Motivation to Play, Player Typologies, and Addiction, Exams of Statistics

Download Motivation to Play, Player Typologies, and Addiction and more Exams Statistics in PDF only on Docsity! Massively Multiplayer Online Roleplaying Gaming: Motivation to Play, Player Typologies, and Addiction Dissertation Presented in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in the Graduate School of The Ohio State University By Michael Scott Lewis Graduate Program in Education – Physical Activity and Education Services The Ohio State University 2016 Dissertation Committee Paul F. Granello, Ph.D., Advisor Darcy Haag-Granello, Ph.D. Jerome D’Agostino, Ph.D. Copyrighted by Michael Scott Lewis 2016 iv Dedication This manuscript is dedicated to three distinct entities: 1) My Mom and Dad who bought me my first video game system at age eight; 2) Nintendo for creating The Legend of Zelda, assuring I’d be lovingly hooked on video games forever; and 3) Laura who allows me to indulge my inner gamer. v Acknowledgments I humbly acknowledge my dissertation committee for helping to support this process, first explored 14 years ago, come to fruition. Thanks to my advisor, Dr. Paul Granello, for encouraging me; to Dr. Darcy Granello for helping me find my writer’s voice; and to Dr. Jerome D’Agostino for giving me tools to execute this study. I want to give gratitude to my parents for always believing in me from a young age and giving me the freedom to fail so that I could learn how to succeed. I am appreciative of my friends and family for offering unwavering support and forcing me to have fun. I’d like to thank Team Ph.D. who continuously grounded and inspired me throughout this process. Lastly, and most importantly, I sincerely thank my wife Laura for being my center. She grounds me, moves me, and makes me the best version of myself I can hope to be. She has never doubted the completion of this manuscript, even if I silently did. With love, I appreciate her efforts. vi Vita May 1995……………………....Greenon High School May 1999………………………B.A. Psychology, Ohio Dominican University May 2003……………………....M.A. Counselor Education, The Ohio State University 2010 – Present………………….Ph.D. Counselor Education, The Ohio State University Publications (In Process) Hunnicut Hollinbaugh, K. M. & Lewis, M.S. (anticipated 2016). Using Dialectical Behavioral Therapy with Adolescents. Columbus, OH: Routledge. Focht, B. & Lewis, M. S. (2011). Physical activity and psychological well-being. In P. F. Granello, Wellness Counseling. Upper Saddle River, NJ: Prentice Hall. Mahaffey, B. A. & Lewis, M. S. (2008). Therapeutic alliance directions in marriage, couple, and family counseling. In G. R. Walz, J. C. Bleuer, & R. K. Yep (Eds.), Compelling counseling interventions: Celebrating VISTAS’ fifth anniversary (pp. 1-10). Ann Arbor, MI: (Counseling Outfitters. Fields of Study Major Field: Counselor Education ix 3.3 Limitations and Threats to Validity ...........................................................73 3.3.1 Generalizability ..................................................................................73 3.3.2 Construct Validity ..............................................................................73 3.3.3 Attrition and Sampling Bias...............................................................74 3.4 Procedure ...................................................................................................75 3.5 Recommended Data Analysis ....................................................................76 4. Results .....................................................................................................................78 4.1 Participants .................................................................................................79 4.2 Tests of Normality .....................................................................................79 4.2.1 Reliability ...........................................................................................83 4.3 Research Question One ..............................................................................84 4.3.1 Age and Sex .......................................................................................84 4.3.2 Race....................................................................................................84 4.3.3 Education ...........................................................................................85 4.3.4 Annual Income ...................................................................................85 4.3.5 Marital Status .....................................................................................86 4.3.6 Number of Months Playing MMORPG’s ..........................................87 4.3.7 Number of Days Playing MMORPG’s ..............................................87 4.3.8 Number of Hours Playing MMORPG’s Per Day ..............................88 4.3.9 Who Introduced You to MMORPG’s ................................................88 4.3.10 Time of Day .....................................................................................89 4.3.11 MMORPG’s Played ..........................................................................89 4.3.12 Independence of Samples Analysis .................................................91 x 4.4 Research Question Two .............................................................................95 4.5 Research Question Three ...........................................................................96 4.6 Research Question Four .............................................................................99 4.7 Post Hoc Analyses .....................................................................................103 4.7.1 Gender ................................................................................................103 4.7.2 Passion Criteria ..................................................................................106 4.7.3 Motivation for Play Subcomponents .................................................109 4.7.4 Time Played .......................................................................................111 4.7.5 Addiction Level .................................................................................112 5. Conclusions, Summary, and Recommendations ......................................................114 5.1 Research Findings ......................................................................................114 5.1.1 Obsessive Passion ..............................................................................114 5.1.2 Gender Differences ............................................................................115 5.1.3 Time Played .......................................................................................116 5.1.4 Clinical Implications ..........................................................................116 5.1.5 Passion Criteria ..................................................................................117 5.1.6 Motivation for Play ............................................................................118 5.2 Research Question Analyses ......................................................................118 5.2.1 Research Question One ......................................................................118 5.2.2 Research Question Two .....................................................................121 5.2.3 Research Question Three ...................................................................122 5.2.4 Research Question Four .....................................................................123 5.2.5 Post Hoc Analyses .............................................................................124 xi 5.3 Methodological Implications .....................................................................125 5.4 Limitations ................................................................................................126 5.5 Future Research .........................................................................................127 5.6 Summary ....................................................................................................128 References ........................................................................................................................131 Appendices: A. Demographics Questionnaire ...............................................................................141 B. Problematic Online Gaming Use Scale................................................................144 C. The Motivations for Play Questionnaire ..............................................................147 D. The Passion Scale ................................................................................................151 E. IRB Approval .......................................................................................................154 xiv List of Figures Figure Page 2.1 Problematic Online Game Use Scale Model: 5 Factors and One Second Order Factor ...............................................................................................................................56 1 Chapter 1 Introduction Research on addiction has traditionally been focused on understanding the characteristics of chemical addiction, such as dependence on a substance like heroin or alcohol (Walters & Gilbert, 2000). Recent studies, however, are attempting to understand a set of behaviors that have similar characteristics to chemical addiction but do not involve the ingestion of a substance (Smith, 2012). These behaviors, commonly known as process addictions, are activities that are at first enjoyable but eventually lead to symptomology such as tolerance, withdrawal, lack of control, and a pattern of remission and relapse, all of which are similar to chemical dependency (Black, 2013). Although process addictions have increasingly gained attention from addiction and mental health professionals as they become more prevalent in our society (American Society for Addiction Medicine [ASAM], 2011), the body of research in this area continues to be less established around understanding the general characteristics of process addictions, including clinical issues. Online gaming, a relatively new clinical concern within the spectrum of process addictions (Chappell, Eatough, Davies, & Griffiths, 2006), appears to be the fastest growing addiction throughout the past decade (Young, 2009). It has been suggested that online gaming addiction is best understood using research and theory that outlines the concept of process addiction. Therefore, process addiction characteristics, history and assessment, diagnosis, and treatment 2 considerations all provide essential components for understanding online gaming addiction. Statement of the Problem Online gaming addiction is one of the fastest growing addictions in the past decade largely due to the increasing availability of computer technology and Internet access (Young, 2009). Massively multiplayer online roleplaying games (MMORPG’s) most rapidly evolved in the late 1990’s. Frequent reports of problems began to emerge such as lost relationships, loss of employment, and mental health diagnoses (Hussain & Griffiths, 2009), and mental health professionals began to relate these problems to addiction (ASAM, 2011). Websites