31 -0.484020955 -0.7029320985 0.1868717578 32 -0.3466822283 0.04339091239 0.1162472036 33 0.6922802066 -0.6790610333 0.1149229491 34 -0.4025335424 -0.04001172202 -0.5679879735 35 0.3348402213 1.856083036 1.203606338 36 -0.003795819712 -0.3167560966 0.7678021187 37 -0.2164394879 0.1561291533 -0.02923217223 38 0.4272139029 -0.7029320985 -0.1331140495 39 0.08209877948 -0.6790610333 -0.6923121908 40 0.4097303828 -0.5189895396 -0.1865999437 41 -0.625665151 -0.6200539815 -0.8912036892 42 0.4467856989 0.2396728797 0.07581117765 43 -0.9689279415 0.7882521133 -0.3288531627 44 -0.8745132339 0.3644672949 -0.4836433924 45 0.1041755765 -0.2734725437 -0.2130805693 46 -0.3303416184 -0.5961829163 -0.4220622641 47 -0.02976530573 -0.6790610333 0.2606077667 48 -0.4424668974 -0.6790610333 -0.5618194487 49 1.587617435 1.075934071 1.190932178 50 0.567595074 -0.2734725437 -0.2393730193 51 -0.616002168 -0.5961829163 -0.2349312426 52 0.04587338209 -0.2219263644 1.124972988 53 0.2368763747 0.09671322393 -0.5150149369 54 1.089956065 -0.4563406155 0.3661112579 55 0.9850810121 0.3848010468 1.013298168 56 0.4344634397 0.7345139058 1.022266875 57 0.2099503792 -0.03364028792 0.484488516 58 -0.7949257508 -0.7029320985 0.2417870407 59 -0.6675502689 -0.3419683844 0.1103988228 60 0.7653597017 0.08029001518 0.7774808403 61 -0.3357487094 0.4027206363 0.1452494525 62 -0.775735609 0.9381658278 -0.4846126397 63 0.04487649406 -0.5835767724 -0.8445991385 64 0.4419272671 -0.3440112522 0.5838234824 65 -0.6263749204 0.0741833813 0.5084382484 66 1.118285475 1.165037944 0.8453456838 67 -0.5233383871 -0.1775663891 -0.3062896708 68 0.4430852654 0.7831068419 0.517685387 69 -0.8552315371 -0.3897105148 0.3747193495 70 0.1924862559 0.1823008683 0.02232528349
Table 4: Averaged Data per Respondent for Game Behavior
cont. Game Behavior
71 0.9125769759 -0.2993864767 0.279249587 72 0.3505727791 -0.2590902674 0.4904354752 73 0.2159795754 0.9953820558 0.4899015623 74 0.1488820266 0.6056427085 0.235084603 75 -0.745989605 0.512879941 0.6488937521 76 -0.3412000696 0.5646708861 0.751244858 77 -0.8028378071 0.3617577674 0.5603025287 78 -0.8596860093 -0.6200539815 -0.4808836816 79 0.1480045822 -0.6790610333 -0.7575585618 80 0.2108750735 -0.2032005706 -0.4320959929 81 -0.5423329652 0.3197003772 0.7334919576 82 -0.5748143299 0.1581720211 0.3655773451 83 -0.746986493 -0.7029320985 -0.8445991385 84 -0.216451942 -0.7029320985 -0.6798388953 85 -0.2708331943 -0.2090605033 0.5803095341 86 -0.5155178857 -0.3883692922 0.1123819997 87 1.444958677 -0.7029320985 0.3548270463 88 0.1802555872 -0.7029320985 0.1665069409 89 1.162489575 0.02086106975 -0.297938008 90 -0.6594869537 -0.7268031637 -0.5679879735 91 0.243968799 -0.1708090972 -0.8818827791 92 1.145400127 2.124269004 1.467923113 93 -0.8047593895 0.5056868541 -1.096263713 94 0.5698579676 -0.02483819284 0.01838255657 95 -0.4287304074 -0.7029320985 -0.6543157316 96 -0.3904518409 -0.7029320985 0.4897226566 97 0.6020173821 -0.4304266825 -0.6270658197 98 -0.644751284 0.4302696097 0.2187772152 99 -0.9348190202 -0.6200539815 0.5322793462 100 -0.2257248141 -0.3450987553 -0.5586670634 101 -0.03621355397 0.1685931551 0.740018294 102 0.6640019727 0.01517632956 -0.5547381442 103 -0.7618137176 -0.6551899681 0.3467528675 104 0.2117679526 0.1392970669 -0.7202749212 105 0.297182161 -0.6439250467 -0.8507676633 106 -0.3958116463 -0.667796112 0.260428861 107 -0.6380510406 0.3338029019 -0.9439767648 108 0.5189017775 -0.7029320985 -0.7332610742 109 -0.6528658111 1.36176545 0.07509835906 110 0.353585043 0.4982249314 0.4463131511 111 -0.931361436 -0.4009754361 -0.9719394952 112 -0.7514409652 -0.7268031637 0.144891641 113 -0.03907390627 -0.3307034631 -0.7391731703 114 0.6047334173 0.4910367922 0.6110733943 115 0.6493578578 1.625159765 0.8982976651 116 1.438710523 1.504083593 1.652083459 117 -0.2050823015 -0.6790610333 -0.8973722141 118 -0.1245796695 -0.7268031637 -1.012375521 119 1.357213911 0.7768360963 1.219076618 120 0.5910037457 1.186800179 1.004893745 121 0.5751262181 -0.310651398 -0.2479246431 122 0.2177902418 -0.6551899681 -0.6177449096 123 -0.5428225547 1.174698913 -0.9939901296 124 0.2457574303 1.665046211 0.4593203594 125 0.5796869379 0.9398838529 0.6858195812 126 0.2953521335 -0.667796112 -0.832125843 127 -0.1002948415 -0.667796112 -0.8694094836 128 -0.5097509 -0.3054911753 -0.3658803745 129 -0.6357507848 -0.6551899681 -0.6177449096 130 -0.8012102505 -0.6790610333 -0.7482376517 131 0.2893616596 -0.7029320985 -1.012375521 132 -0.7880105826 -0.7029320985 -0.4624982901 133 -0.8440450064 2.072159286 -0.8510240921 134 0.3618229758 -0.05819804704 0.2910100639 135 -0.4712386829 -0.7029320985 -0.8600885735 136 -0.7383522133 1.163938475 -0.1959208538 137 -0.7583561334 -0.2369953346 0.5907048258 138 0.4611272248 1.782779678 1.119897052 139 0.2948130198 -0.3532333057 0.2419659464
Table 4: Averaged Data per Respondent for Game Behavior
Enjoyment of Game Elements
Resp. Achievement Social Immersion 1 0.8377264153 0.4502243721 0.2040125897 2 0.1040782348 -0.2825476796 -0.1125233221 3 -1.012981581 0.8404511947 0.8507225802 4 -0.2634894984 -0.181809283 1.174950771 5 -1.80484711 -1.378650237 -1.900044201 6 -0.5745957757 -0.5292676855 0.6710850253 7 1.243395551 0.9507017005 -0.4329325529 8 0.1388557013 -1.363762644 0.5252830366 9 -1.559306122 -1.62186253 -1.9118251 10 -0.314825014 0.04141400699 0.842760641 11 0.5733925733 0.9440072308 -0.272122261 12 -0.01268373828 0.09567447324 1.174950771 13 0.326723738 -0.2846604027 -0.9223554335 14 -0.0842516548 -0.8678259491 -1.900044201 15 0.3327248899 0.7031637176 0.5294521333 16 -0.5187996787 -1.257044091 -0.4263972681 17 -0.3276110524 -1.261753672 -1.232140759 18 0.2426730146 -0.64393441 -0.9258786155 19 -0.1276369565 -0.3880401933 1.007094115 20 0.3768279168 -0.6458878295 -1.090212089 21 