devoted to discussing and treating the issue, such as Online Gamers Anonymous (olganon.org, 2012), and Everquest Widows (groups.yahoo.com, 2013), began to give people a place in which to speak about how MMORPG’s had impacted their lives or the lives of their loved ones. Everquest, one of the first popular MMORPG’s, earned the nickname “EverCrack” in reference to its immersive and addictive nature (Hussain & Griffiths, 2009; Ng & Weimer-Hastings, 2004). Given the rapid growth of the games, as well as the concerning consequences that have become associated with them, it has become clear that online gaming has the potential to cause significant harm to game players. For example, the difficulties associated with online gaming has reached near pandemic levels in South Korea and other Asian countries, resulting in legislation that governs use of online games, hours played, and time in Internet cafes where most players play (Seok & DaCosta, 2012). There have been many documented consequences associated with online gaming ranging 5 who oft define themselves by their gameplay and what the gameplay means for them personally are more apt to continue playing despite negative consequences. Although the concepts of player motivations and dualistic passion lend insight to underlying factors related to MMORPG players they do not fully explain how they relate to or predict addiction to MMORPG games. In addition, researchers have been attempting to better classify and understand the prevalence of online gaming. There are over 19 million active players documented (mmodata.net, 2012c) and several additional million of players that play “free-to-play” MMORPG’s (which do not require a subscription and therefore are more difficult to track). Between 7% and 12% of people who play online game meet criteria indicative of addiction, such as playing longer than intended, difficulty stopping play over extended periods of time, increased time devoted to play, irritability and anger associated with being unable to play, and real world consequences linked to continued play (Gentile, 2009; Grusser, Thalemann, & Griffiths, 2007; Hussain & Griffiths, 2009). If these statistics are correct, it is possible that over 2 million people meet the criteria for addiction, and numerous family, friends, and peers may also be impacted. Therefore, given that the number of individuals who engage in online gaming is on the rise, both domestically and abroad, there is clear justification to further research to understand online gaming in order to properly diagnose and treat this type of process addiction. Purpose of the Study The present study seeks to understand the underlying factors that contribute to addiction among online role-playing gamers. Many researchers believe there are psychological and motivational components that differentiate players who become 6 addicted to online gaming from those who do not (Lafreniere, et al., 2009). The current study intends to combine the above concepts of dualistic passion (Vallerand, et al., 2003; Vallerand, 2010) and player typology (Yee, 2006), and to investigate how they interact and relate to each other. The researcher is also interested in how these constructs may contribute to and predict online gaming addiction through utilizing Kim and Kim’s (2010) Problematic Online Gaming Use Scale. The results will be analyzed first describing demographic statistics, correlations on the independent variables will be reported, and lastly a multiple regression model will explore these constructs and their relevance to online gaming addiction. Research Questions The researcher posits the following four questions: 1: What are means and distribution of MMORPG player’s age, gender, race, education, personal income, and frequency/duration of game play? 2: To what degree are people motivated by harmonious passion, obsessive passion and the three types of Player Typologies (immersion, socialization, and achievement)? 3: What is the correlation between the independent variables harmonious passion, obsessive passion, immersion, socialization, and achievement? 4: Are there patterns of play and passion according to Vallerand’s Dualistic Model of Passion (IV) and Yee’s player typologies immersion, socialization, and achievement (IVs) that can predict levels of addiction according to Kim & Kim’s Problematic Online Game Use Scale (DV)? 7 Significance of the Study The domestic and worldwide impact of MMORPG’s clearly demonstrates the need for an expert understanding of online gaming addiction. There is also a critical need to unify the discussion around online gaming addiction. As such, the results of this study will help provide practicing mental health clinicians and the research community a better understanding as to why players continually play games to the detriment of their quality of life. Second, the present study will contribute to and expand upon the language and definitions related to online gaming addiction. As Seok and DaCosta (2012) have remarked, there is a need to arrive at a common method to assess and measure the ways in which gamers become addicted, yet much of the research on diagnosis and assessment has been conflicting and incomplete (Kim & Kim, 2010). This study also intends to provide insight on how to assess and predict addiction. In general, the present research intends to provide further depth and understanding to the existing body of literature in this topic area and to provide evidence of the need for additional future research opportunities. Limitations of the Study As with most research, there are several limitations the researcher can identify within the current study. First, due to the nature of sampling an online community, generalizability is at risk. It can be inherently difficult to work with a sample obtained online as it cannot be predicted who will be attracted to the online portals where the study is housed. Although the sample of this study may be representative, it is also possible that more dedicated players would be likely to visit such online gaming websites and complete the survey, thus skewing the sample towards a more devoted gaming 10 (alteredgamer.com, 2012). The six criteria that make a MMORPG unique to other games are persistence, physicality, social interaction, avatar-mediated play, vertical progression, and perpetuity (Hussain & Giffiths, 2006). Ultima Online (UO) Ultima Online (UO) was one of the first graphically based and widely dispersed MMORPG’s. UO was released in 1997 and is still active today (uo.com, 2014). It was also the first MMORPG to reach 100,000 active subscribers and peaked at over 250,000 in 2003. World of Warcraft (WoW) World of Warcraft is one of the most popular MMORPG’s to date boasting over 12 million active users at its peak (mmodata.net, 2012b). WoW was released in 2004 and has since become the standard to which big market MMORPG’s are held. The game’s manufacturer, Blizzard Entertainment, has released five expansions, the most recent in 2014. WoW has also become the most financially successful MMORPG generating over 10 billion in sales revenue (businessinsider.com, 2012). Over 100 million lifetime subscriber accounts have been created (polygon.com, 2014). Guild A guild is a collection of players who form a like-minded or purposed group. Guilds often team together in online play and assist in player advancement and socialization. Guilds can function as a team in order to take on particularly challenging parts of the game, commonly known as guild raids. These activities require planning, coordination, and strategy and often require dozens of players to coordinate in order to successfully complete the tasks. 11 Experience Experience is a functional part of MMORPG’s as it allows players to measure their progress towards the next level of their avatar. Experience grows through activities such as killing monsters, exploration of the world, crafting items, or other game achievements and moves players to more advanced levels as the game progresses. For example, it may take 500 experience points to advance from Character Level One to Character Level Two or 10,000 points to advance from Character Level 10 to Character Level 11. In essence, advanced achievements often offer more experience, and therefore the game playtime required can significantly increase in order to achieve higher levels. Avatar An avatar is a controllable physical representation of the player used to manipulate and interact with the game environment (Yee, 2006). Players create avatars by selecting adjustable features such as gender, body characteristics, and dress. Avatars are also informed by statistics that account for a player’s attributes such as strength, wisdom, luck, and fortitude among other traits. Players are often required to assign their avatars a player class and race, which will further represent how the avatar interacts with the world. Player Class Player class refers to professional archetypes and professions that players assume in games such as a fighter, healer, or wizard (Yee, 2006). These roles inform how the player will play the game. For example, a fighter might play in close combat with enemies whereas a wizard is more proficient at fighting from afar with magic. Typically, 12 players choose to join with other players to form a group of diverse roles that complement one another during game play. Player Race Player race denotes fantasy roles that players assume in a game such as elves, gnomes, ogres, trolls and hybrids of other races (Yee, 2006). Racial differences can be demonstrated as well as physical appearance and often inform what class the character can choose. For example, an ogre might have more physical strength to begin a game, a gnome would have more wisdom. Therefore, an ogre would be better suited to become a fighter than a wizard. Some games, however, prevent players from making certain class and race combinations. Player vs. Environment (PvE) Player vs. Environment (PvE) is a style of play common to MMORPG’s in which players play alone or with groups to combat game made creatures in order to gain experience and treasure (alteredgamer.com, 2012). This type of environment often encompasses interaction with towns or large cities, dungeons, and differing overworlds, which can change by region or by what type of inhabitant lives there. For example, northern regions of the game may be snow covered to simulate the change in temperature or an Elvan city may be in the treetops. Players in PvE’s interact with non-player characters such as townspeople in order to obtain quests and seek out new environments in order to complete the quests. Players can team with other players online in order to complete quests together. As the game continues, players are often encouraged to explore more of the game’s environment and new parts of the virtual world. 15 Achievement Player An achievement game player is defined as a player who enjoys advancing their avatar to the maximum levels, defeating the most difficult challenges, and receiving the best treasures. These players are likely to be recognized for their accomplishments such as obtaining a rare treasure, reaching high character levels, or asserting leadership roles within guilds (Yee, 2006). Exploration Player An exploration game player is primarily interested in learning more about the virtual game world and experiencing what it has to offer him/her (Yee, 2006). This player may spend time reading the game’s virtual history and stories, learning about the game’s landscape, and engaging with the game’s lore (i.e. the virtual world’s history, NPC backgrounds, legends of heroes) through role-play. Social Player A social player is a type of game player who primarily plays MMORRPGs in order to interact with and become acquainted with other players online. This often evolves through formed groups, guilds, or one-on-one interaction. Social players often form substantial relationships in the game (Yee, 2006). Summary Process addictions are on the behaviors that mimic chemical addiction including tolerance, withdrawal, and continued engagement despite consequences. The fastest growing process addiction over the past decade has been online gaming. Online games, specifically Massively Multiplayer Online Roleplaying Games (MMORPG’s), have been linked to significant problems for many users including job loss, relationship strife, 16 violence, health issues, legal problems, and in more extreme circumstances death. The total number of MMORPG players continues to grow making online gaming addiction a topic worth investigating. This study posits that there are underlying motivations related to the concepts of dualistic passion and player typology that may help explain why players are motivated to play MMORPG’s despite the negative effects and consequences. 17 Chapter 2 Literature Review The concept of process addiction is not a completely new domain of clinical research interest. Problematic gambling, for example, was included as a diagnosis in the Diagnostic Statistical Manual of Mental Disorders – Fourth Edition (DSM-IV) (APA, 1994) and since then researchers have written about and explored other behavioral addiction areas such as shopping, sex, exercise, and eating (Young, 1998). Research concentrated specifically on the online world or online games with even less attention to MMORPG’s was slow to follow. In the mid-1990s researchers began to investigate Internet addiction. The attention given to this topic then is likely related to increased access and expansion of Internet services, which has allowed people broader access to different methods in which to engage in addicted behaviors online. As a result, the Internet quickly became a conduit for shopping, sexual, and gambling behaviors. Video games have also evolved with these technological advancements. In the late 1990’s and early 2000’s, MMORPG’s started to gain popularity among computer game players, and games such as Ultima Online, Everquest, and World of Warcraft emerged. During this time, process addiction concerns began to extend to online gaming, and research on the topic increased throughout the mid 2000’s. This chapter outlines the existing research on process addictions and its relationship to online gaming addiction, explains how MMORPG’s are played, identifies why players engage in these games, 20 individuals are motivated through the presence of withdrawal and cravings, which negatively reinforces the continued behavior. Hedonism, however, is motivated by emotional responses and pursuit of pleasurable emotions or retreat from negative ones; motivation from reinforcement is learned through experience and strengthened through repeated measures. Lastly, chemical dependence can also be understood through incentive motivation theories (Yee, 2006). Incentive motivation suggests that the hedonistic motivators are not sufficient to motivate and continue the behavior. Instead, after repeated usage, the brain becomes rewired to desire the drug, and the brain’s reward center is then altered to view the drug as necessary to survival (Freimuth, 2009; Smith, 2012). In addition, the incentive threshold, the level at which the drug begins to rewire the brain, is variable among drugs (Petri & Govern, 2004). For example, the threshold for heroin is much lower than marijuana as it requires less usage and dosage to create a physiological change. The threshold is also variable among individuals, which explains why some individuals are more prone to developing dependence than others. Process Addictions In the past, the American Society of Addiction Medicine’s (ASAM) definition of addiction was exclusive to chemical dependency; however, in their most recent Public Policy Statement, process addictions are recognized and included for the first time (ASAM, 2011). As previously mentioned, process addiction is not an entirely new concept; the diagnosis of pathological gambling was first formally recognized in the DSM-IV as an impulse control disorder (APA, 1994) and updated in the latest edition (DSM-5) as the newly added category of behavioral addiction (APA, 2013). Although 21 the American Psychiatric Association (APA) recognizes gambling addiction as the singular behavioral addiction, further research has been requested by the APA to understand other behavioral addictions, including Internet gaming disorder, which has been listed as a condition warranting further clinical research (dsm5.org, 2013). Although acceptance of process addictions as a group of disorders has been slow to materialize, there has been substantial growth in research in this area over the past two decades (Smith, 2012). A search for peer-reviewed articles on a university library search engine using the term “process addiction” returned 375 articles published in 1994. In 2004, there were 1099 results discovered and in 2014, 2986 articles were included in the results. These searches indicate a significant growth of 171% from 2004 and 2014 and 696% between 1994 and 2014. In addition, this increase in attention appears most noticeable in the areas of sexual addiction, pathological gambling, compulsive shopping, and Internet-related addictions (Rooij, Schoenmakers, van de Eijnden, & Mheen, 2010). These results demonstrate a recognized need for continued and focused research in the area of process addictions, especially since chemical and process addictions have many parallel characteristics. To begin, both involve displays of tolerance, withdrawal, negative consequences related to use, a pattern of remission and relapse, and a physiological and/or psychological preoccupation with use (ASAM, 2011; Smith, 2012). Individuals with process addiction report some of the same thought patterns, cravings, and symptomology (i.e. tension, uneasiness, preoccupation, or a “hunger” for use) as chemically dependent individuals (Smith, 2012). The action is simply domain specific (i.e. ingestion or behavioral) (Grant, Potenza, Weinstein, & Gorelick, 2010). 22 There is also substantial evidence that process addictions stimulate dopamine secretion in similar ways to drugs or alcohol (Grant, et al., 2010; Oberg, Christie, & Tata, 2011; Smith, 2012). The motivational models for continued chemical use outlined above can also be applied to process addictions even if they are manifested in different ways (Grant, et al., 2010). For example, although the withdrawal symptoms for chemical use are typically physical, such as perspiration, cramping, or headaches (Charlton & Danforth, 2007; Griffiths & Meredith, 2009), the withdrawal symptoms for process addiction are often psychological as demonstrated by irritable moods, preoccupation with the behavior, anger, or depression (Freimuth, 2009). In addition, the consequences resulting from use is another area in which process and chemical addictions are comparable. The results of addictive behavior can take several forms, such as chronic health concerns or maladies, interpersonal relationship strife, illegal acts, vocational or school difficulties, and discipline at school or work (Anand, 2007; Freimuth, 2009). The presence of such negative consequences is one of the key indicators for both types of addiction. Although individuals may experience the consequences differently, the same level of engagement can contribute to such adverse consequences (Charlton & Danforth, 2007). In sum, it is somewhat perplexing why it has been so challenging for process addictions to gain active research momentum given their similarities between chemical and behavioral addictions. However, Young (2009) notes that the recent growth of research in process addictions is connected to the rise in the accessibility of the Internet. Increased access to pornography, cyber-shopping, and online gambling sites has led to 25 indicated a mean use of 4.9 hours per week (SD = 4.7), while the dependent group averaged 38.5 hours of use per week (SD = 8.04). Young also noted that the dependent group reported increased use over time, which appeared to be similar to the experience of tolerance. Most notably, the dependent and nondependent groups differed in what applications they used through the Internet. The dependent group engaged in applications that allowed them to interact more with other individuals or groups than did the nondependent group. The top three applications used by the dependent group were chat rooms (38%), multiuser dungeons (28%), and interactive newsgroup forums (15%). Conversely, these were the three least used applications among the nondependent group at 7%, 5%, and 10% respectively. Instead, the nondependent group more frequently sought information-gathering applications such as online libraries or databases and home pages (www addresses) at 24% and 25% respectively. Comparatively, the dependent group only used such applications 25% and 7% of the time. Finally, email use indicated the highest overlap between the two groups, with 30% of the nondependent group and 17% of the dependent group operating email regularly. All subjects in the dependent group in Young’s (1998) study reported some negative consequence related to their Internet use. Young categorized the impairments into five categories: academic, relationship, financial, occupational, and physical. Fifty- eight percent of the subjects in the dependent group reported severe problems with academic problems such as late assignments, lowered grades, or academic probation/expulsion. Since the mean age of the dependent group was 29 for males and 43 for females, this finding suggests that almost all of the college students in the group 26 experienced this consequence. Nighty-eight percent of subjects in this group reported moderate to severe disruption in relationships as described by ignoring personal relationships in lieu of online ones and online infidelity with cybersexual partners. In addition, severe financial concerns were reported by 58% of the subjects in this group related to paying for Internet service. At the time of the study individual internet use was paid for through data use (as opposed to the current format of flat rates), and a bill for online use could reach hundreds of dollars a month depending on how much data was used. Also, severe vocational difficulties were reported by 51% subjects which included tardiness or absenteeism, lack of focus at work (preoccupation with Internet user), and misuse of work time. Physical complaints were the least reported category with 75% reporting no problems and 25% reporting mild or moderate levels of concern. The most common reported complaint was sleep disturbance (attributed to staying up late to engage in online activities) and lack of exercise. Despite these negative consequences, 54% of subjects in this survey reporting problems indicated no intention to decrease their use of the Internet. The remaining 46% attempted to restrict their use by enforcing time limits or cancelling their Internet service altogether. Many of these participants reported an inability to refrain from continued use (this percentage was not reported by the researcher). Although this initial study was quite informative, it should be noted that there were several shortcomings and limitations. To begin, Young’s study was conducted nearly in the late 1990’s during the time period in which the Internet was a still very new and growing entity. Young (1998) specifically measured when subjects began using the Internet, and among the dependent group, 81% had been on the Internet for less than a 27 year. This suggests how original the Internet was at the time of data collection, which may suggest novelty may have played a role. Should Young’s study be replicated in present day it could be hypothesized that results might differ given that the Internet is more familiar to users and integrated into daily living. Secondly, the method Young used for subject recruitment resulted in a skewed sample. Young (1998) populated and advertised the survey on websites and used search terms that would easily engage individuals who already self-identified as regular users. This may have created data that was more reflective of a higher negative consequence set than the population as a whole. Also, the sample size was relatively small compared to the current three billion Internet users worldwide (liveinternetstas.com, 2015) and therefore generalizability must be applied with caution. Lastly, the instrument itself may not be satisfactory. Young’s use of adapted pathological gambling criteria, although perhaps helpful to use as an initial baseline, may not have been specific and inclusive enough in order to adequately measure broad Internet use and dependency. Young herself noted that addiction is likely domain specific to applications accessed via the Internet rather than the entire Internet itself. More specifically, the dependent group used applications that allow for more frequent interactivity. In spite of these limitations, Young’s (1998) work provided a helpful foundation for future research on process addictions by finding a link between Internet use and addiction. In the decades that followed, research relevant to the Internet became more specific to the individual aspects of Internet use. Subsequent studies have not only better defined Internet addiction, but also have identified subsets of Internet use such as gaming (Caplan, Williams, & Yee, 2009; Chappell, et al., 2006; Doan & Strickland, 2012), 30 addition is the development of online computer video games, which has resulted in the growth of yet another type of process addiction known as online gaming addiction. Online Video Gaming Online gaming addiction has become the fastest growing process addiction in the past decade (Young, 2009). In 2013, the video game industry was a $21.53 billion dollar industry making it one of the most lucrative entertainment industries (Entertainment Software Association, 2014). Comparatively, major motion pictures produced in 2013 in the U.S. grossed $10.82 billion (the-numbers.com, 2015). Thousands of digital games are produced each year in various formats such as game consoles like the Nintendo Wii or Microsoft Xbox One, computers and laptops, and handheld systems or smartphones. It is estimated 59% of Americans are engaging in some type of a regular digital game play. The average age of a video game player is 31 years, which has been trending upwards for over a decade. Males comprise the majority of the video gaming population but females represent 48% of the total video game market and in certain gaming subsections, such as mobile gaming, are the majority. Also, with the advent of accessible, reliable, and mainstream Internet connections, many of these games can also be played online with people all around the world (Yee, 2006). Although the online gaming market in general continues to grow steadily, it only represents 11% of the total share of the game software market (Entertainment Software Association, 2014). This may seem small but considering the genre did not exist two decades ago and how varied gaming is in general, 11% of the market is quite significant. This percentage is primarily maintained by one particular type of game – the Massively Multiplayer Online Roleplaying Game (MMORPG). Of the top 20 console or computer 31 games sold in 2010, six were MMORPG’s, the top of which ranked at numbers two, four, and six (Entertainment Software Association, 2011). In addition, MMORPG’s are often chosen for research on online gaming addiction because of their unique characteristics, variety in play, and other potentially addictive qualities. Understanding MMORPG’s Massively Multiplayer Online Roleplaying Games (MMORPG’s) are games that are played on the Internet to connect the player to a persistent virtual world. In this type of space, thousands of other players can interact with each other at the same time (Yee, 2006). Players create an account through the game manufacturer’s software and construct an “avatar”, a game character that will represent them in the game. Once a player is connected in the game, he/she can interact with the environment, game characters, and other players using already scripted action buttons, interactive chat lines, or emotive commands (such as waving one’s hand at another person). MMORPG’s often construct their games around a particular genre (i.e. Tolkien fantasy, space fantasy, old west) to create the setting, creatures, people, and other objects of the universe such as dragons, spaceships, saloons, etc. (Yee, 2006). Initially there are specific tasks a player must complete, but typically the player decides how to begin, which quests to complete and when, and with whom to interact. The gaming world is essentially wide-open for a player to explore at his/her pace. Players are also forced to create character roles when they choose their avatar, which will partially inform and prompt how their play. These roles or classes typically fall into the categories of fighter, healer, wizard or some variation of all three (Yee, 2006). For example, in the MMORPG Everquest a player can choose to belong to the 32 available classes of warrior, mage, cleric, rogue, wizard, druid, or ranger, among others (everquest.com, 2013). Each profession has strengths and weakness. Warrior roles, for example, are designed to be good at fighting up-close to enemies and can handle a large degree of damage before dying; they cannot, however, usually cast magic spells, heal themselves, or effectively use projectiles. Clerics, on the other hand, can heal themselves or others but are not as proficient at fighting. Such varied and defined roles make it beneficial for players to engage with other players in order to maximize each player’s individual success during the game (Yee, 2006). For example, a team of warriors that can fight enemies alongside a cleric who can heal the warriors when they are hurt is a more formidable and successful team than fighting alone. Additional members can be added over time to better prepare the group for larger tasks, quests, and enemies. Similar to character roles, players also choose races. Races in MMORPG/s are explained as player archetypes that come with particular benefits and disadvantages, such as humans, elves, gnomes, and dwarfs (Yee, 2006). Racial differences are often important when choosing what type of class the player would like to play. For example, a dwarf character may be considered physically stronger and therefore would make a better warrior than a gnome; however, the gnome may be considered wiser and therefore, make a good wizard. Combinations of race and class vary widely across MMORPG’s. MMORPG characters also have other various statistics and factors relevant to successful gameplay. These statistics are relatively consistent across MMORPG’s and include components such as strength, fortitude, wisdom, stamina, magic power, and luck. These statistics impact game outcomes such as how much damage the character can cause to an enemy, how well the character can read languages, how many magic spells 35 interact with one another in real world ways such as text based messaging or voice interaction. Fourth, MMORPG’s can be characterized as avatar-mediated play, which suggests that players create representations of themselves to interact with the game environment. The fifth factor, vertical game play, means that progression through the game is marked by growth of the avatar. This is typically achieved through character levels or virtual wealth, although some games have other novel ways of character growth and demarcation. Lastly, the sixth characteristic of MMORPG’s is perpetuity, which refers to the idea that the game has no end point, that there are always new goals to achieve and places to explore, and that game makers can always create new content to keep players engaged. In sum, there are countless ways that players can engage in MMORPG’s games (Yee, 2006). Such experiences can essentially be described as “virtual sandboxes” meaning players can play however, whenever, and with whomever, they please. Players choose their heroes’ defining characteristics such as gender, race, profession, skills, and talents and help their character to develop and grow by gaining experience points and achieving progressive levels. They choose how they want to interact with the environment and with other people by exploring the world, becoming parts of larger groups or guilds, or competing with other players. This is perhaps why MMORPG’s attract a variety of different players from across nearly all demographic categories (Williams, Yee, & Caplan, 2008). Players of MMORPG’s It can be challenging to obtain the most accurate and recent statistics regarding the prototype of someone who regularly engages in MMORPG play since software 36 companies do not typically release player demographics to the public (Williams, et al., 2008). The historical stereotype of a person who plays videogames has greatly evolved over the last 20 years and may no longer be accurate (Barnett & Coulson, 2010). As a result of the mainstream popularity of video games, players of MMORPG’s come from nearly every demographic of race, sex, and age (Caplan, et al., 2009; Williams, et al., 2008). However, several large-scale studies conducted on MMORPG gamer demographics have provided some helpful insights (Griffiths, Davies, & Chappell, 2004; Williams, et al., 2008; Williams, et al., 2009). To begin, Williams et al. (2008) and Caplan et al. (2009) may best demonstrate this population diversity in some of the largest MMORPG studies to date. These two studies evaluated different variables but involved the same researchers and sample. The research team surveyed a total of 7,129 Everquest 2 players recruited with assistance from the game’s manufacturer Sony Entertainment who allowed the researchers to contact players from across four of the games virtual world servers. A final useable sample totaled 4,278 players. It is important to note that the level of cooperation between researcher and manufacturer seen in this study had not been witnessed to this point or since in research on this topic. Participants were incentivized with an in-game item that was created by Sony Entertainment for the purpose of this study. The survey consisted of demographic questions such as age, race, gender, education, length of video game play, and mental and physical health measures. Researchers also asked questions relative to motivation for play as a follow up to an earlier study on the same topic designed by Yee (2006). 