0.413759584 0.1824748978 0.04441365083 22 0.4868071612 1.079406572 0.8616422631 23 1.172097871 1.561598414 1.010025741 24 0.3395445805 0.562626299 0.2007851856 25 -0.2642591874 0.5900845365 -0.9297519344 26 0.02500787769 -0.3903509866 -1.254841341 27 -0.9388763391 -1.135437944 -0.1467090253 28 0.2214254179 1.073578422 0.2078859086 29 0.377335662 1.694829393 1.174654993 30 0.3718831123 0.09911347965 0.6852321123 31 0.4491077347 -0.2786729485 1.010321519 32 -0.4778992834 -0.6542813696 0.6713808031 33 0.4026932067 0.7121916915 0.3651186595 34 0.02077088603 0.10020058 -0.280379978 35 0.405448817 0.7018374343 1.339284245 36 0.1543418055 0.0838588685 0.6746082073 37 -0.1203126032 -0.6207220518 0.850157142 38 0.9642238189 1.071441528 0.5326795375 39 -0.02914593875 -0.1387789689 0.5099789555 40 -1.027280419 -0.4910893832 -0.9223554335 41 -0.3354646283 -0.293920262 -0.268599079 42 -0.07597286168 -0.6727526776 -1.090212089 43 0.7459964712 -0.3841112825 -0.6058195216 44 0.6918675351 0.9495831315 1.007094115 45 1.167723069 0.9483052587 1.339284245 46 0.4401557502 0.3294507297 0.2188055916 47 0.3353925297 0.7143117123 0.5294521333 48 -0.7944353712 -0.4157786583 -0.7698028587 49 -0.8015028877 -0.4138494099 -0.1511738999 50 -1.606064173 -0.7666591255 -1.105300869 51 0.4941555544 -0.2700565281 0.03586015595 52 -0.4767663591 -0.8987957062 -0.1202156009 53 -0.01265233469 -0.1467198426 -0.1243042212 54 -0.09955044995 0.2034874244 1.174950771 55 0.1273465365 0.7358308378 1.010025741 56 0.3084445216 0.6842082744 0.3565651646 57 -0.5947092229 -1.118795641 -1.587892687 58 -1.106970163 -1.371710815 -1.900044201 59 0.4582782133 0.2067501853 0.6855278902 60 0.5613185922 -1.031867381 0.6813587934 61 -0.2125399325 -0.1813977294 1.003570933 62 -0.1217075666 -0.6312839549 1.171427589 63 0.1021250286 0.450214728 -0.280379978 64 -1.118592441 -0.6762158551 -1.265465246 65 -0.1648090828 0.4536464367 -0.4307816662 66 -0.1674388046 0.2434924531 0.6937856072 67 -0.01322549462 0.2125541647 0.04325665686 68 0.2628600109 0.2067260142 0.8424648632 69 -0.3242968519 -0.3968473862 -0.4450092298 70 -0.1813026869 -0.5223282636 0.3651186595
Table 5: Averaged Data per Respondent for Enjoyment of Game Elements
cont. Enjoyment of Game Elements
71 -0.2394791828 0.469780434 0.05267136785 72 -0.4675420659 -1.02520438 -0.7786260139 73 -0.4043026597 0.3124189445 0.2090429026 74 -0.03787373545 0.4661375882 0.3733763765 75 -0.8892603801 -0.8841918316 -0.6128658857 76 -0.4590240892 -0.132958116 0.1813120078 77 0.3948878195 1.185997291 0.5368486343 78 0.4693140084 0.5418936819 0.861938041 79 -0.3614626963 0.2192171657 -0.2806757559 80 0.02124532331 0.4489368552 -0.280379978 81 0.1197720827 -0.1701602796 0.1972620036 82 1.243426955 1.434171415 1.339284245 83 0.01364768288 0.3111652429 -0.2680075233 84 -0.3180127757 0.5752525831 0.3725151604 85 -0.1249345309 1.17562616 1.183208488 86 0.08001949298 -0.4444279781 1.339284245 87 0.9954804744 1.439992268 0.3851572755 88 1.240674016 -0.1779030832 1.339284245 89 0.2136027962 0.08386616616 0.04355243474 90 0.02977008807 -0.4054316987 -0.280379978 91 0.1799540624 0.3283321606 0.1972620036 92 -0.4369045219 -0.5176113845 0.01315957398 93 -0.3722173345 0.2170971449 -0.280379978 94 0.9312051777 0.6971278528 -1.063988325 95 -0.09620293484 -0.2526035453 0.2078859086 96 -0.1967025698 0.173391284 -0.1125233221 97 -0.001028243235 0.4560355808 -1.090212089 98 -0.3965669952 -0.8940861247 -1.090212089 99 0.005774323549 -1.3880694 -1.900044201 100 -0.2740531624 -0.04938898362 -0.4406248317 101 0.09188929305 0.3248931542 0.3648228816 102 0.0229586638 0.6986207374 1.010617297 103 -0.8719807907 -1.146810527 -1.246287846 104 -0.02366994373 -0.5160643885 0.8424648632 105 0.5239432014 0.1964443386 1.174654993 106 -0.083882029 -0.2904885533 0.06474804475 107 -0.4999243517 -0.1622194059 -0.7615451416 108 0.3637327051 0.4443865774 0.5303133494 109 -0.2535020245 0.07478255237 0.3648228816 110 -0.09685624513 0.3447149166 0.6940813851 111 0.2000523755 0.250431875 -0.5904610816 112 0.1644275806 0.9612321348 0.2087471247 113 -0.6600441397 -0.04017023721 -0.7689416426 114 0.4604132534 -0.1638964044 -0.9341363325 115 0.4282737989 -0.5171418446 -0.2736293919 116 0.04888602258 0.1113727458 -0.1125233221 117 0.1977458083 0.6822548549 -0.5975618046 118 -0.6532954099 -1.62186253 -1.430955714 119 -0.4271280933 -0.3948939667 -0.7686458647 120 0.2203226563 0.8049481016 -0.6016504249 121 0.2727554485 0.478621442 -0.6051736069 122 -0.02514210513 0.8507981543 -0.2795187619 123 0.09348775152 0.199619991 0.5294521333 124 0.01028555968 0.4652954401 -0.4405443551 125 0.7175528361 0.4594431185 0.5208986385 126 0.6271097001 0.9528144237 0.2004894077 127 0.6271097001 0.9528144237 0.2004894077 128 1.383332185 -0.05198584195 0.3683460636 129 -0.8127702122 -1.100531979 -0.4450092298 130 -1.557934776 -1.280866712 -1.900044201 131 0.2190867719 -0.1906019489 -1.408255132 132 0.3714894671 0.3262121398 0.5170253196 133 0.2626980247 0.4570443303 0.6778356114 134 0.8214892536 0.4491276277 0.5220556324 135 -1.283053324 -0.4279933888 -1.900044201 136 -0.3549595274 -0.523439535 1.014490616 137 0.1582847889 -0.01606155111 0.2211195795 138 -0.2384974606 -1.014666648 0.6831617022 139 0.4927070302 -0.7587109541 -0.6093427037
Table 5: Averaged Data per Respondent for Enjoyment of Game Elements
Usefulness of Game Elements
Resp. Achievement Social Immersion 1 0.09170650954 1.156796603 -0.5547677471 2 -0.3020317123 -0.7487496275 -0.07783286299 3 -0.7929147823 1.163455524 0.5831822028 4 0.02894779482 0.8894638094 0.5793427319 5 -0.5269595387 -0.2722861869 -1.52117615 6 -0.2736201056 -0.3609747621 -0.07628532786 7 1.167257825 0.6698464185 -0.2343924004 8 0.4523058299 -0.9811579659 0.4175086226 9 -1.851209048 -1.716768682 -1.863297323 10 -0.4104691514 0.5379993156 