37 The average age of a MMORPG player in this study was 31.16 years of age with a range of 12 to 65 years (Williams, et al., 2008). What many people considered at that time to be the primary demographic of online game players - the teen and college aged population (18-22) - in actuality only accounted for 19% of the total players; most players were in their late twenties (26%) and thirties (36%). There was an approximate 80/20% split between males and females respectively. The mean education and annual income of this sample far exceeded the general population with nearly 74% having obtained a college education with an average household income of over $84,000 (Williams, et al., 2008). In addition, the average player engaged in 25.86 hours of play per week with females (M = 29.31) playing more than males (M = 25.03). In addition, MMORPG players self-reported a BMI of 25.19 (SD =8.19), which is considered slightly overweight by the American Heart Association (heart.org, 2014) yet healthier than the average American who has a BMI of 28.6 (cdc.gov, 2014). Players in this sample also reported more mental health issues, particularly depression and anxiety, than what is typically observed in the general population (Williams, et al., 2008). In this study, 22.76% of the sample reported a prior diagnosis of depression and 18.1% reported prior diagnoses of anxiety disorders. These rates were higher than the average U.S. rate at 16.1% for depression and 11.3% (SD = 37.21%, t = 3.32, df = 6776, p < .005) for anxiety disorders (cdc.gov, 2011). Rates of depression were higher for females (36.52%, SD = 48.17%) than males (19.38%, SD = 39.63%) (t = 13.567, df =6776, p < .001). Conversely, reported substance abuse and dependence rates were lower than the general population (M = 5.56; SD = 22.91%, t = 2.73, df = 6798, p < .01; Population = 8.2% (samhsa.gov, 2013). 40 MMORPG’s. He posited that the built in nature of variety inherent to a MMORPG might be precisely what appeals to players. Yee hypothesized that players may have differing motivations related directly to styles of play in the game (such as creating relationships or exploring the games dungeons) and that these motivations may impact usage patterns or behaviors in the game. Yee (2006) recruited 3000 participants through online portals relevant to MMORPG’s and other specific popular MMORPG games at the time such as Everquest, Ultima Online, and Dark Age of Camelot to take the questionnaire. The items were generated based on Bartle’s (1996) prior work and updated to more current language. He designed the survey using a 5-point Likert scale ranging from “Not Important at All” to “Very Important”. Additionally, Yee added a qualitative component allowing participants to comment on their motivations for playing MMORPG’s. Using a factor analysis Yee (2006) noted that three primary components emerged as motivations for play: 1) achievement, 2) immersion, and 3) social interaction. Imposition on others, seen in Bartle’s (1996) typology, was not supported in this factor analysis as a motivation for play. Ten subcomponents that more explicitly define the primary components emerged as well. The achievement motivation suggested that players desire to be the best or to accomplish specific tasks. Subcomponents included advancement (gaining levels, power, wealth and status), mechanics (understanding how to optimize avatars or how to best approach tasks), and competition (competing with others either directly or indirectly). The immersion component referred to players who desired to become deeply involved in the playing experience. Subcomponents to immersion included discovery (exploring new and unique places and experiences), role- 41 playing (creating a unique avatar with a back-story or playing as though they are the character), customization (making the avatar unique in appearance) and escapism (avoiding real-life stressors by playing). Lastly, the social component addressed how a player wanted to engage with other people within the game. Subcomponents of this theme included socializing (chatting and helping others), relationships (forming meaningful long-term interactions with others), and teamwork (group interactions or being a part of a larger whole). Together the ten subcomponents accounted for 60% of the variance among participants for motivation to play MMORPG’s. Grouping the subcomponents into the three primary components helped prevent a large amount of cross correlations (Yee, 2006). Correlations among the three components were (r < .10) suggesting that they are unique. Contrary to Bartle, Yee discovered that players could play MMORPG’s for all three of these primary reasons, or in varying combinations, and a preference for one does not suppress another. Yee (2006) also explored gender differences and found several important distinctions between men and women regarding their play style preferences. Males scored higher on all achievement subcomponents whereas females scored higher on the relationship subcomponent, but the other areas were relatively equal. This seems to support a common gaming stereotype that suggests men want to dominate and women want to play together, but there were equal results for the socialization subcomponent. This indicates that building connections online is important for both genders even though the reasons for socializing may be different. Additionally, Yee (2006) explored the relationships between motivations to play and problematic usage as defined by Young’s (1998) Internet Addiction Scale (IAS). He 42 used an adaptation of Young’s survey to determine if any of the motivation typologies were potentially more problematic than others. Yee found that the escapism subcomponent was the best predictor of problematic usage (b = .31, p < .0001) indicating a strong relationship between this component and problematic usage. This was followed by the number of hours played per week (b = .30, p < .001) and the advancement subcomponent (b = .17, p < .01). No other components or subcomponents exceeded a coefficient of b = .10. This potentially outlines factors that may lead to addiction and could also indicate that games serve as a coping technique for other problems or anxieties. Furthermore, this data might suggest that some players play online games as a means to relieve anxiety or delay facing real-world problems. Yee suggested that this connection might support the correlation between problematic game usage and prior mental health issues, such as depression and anxiety disorders. Yee’s (2006) study provides a firm foundation for future research in the area of gaming motivation and addiction for a few important reasons. First, he created a protocol that can be used for understanding motivations and game play. Second, he grounded the idea that there are multiple motivations that prompt the play of MMORPG’s. Third, he expanded prior ideas and research that gave the concept stronger legitimacy. Lastly, Yee established a connection between motivation for play and problematic consequences. However, it should be noted that Yee’s (2006) study had a variety of limitations. To begin, the assessment adaptations (Young, 1998) created from DSM-IV criteria (APA, 1994) used to measure Internet addiction may not accurately assess the intended construct, as it does not explicitly address online gameplay. In fact, Young’s survey predates the MMORPG’s used in the study by at least seven years and was published 45 Although it is unclear exactly how many players may be impacted by online gaming addiction, research indicates that prevalence rates range from 7% - 12% of the MMORPG gaming population (Grusser, Thalemann, & Griffiths, 2007; Hussain & Griffiths, 2009; Gentile, 2009). However, what is perhaps even less apparent is why online gaming addiction affects some players and not others. One hypothesis is that personal factors among some players that match gameplay aspects internally motivate them to continue play despite the problems they encounter (Ducheneaut, Nickell, Yee, & Moore, 2006; Charlton & Danforth, 2007). As previously mentioned, Yee (2006) suggested that there are different factors within MMORPG’s that motivate players to continue play, such as the immersion of the gaming universe, a desire to achieve top levels, obtain valuable treasures, or the opportunity to socialize and build relationships with others. Research on how these motivations correlate with online gaming addiction are mixed. Yee (2006) noted that players who primarily play for immersion are more likely to demonstrate negative consequences over players motivated by different factors. He suggested it could be that these players engage in the game to escape real-life problems; therefore, playing MMORPG’s offers them relief and a way to cope with everyday life. Similarly, Hussain and Griffiths (2009) found that 41% of their subjects from a self- selected adult sample played MMORPG’s as a form of escapism; however, escapism was not necessarily correlated with online gaming addiction. Yee (2006) found a correlation with the amount of time played and online gaming addiction. Conversely, Kesici and Sahin (2009) suggested that socialization was correlated with extended playing time but did not correlate it with online gaming 46 addiction. Additionally, players who play for competition, a subcomponent of the Yee’s (2006) achievement variable, did not show significant problems related to online gaming addiction (Hart, Johnson, Stamm, Angers, Robinson, Lally, Fagley, 2009). Thus, since contrasting information on motivations for play and research limitations exist in the above mentioned studies (i.e. self-selected sample, low number of participants, and differing scales and measures used), it is clear that a comprehensive understanding of these motivations and their impact on online gaming addiction has not yet been fully explored and warrants continued research. MMORPG play is also related to behavioral learning (Charlton & Danforth, 2007; Doan & Strickland, 2012; Ducheneaut, et al., 2006). For example, the principles of operant conditioning are present in MMORPG’s where the games use an interval-ratio reward system to encourage players to advance (Charlton & Danforth, 2007; Doan & Strickland, 2012; Ducheneaut, et al., 2006). When players first begin playing the rewards are readily available and come quickly; however, as players advance they become much less frequent. There are many rewards players can achieve such as new items, advanced levels, and new world areas to explore, and thus, the next possible reward always appears within the player’s reach. This entices players to keep playing longer than intended or to return to play as soon as possible. However, operant conditioning may not satisfactorily account for why some players are negatively impacted and others are not. Instead, the concept of dualistic passion (Vallerand, et al., 2003) has been suggested as a way to explain the dichotomy (Wang & Chu, 2007). The dualistic model of passion indicates that individuals are most attracted to hobbies and leisure they find independently interesting (Vallerand, et al., 47 2003). Over time these activities become integrated into an individual’s identity in such a way that eventually the player identifies with and may describe him or herself using the label “gamer.” The Role of Passion The dualistic model of passion (DMP) (Vallerand, et al., 2003) expanded upon the concepts of self-determination theory that were originally developed by Deci & Ryan (2000). Self-determination theory, based on a meta-analysis of motivational studies, hypothesizes that people have an innate motivation to grow and develop themself. It states that humans are intrinsically motivated by three overarching psychological needs: 1) autonomy or the feelings of independence and control of one’s own choices and behaviors, 2) competence or the ability to learn and master skills, and 3) relatedness or the desire to relate, belong, and connect to others. According to this theory, people self- select activities that feel comfortable and fulfill these psychological needs. Although extrinsic factors may also influence an individual’s behavior, they are still primarily motivated by these three needs. According to Vallerand, et al. (2003), the DMP suggests that people self-select activities that feel comfortable and fulfill the three basic psychological needs outlined by self-determination theory (Deci & Ryan, 2000). The DMP further conceptualizes that these activities develop into passions and typically become part of the individual’s identity (Vallerand, et al., 2003). For example, DMP theory would state that people do not just simply engage in sports, reading, or strumming a guitar, but rather may become passionate about these activities and eventually become athletes, readers, or guitarists. 50 negative behaviors or effects. This contributes to feeling compelled to engage in a passionate activity even when it is not advisable, may be dangerous, or could have serious consequences. For example, a passionate cyclist may want to ride, but snow and ice have made the terrain dangerous and unsuitable. If this person has internalized this activity as a controlled choice, he/she will feel compelled to find a way to engage in it nonetheless. Lafrenière et al. (2009) explored how the DMP (Vallerand, 2010) relates to online gaming. They sought to understand how harmonious and obsessive passion could help understand the difference between players who experience positive and negative outcomes. More specifically, they researched how these were linked to the gaming experience, the emotions felt during gameplay, and the overall adjustment, as well as how these variables potentially relate to problematic gaming (Lafrenière, et al., 2009). Lafrenière et al. (2009) used a sample of 222 participants comprised of 191 males and 31 females who played MMORPG’s. The independent variables were measures of harmonious passion, obsessive passion, age, and gender. The dependent variables consisted of hours played per week, overall life satisfaction, a measure of self-realization, positive and negative affect, physical symptoms, and problematic behaviors. The instrument the researchers used involved shortened versions from the following six different constructs. Passion was measured using the passion scale (Vallerand, 2010) and the positive and negative experiences were assessed using items from an affect scale (Barrett & Russell, 1998). A problematic gaming scale (Tejeiro & Morán, 2002) was used to identify problem behaviors associated with gaming such as tolerance, withdrawal, and preoccupation. An edited life satisfaction scale (Diener, Emmons, Larsen, & Griffin, 51 1985) measured items related to contentment with current life areas such as family, job/school, and relationships. An overall health questionnaire (Knäuper, Rabiau, Cohen, & Patriciu, 2004) evaluated physical concerns associated with excessive gaming such as sleep disturbance, dry eyes, appetite loss, and dizziness. Self-realization was measured using items asking personal satisfaction, life purpose, personal growth, and positive relationships (Miquelon & Vallerand, 2006). Overall, 40 items were assessed using a 7- point Likert scale (1 = never to 7 = almost always) to gather scores on the various constructs (Lafrenière, et al., 2009). Lafrenière, et al. (2009) then used canonical correlation to assess the variable relationships, a type of multivariate regression analysis used to measure the relationships between independent and dependent variables. Different loadings of the variables were then input to determine which variables produced high factor loadings resulting in the variable of interest. In this case, the variables of interest measured are harmonious and obsessive passion. The first significant canonical correlation set was 0.81, F(28, 754) = 11.68, p < 0.001 (Lafrenière, et al., 2009). The primary independent variable in this set was obsessive passion, which had a loading of 0.99. Dependent variables that had a significant factor loading in this canonical correlation (> +/- .40) were problematic behaviors (.91), positive affect (.65), negative affect (.54), hours played per week (.54), physical health symptoms (.40) and self-realization (-.40). The second significant canonical correlation set was 0.45, F(18, 594) = 3.08, p < 0.001. The primary independent variable in this set was harmonious passion with a factor loading of 0.94. Significant dependent variables in this set included self-realization (.71), life satisfaction (.70) and positive affect (.68). 52 As hypothesized Lafrenière, et al. (2009) found that harmonious passion was positively correlated with self-realization, life satisfaction, and positive affect while gaming. Conversely, obsessive passion was positively correlated with both negative and positive affective experiences, problematic behavior, health, and hours per week played and negatively correlated to self-realization. Both harmonious and obsessive passions were positively correlated with positive affect while playing, but the researchers suggest the nature of these experiences may differ. They concluded that obsessively passionate players experienced relief (positive affect) from playing after a period of not playing, and that harmoniously passionate players experienced joy (positive affect) from the simple act of playing. However, future research in this area would be needed to further extrapolate these results. One of the more significant findings in this study was that obsessive passion was positively correlated with problematic behaviors related to play (Lafrenière, et al., 2009). Since Vallerand (2010) suggested that obsessive passion includes a sense of being out of control, it is not surprising that the two are associated. However, what appears less obvious are the underlying factors and motivations, which differentiated the players who identified as harmoniously and obsessively passionate. It is important to note that although this study provided important information regarding the profiles of harmonious and obsessive passion, it also has several limitations, which may inform similar future research. First, since a correlational study was conducted, causality can only be inferred. A regressional or longitudinal type of study may better inform how predictive these factors are to passion and problematic gaming. Second, the population sample was largely skewed towards men. Third, several 55 Charlton and Danforth (2007), Griffiths et al. (2003), and Ng and Wiemer-Hastings (2005). Kim and Kim (2010) then connected their items to prior research on behavioral addiction, most specifically the six specific core concepts that define addiction as outlined by Brown (1991) and Brown (1993). The first concept is salience, which is when the activity or behavior assumes a dominant and important role in one’s life. Second, euphoria is a “high” experienced from engaging in the activity. Third, the concept of tolerance refers to needing greater periods of engagement in order to achieve the same “high.” The fourth core idea is withdrawal, or when people experience withdrawal symptoms, such as irritability of preoccupation, when not engaging in the activity. The fifth concept indicates there are instances of interpersonal or self-conflict because of continued use. The last concept describes periods of cessation of the behavior followed by a relapse back to the behavior. The researchers then conducted a pilot study creating items based on these six core concepts (Brown, 1991; Brown 1993) and others adapted from previous assessments for gaming use, Internet addiction, and gaming addiction (Kim & Kim, 2010). Some items were eliminated for a variety of reasons, such as lack of variability in answers, a lower Cronbach’s α relative to other items, and questions that had little commonality when compared using a principal axis factoring model. Twenty items remained useable and all spanned across five distinct dimensions defined by the researchers. These dimensions were euphoria, health problems, conflict, loss of self-control, and preference for a virtual relationship. These 20 items comprised the POGUS. Each dimension 56 contained 3-5 questions and the entire survey uses the POGUS. A visual of this scale can be seen in Figure 2.1. Figure 2.1 Problematic Online Game Use Scale: 5 Factors and One Second Order Factor The sample for the POGUS study was extended to three phases and each one tested different age groups of children and adolescents (Kim & Kim, 2010). In the first phase, the sample size was 1422 fifth-grade students from seven different elementary schools (674 females and 748 males) (Mage = 11.88). In phase two the sample size consisted of 199 eight-grade males (Mage = 14.74) and phase three included 393 11th graders (198 females/ 195 males; Mage = 17.87). All subjects were from Seoul, South Korea. The five dimensions of the POGUS accounted for 65.13% of the total item variance and all items in the scale had factor loading of greater than .62 (Kim & Kim, 2010). The researchers also assessed the dimension subscale reliability scores and judged 57 all dimensions to be acceptable (euphoria α = .836; health problems α = .777; conflict α = .811; failure of self-control α = .822; preference for a virtual relationship α = .866). They conducted a goodness of fit test and two fit indices (Tucker Lewis Index (TLI) and the Root Mean Square Error of Approximation (RMSEA)) and found a X2 (165, N = 1422) = 1119.729, p < .001, TLI = .913, RMSEA = .064. They also tested two other models; the first was a single factor analysis (eliminating the five dimensions) where they found a goodness of fit X2 (170, N = 1422) = 3361.361, p < .001, TLI = .718, RMSEA = .115 and the second was a uni-dimensional model (eliminating the Problematic Online Game Use second order) where they found a goodness of fit X2 (160, N = 1422) = 882.228, p < .001, TLI = .932, RMSEA = .056. Although Kim & Kim (2010) discovered a five factor, one second order model and the uni-dimensional model to be good fits, they choose the more restricted model (the first). The researchers used several measures to assess reliability and validity of the POGUS (Kim & Kim, 2010). The scale demonstrates reliability as a confirmatory factor analysis and was used between samples, which contributed to finding a factor loading above .70 for all items. The first sample had a Cronbach’s α = .911, X2 (165, N = 199) = 332.884, p < .001, TLI = .886, RMSEA = .072 and the second sample had a Cronbach’s α = .948, X2 (165, N = 393) = 717.943, p < .001, TLI = .880, RMSEA = .082. They thus determined that the scale could be generalized to similar populations and cultures beyond these samples. The researchers also assessed measures of discriminant validity and convergent validity (Kim & Kim, 2010). To test this, the researchers used assessments for four additional components that have been linked to problematic game play in prior studies. 