1.382738017 11 -0.02607294519 -0.3658573309 0.2426085653 12 -0.05069156334 -0.2818009627 1.058457118 13 0.2795881118 -0.264076119 -0.3848861695 14 -0.08779717306 -0.2743799033 -0.7340736034 15 0.2843094889 0.5630245355 0.571229182 16 -0.009803850129 -0.3843987108 0.2589013036 17 -0.02614997076 -1.00404221 -0.06315329832 18 0.3749036942 -0.2578517817 -0.3973394368 19 0.1136286502 -0.2876318533 0.9008970873 20 0.5278249059 0.2102936053 -0.2402796529 21 -0.01203663535 0.1940934895 0.08784071725 22 0.3786635914 1.398365292 1.217064437 23 -0.1913972214 0.8921067627 1.223630451 24 -0.02341661767 0.3001796529 0.2410610302 25 0.01324064383 0.5387733067 -0.3940002123 26 -0.8772910036 -0.4214051223 -1.528110691 27 -1.099411709 -0.7721866321 1.051390856 28 1.155908023 1.624534276 0.3922335615 29 0.5849797659 1.496328872 1.223630451 30 0.2773346899 0.06742904169 0.7364025158 31 0.4294418384 0.2036232827 1.217064437 32 0.1388011841 -0.5091031892 1.04600385 33 0.3595779488 0.7892085365 0.4055556017 34 0.3656393493 0.6640688167 -0.2316658566 35 0.2789224317 -0.04693263052 1.382738017 36 0.03360940763 0.5247434505 0.7277887194 37 -0.5727467039 -0.3068029283 0.8955100812 38 0.9213773351 0.7649422905 -0.2303744137 39 -0.05603706977 -0.6438752035 -0.4113402393 40 -0.8898897596 -0.3930705316 -0.8893416979 41 -0.08108277682 -0.2777748046 -0.3993872184 42 -0.1399640525 -0.8906154458 -1.191007998 43 0.1500374397 0.07873157132 -0.3940002123 44 0.4965485932 0.4275810314 1.382738017 45 0.8720444726 0.8971486449 1.217064437 46 0.3678495303 -0.1727024038 0.09143603398 47 0.2965981739 0.4327404322 0.571229182 48 -0.5833571127 -0.3046067105 -1.216283059 49 0.1531540375 -0.02546744784 0.2373339934 50 -0.597529028 0.641526845 -0.0746060727 51 0.2457493555 -0.1778085157 -0.2407798994 52 -0.6222928616 -1.011742063 -0.3940002123 53 -0.1463939333 -0.4095986532 -0.06921906663 54 -0.4307288849 0.2824157873 1.382738017 55 0.2468156197 0.6525443057 0.8988493057 56 0.528289582 0.540233805 0.4050553553 57 -0.7521266839 -1.47405656 -1.387475366 58 -1.858475756 -1.716768682 -1.863297323 59 0.4668129441 0.07399565184 0.729836501 60 0.5533872455 -0.314159758 0.746307755 61 -0.03614587691 0.2765848967 0.8868962849 62 0.3239328916 -0.4142963011 0.7277887194 63 1.015503385 0.6253953349 -0.2402796529 64 -0.3308597129 -0.7459201915 -1.074403087 65 0.5663752433 0.6423430227 0.4082821456 66 0.5002912373 0.6709796611 1.211677431 67 -0.6172883759 0.2117158319 -0.8861149076 68 0.05999441019 0.1037757235 0.2253809726 69 -0.3921101178 -0.1513338999 -0.3993872184 70 -0.1600197685 -0.6349936458 -0.2348926469
Table 6: Averaged Data per Respondent for Usefulness of Game Elements
cont. Usefulness of Game Elements