60 more impactful on adults such as careers, higher education, marriage, and children and how these life experiences may impact play. The addition of these types of questions would be an area for further study. Lastly, the study did not report gender differences for the POGUS. Significant differences in instrument scores could skew future research studies, especially considering that males comprise a larger portion of the sample in this study. Nonetheless, even with the aforementioned limitations, the POGUS is currently one of the most specific and validated measures available for testing online gaming. Current Study The current study intends to combine the above concepts of dualistic passion (Vallerand, et al., 2003; Vallerand, 2010) and player typology (Yee, 2006), and to investigate how they interact and relate to each other. The researcher is also interested in how these components may contribute to and predict online gaming addiction by using Kim and Kim’s (2010) Problematic Online Gaming Use Scale. The result will be assessed using correlations and a multiple regression model, which can contribute insight toward how these constructs relate and their relevance to online gaming addiction. Summary This research literature review is intended to provide a framework and context for the current proposed study. Chemical addiction is a disease with characteristics such as tolerance, withdrawal, and continued use despite negative consequences. Models of addiction include hedonism, negative and positive reinforcement, and incentive motivation. Process addictions, such as pathological gambling, compulsive shopping, and sex addiction, are comparable with chemical addictions. Process addictions share 61 many similarities with chemical addiction such as the characteristics tolerance, withdrawal, and continued use despite negative consequences. Research on process addictions is not as prevalent as they are for chemical addition but there has been growth in the area over the past two decades. The American Psychiatric Association has asked for more research on process addictions, namely Internet gaming. Massively multiplayer online roleplaying games (MMORPG’s) are a genre of games that opens a virtual world allowing dozens to thousands of players to play together at the same time. There are six factors that make MMORPG’s different from other type of games: persistence of the gaming universe, physics relative to the gaming world, social interactions with other players, avatar mediated play (players use a representation of themselves in the game), vertical gameplay (i.e. avatars advance in level, strength, wealth), and perpetuity meaning that the game universe is always on regardless of whether the player is playing in it or not. Yee (2006) offered three primary motivators for playing MMORPG’s. First is achievement in the game through character advancement, defeating other players, or optimizing their character to be the best at their profession or role. Second is socialization with other players through groups or guild and making connections with players outside of the gameplay. Last is immersion in the gaming world as demonstrated by exploration of the game areas, seeking out little known dungeons, or and role-playing as the character. A 38-item Motivation for Play survey was created to assess what factor(s) motivate players. Passion is defined as a strong inclination towards a self-defining activity that one likes, finds important (or highly values), and invests time and energy toward. Passion is 62 dualistic with harmonious and obsessive types. Harmonious passion happens when an activity is freely chosen and a person willingly engages in it without connection to other needs or extrinsic motivations. Obsessive passion is connected to other values, feelings of self, desires, needs, or other meanings about oneself. The person is engaged not only for the enjoyment of the activity but also for what it personally represents for them externally (reputation, status, credibility, etc.) or what it gains for them other than person satisfaction (wealth, status, a job, a partner, etc.). Obsessive passion has been connected to negative consequences and process addictions. Online gaming addiction is defined as a type of process addiction, which involves compulsively playing online games such as MMORPG’s. Kim and Kim’s (2010) Problematic Online Gaming Use Scale (POGUS) is a 20-item instrument used to assess online gaming addiction. The instrument’s validity and reliability measures indicate the instrument can be used in the present study. However, this will be the first time the assessment is used with adults. The researcher is interested in how motivations for play (achievement, immersion, and socialization) and passion (harmonious and obsessive) might predict online gaming addiction. The Motivation for Play Scale (Yee, 2006), Passion Scale (Vallerand, 2010) and POGUS (Kim & Kim, 2010) will be used. Player demographics and playing behaviors will also be recorded. 65 The POGUS demonstrates adequate reliability and validity with adolescents and children. Kim and Kim (2010) conducted a confirmatory factor analysis and found a factor loading above .70 for all items. The first sample they evaluated had a Cronbach’s α = .911, X2 (165, N = 199) = 332.884, p < .001, TLI = .886, RMSEA = .072 and the second sample had a Cronbach’s α = .948, X2 (165, N = 393) = 717.943, p < .001, TLI = .880, RMSEA = .082. They concluded that the scale could be generalizable across similar populations. Discriminant validity and convergent validity was also measured (Kim & Kim, 2010). They compared four components linked to problematic game play, loneliness, academic self-efficacy, anxiety, and life satisfaction, to their current measures. Discriminate and convergent validity can be inferred based upon the strong correlations between these four factors and their relative items. They suggest the POGUS is a distinguishable and unique assessment. Adding additional credibility to the instrument is its inclusion by multiple researchers listing Internet and gaming addiction assessments (King, Haagsma, Delfabbro, Gradisar, & Griffiths, 2013; Kuss & Griffiths, 2012; Pontes & Griffiths, 2014). The scale has only been used for adolescents and children so there is concern regarding its use with this adult sample. However, when this researcher explored all viable options for an assessment, the POGUS was evaluated to be the most comprehensive and best suited for the purposes and scope of this research. The full assessment of the POGUS can be viewed in Appendix B. Motivation for Play The first independent variable is Player Typology as defined as a given score on Yee’s (2006) Motivation for Play (MFP) questionnaire. There are 38 items on the MFP and each is scored on an adapted Likert-scale with seven points (0 = Not Agree at All; 1 66 = Very Slightly Agree; 2 = Slightly Agree; 3 = Moderately Agree; 4 = Mostly Agree; 5 = Strongly Agree; 6 = Very Strongly Agree). The MFP includes three subscales. Participants receive scores in three primary components related to style of play: social, immersion, and achievement. Participant’s scores range from 0 to 66 on the social subscale, 0 to 84 on the achievement subscale, and 0 to 84 on the immersion subscale. Additionally, there are 10 total subcomponents under the three subscales. The achievement subscale has three subcomponents (advancement, mechanics, and competition) and comprises 14 total items. The advancement subcomponent (6 items) asks participants about creating powerful characters, accumulating wealth or rare items, and being well known in the game. The mechanics subcomponent (4 items) includes questions centered on creating “ideal” characters and using the in-game character building system to optimize character performance. The competition subcomponent (4 items) includes questions regarding playing against other players or purposely annoying or provoking other players. The social subscale has three subcomponents (socializing, relationship, and teamwork) and comprises 11 total items. The socializing subcomponent (4 items) contains questions regarding player enjoyment with interacting, chatting, or helping other players. The relationship subcomponent (3 items) includes questions about forming more meaningful connections with other players. The teamwork subcomponent (4 items) includes questions regarding cooperation with other players and joining groups. The immersion subscale has four subcomponents (discovery, role-playing, customization, and escapism) and comprises 14 total items. The discovery subcomponent (3 items) includes questions regarding exploration of the game world and 67 learning about the world’s history. The role-playing subcomponent (4 items) includes questions about players adopting their characters persona in the game, experimenting with different personalities, or creating a history for their character. The customization subcomponent (3 items) includes questions about changing and coordinating a characters visual appearance such as color coordination, tattoos, hair and eye color, or accessories that do not serve a function in the game. The escapism subcomponent (3 items) includes questions pertaining to using the game as a means to relax, relieve stress, or escape from real world problems. The MFP was chosen as an assessment because of the potentially rich data that can be gained given the three primary subscales that are defined as well as the 10 subcomponents that provide further exploration. The primary subscales will provide three independent variables for this study and the subcomponents can provide further insight into more specific components that motivate players. Subcomponents will be analyzed in post hoc analyses. Sample items include: Rate how important it is to you: To acquire rare items that most players will never have? (Achievement; Advancement) That your character is as optimized as possible for their profession/role? (Achievement; Mechanics) To be able to talk to your online friends about personal issues? (Social; Relationship) To explore the world just for the sake of exploring it? (Immersion; Discovery) 70 Agree; 5 = Strongly Agree; 6 = Very Strongly Agree). For the purposes of this study only the harmonious and obsessive passion subscales will be used. Vallerand (2010) added the passion criteria items to differentiate between subjects that are passionate or not passionate on the given subject. Although this could be used as a qualifier, most research using the DMP scale does not use it in such ways and no reports of validity have been found to substantiate use of this subscale. As such, for the purposes of this study subjects will be evaluated on harmonious and obsessive passion. Post hoc analyses will report the differences between the passionate and non-passionate group if there are any. There are four items on the passion criteria subscale and scores on this scale range from 0 to 24. The questions assess activity valuation, time and energy investment, and love or passion for the activity. These questions include I love playing online games Playing online games is important for me The harmonious passion subscale has six items and scores range from 0 to 36. Sample items from the harmonious subscale include Online gaming is well integrated in my life Online gaming is well integrated in my life Online gaming reflects the qualities I like about myself The obsessive passion subscale has six items and scores range from 0 to 36. Sample questions from the obsessive passion subscale include I have the impression that online gaming controls me Online gaming is so exciting that sometimes I lose control over it The developers of the initial Passion Scale (Vallerand, et al., 2003) reported that 71 the scale demonstrates sound factorial validity supporting the use of the harmonious and obsessive passion scales. An exploratory factor analysis of the initial 34 items helped eliminate 20 items that either loaded on both the harmonious passion and obsessive passion subscales or items that had weak loadings. A second confirmatory factor analysis found 14 items had significant loadings and the model was a good fit X2 (76, 235) = 171.70, p <.001. Reliability was measured using the Cronbach alpha of the two subscales (Obsessive Passion α = .89 and Harmonious Passion α = .79). Since the initial validity and reliability tests, the scale has been amended. Twelve of the original items are still used (six for each subscale) and four qualifying questions were added. The researchers added the qualifying subscale, passion criteria, to help quantify passion among participants before dividing them further with the obsessive and harmonious passion subscales. According to the researchers, subjects who reach a threshold of 16 on the passion criteria scale are labeled as passionate about the activity or behavior. The subscale has not been widely used and no validity measures could be determined for the subscale so for purposes of this study the variable will be used in post hoc analyses only. The full Passion Scale can be seen in Appendix D. Demographics and Game Play Assessment Participants were asked for demographic information. Specific information requested was subject’s age, sex, level of education, marital status, personal income, and racial identification. Categories and tiers for these questions are consistent with the U.S. census. Game play questions include what specific MMORPG’s subjects have played in the past six months, how many total months they have been playing MMORPG’s, how many days they play MMORPG’s in a week, how many hours they play MMORPG’s in a 72 day, what time of day they typically play MMORPG’s, and who initially introduced them to MMORPG’s. This information will help to create comparative groups such as male and female players or younger and older groups, which can be informative in post hoc analyses. Demographic questions can be seen in Appendix A. Participants The population for this study is adult MMORPG players. This study is not concerned with which online game players prefer but rather their playing behavior. A MMORPG player is defined as an adult online game player who engages actively (logs in and plays) a defined MMORPG and has done so at least once in the past six months. The data collection software, SurveyGizmo.com, qualified participants by asking questions regarding age and games played. If a participant indicated that they are under 18 years old or have not entered a qualified MMORPG (as listed on xmmorpg.com, 2013) they will not be permitted to complete the survey and will be notified that they do not qualify for this study. Since there was an option to enter additional MMORPG’s in a narrative space or MMORPG’s played, the researcher evaluated these answers for qualification. Any questionnaires returned without a qualifying MMORPG were removed from analysis. The sample was gathered by advertising on popular online gaming website portals such as wow.allakhazam.com (2013). According to websitelooker.com (2013), this website attracts 30,484 visitors each day, which makes it one of the largest non-specific MMORPG related website. It was chosen for the amount of traffic it attracts and that it attracts players across different MMORPG games. This ideally would give the sample a 75 Procedure The study began in the winter of 2014 and data collection was completed April of 2014. Data analysis spanned April 2014 to April 2016. Submission for approval to the Internal Review Board (IRB) at The Ohio State University was submitted and can be seen in Appendix E. Upon approval the completed questionnaire was fully developed using the adapted items. The questionnaire was be digitized so that it is accessible and easily used online. The web service site surveygizmo.com (2013) was used for data collection. This site was chosen for its ability to transfer data securely while offering the needed flexibility in survey development. The website is able to download the attained data to a secure and confidential database which was encrypted and only assessable by the researcher. Participants were provided a disclosure statement at the beginning of the survey and asked to provide consent to participant in the study in accordance with IRB standards. After giving consent the participant was asked to provide demographic information and questions from the POGUS, MFP, and Passion Scale. To prevent participants taking the survey multiple times, surveygizmo.com tracked IP addresses of computers used to take the survey and did not allow the same computer to take the survey twice. The researcher could also see IP addresses of participants and checked that no duplicates existed. The questionnaire was advertised until the sample was obtained (approximately six weeks) on the above-mentioned websites and online portals. The advertisement incentivized participation with the potential gift of a $100 Visa Giftcard, which was awarded to one winner. Contact information was obtained separate from the participant’s 76 answers and was not linked to individual answers. The researcher conducted a random drawing using a random number generator. The winner was identified and sent the prize via certified mail. Cost of advertising was approximately $277 for websites and $76 per month for data gathering services via surveygizmo.com. The researcher covered all expenditures. Individual surveys were assessed for usability. Any questionnaires that did not meet participant requirements or appear to be random in answering (nonsensical patterns such as answering all 1’s) were discarded and not included in the data analysis. The data was entered into the Statistical Package for Social Sciences (SPSS)-21 statistical analysis software system. Recommended Data Analysis Before data analysis can begin, the data set will need to be assessed for normality. Tests for independence, homoscedasticity, linearity, and reliability will be conducted. Assuming the data set is normal each of the four research questions will be answered. Descriptive statistics will be reported to help answer research question one. This will include means, standard deviations, and frequencies for variables age, sex, education, race, and marital status. Game play statistics such as total length of play, weekly and daily play, and when time of play will be reported, as well. Crosstabs will help report the descriptive statistics. Means and standard deviations of the independent variables Player Typology (immersion, socialization, and achievement) and Passion (harmonious passion and obsessive passion) will be reported. A Pearson product-moment correlation coefficient will be conducted on these variables as well to determine correlation relationships. This 77 will help answer research questions two and three. For research question four, the chosen method of data analysis is multiple regression. This method will allow the researcher to investigate how independent variables (achievement, immersion, social, obsessive passion and harmonious passion) predict the dependent variable (level of gaming addiction). It also allows for the differing variables to be examined individually through part and partials. The ability to closely evaluate individual components is a strength of multiple regression analysis and will allow for a variety of predictive models to be created as an outcome of this research. The “enter” method of entry will be used for the regression. This will allow for evaluation of each variable’s beta weight. Since the primary goal of this study is to understand the predictive strength of player typology and passion on addiction, it made multiple regression an ideal analysis for this study. 80 Table 4.1 Descriptive Statistics Variable Mean Std. Dev. Skewness Std. Error Kurtosis Std. Error Achievement 36.42 17.06 .371 .155 -.361 .309 Social 39.93 15.70 .106 .155 -.672 .309 Immersion 36.80 17.12 .021 .155 -.827 .309 Harmonious Passion 17.59 7.92 .213 .155 -.426 .309 Obsessive Passion 6.70 7.33 1.369 .155 1.356 .309 POGUS 32.67 22.03 .914 .155 .343 .309 n = 246 The samples appear to represent a normally distributed data set. Evaluating the histograms of the seven variables evidences this. Additionally, skewness and kutosis measurements were calculated and all were demonstrated to be within normal and acceptable range (< 3). Obsessive passion was slightly positively skewed (1.369) and platykurtic (.914) representing the variable with the highest scores on either measure. Table 4.2 displays the histograms and normalcy curve for each variable. P-P plots were calculated for each variable. Results can be seen in table 4.3. The plots appear normal for most variables except obsessive passion. However, similar to the histogram, given the nature of the variable (likely to impact a smaller, more concentrated sample) it is expected to be skewed and slightly irregular. Otherwise, there appear to be no concerns. Additionally the scatterplot appears to demonstrate a homoscedastic data set. The scatterplot can be seen in table 4.4. There appears to be independence of subjects. Overall, the data set appears to meet normality standards Frequency Table 4.2 Histograms for Independent and Dependent Variables Achievment Social Score ‘Achievement Score Harmonious Passion Obsessive Passion > - . i} Geese ae Fe crnitiicon vamonious rasan Sore immersion problema One Canin Usseale / Bigsie / : {| z f z i i. N ‘ li 81  82 Table 4.3 P-P Plots of Independent and Dependent Variables