71 -0.09379870933 0.6019266866 -0.3955477475 72 0.04367679894 -0.6648119891 0.2328157603 73 -0.4686473163 0.276203602 0.571229182 74 -0.02507245654 -0.7683584508 -0.8807279016 75 -1.04873494 -0.3720162274 -0.7275075886 76 -0.2910669328 -0.0361406743 0.2464480363 77 -0.5171594928 0.4265010775 0.2377218057 78 0.6930554808 0.8836329188 1.382738017 79 -0.434006148 0.0762723278 -0.2407798994 80 0.152545859 -0.05526253681 -1.863297323 81 -0.005848333321 0.1945012032 0.4141693981 82 1.166745792 0.9239805508 0.7336759719 83 0.01472183687 0.6632798083 0.571229182 84 -0.6959854682 0.4327404322 -0.7236681177 85 0.3773868639 1.024682559 1.058457118 86 0.05074612235 -0.3170042113 1.382738017 87 1.172530219 1.496328872 0.7357237535 88 0.7920216343 -0.2434173765 0.5872001895 89 0.1618688348 -0.6638132438 0.08784071725 90 0.003877463788 -0.6320837517 0.07538744995 91 0.2866887146 -0.1825984744 0.2291080094 92 0.02652938261 -2.325925438 -2.674806158 93 0.1503763056 -0.04611720308 -0.2402796529 94 1.067254858 0.763718399 -1.179054978 95 0.09287114391 -1.20768424 -1.039835467 96 -0.9721008825 0.1941589304 -0.6961667593 97 0.001647579161 0.1980171463 -0.7257158993 98 -0.1623531302 -0.63873082 -0.8931811689 99 0.5589416699 -0.3922815231 -1.863297323 100 -0.3188968648 -0.299462327 -0.3955477475 101 -1.489454855 -0.154095583 -0.7340736034 102 0.793403182 1.624534276 0.746307755 103 -1.344418941 -1.474845568 -1.025834665 104 0.2643968077 -0.5923836048 0.7318842826 105 0.4232460954 0.2102936053 -0.2402796529 106 -0.2521855274 -0.153672852 -0.3810466986 107 -0.6571771421 -0.3211348322 -0.5547677471 108 0.5010168445 0.3059572546 0.7407422332 109 0.08414427422 0.1900283095 0.4121216165 110 -0.03139878853 -0.8305423756 0.7483555366 111 0.509547 -0.1585302052 -1.191007998 112 -0.2541041036 0.5435482481 -0.06921906663 113 -0.7014267658 -0.1650175683 -0.5579945374 114 0.3578312902 -0.03473346496 -0.7275075886 115 0.4140463877 -0.6324531938 0.1013412732 116 0.3297891587 0.2113073679 -0.2509412309 117 0.1541288395 0.4176195199 -0.8687748807 118 -0.5588802341 -1.716768682 -1.863297323 119 -1.135984336 -0.5156697993 -0.5790616011 120 0.6260751424 0.4562961665 0.2604488387 121 -0.09885745316 -0.159201695 -0.7275075886 122 -0.2437431783 0.3074831937 -0.06921906663 123 -0.2101514841 -0.3823200117 -0.06767153149 124 -0.5024020905 0.08865481123 -0.08117208747 125 0.9508625557 0.6691106989 0.08938825239 126 0.6700955228 0.8871871334 0.571229182 127 0.6700955228 0.8871871334 0.571229182 128 0.5032011677 -0.3922815231 -1.051788488 129 -0.9597500783 -0.2952784173 -0.2342138846 130 -1.682726974 -1.490599699 -1.863297323 131 -0.02911870693 -0.2938944622 -1.025834665 132 0.4229218188 -0.1269344601 0.4083945798 133 0.005352190416 0.334330117 0.5621151392 134 0.5838684159 -0.03394445654 0.412621863 135 -0.385901905 0.06838951545 -1.051788488 136 -0.9325569858 -0.8857328769 0.4001685957 137 0.2679198354 0.3012738737 0.8927835374 138 -0.4545847286 -1.012069319 0.7178834802 139 0.4797290276 -0.5119326252 0.08734047079
Table 6: Averaged Data per Respondent for Usefulness of Game Elements
