Pokémon GO 2016: Exploring Situational Contexts of Critical Incidents in Augmented Reality

This is an electronic reprint of the original article. This reprint may differ from the original in pagination and typographic detail. Author(s): Title: Year: Version: Please cite the original version: All material supplied via JYX is protected by copyright and other intellectual property rights, and duplication or sale of all or part of any of the repository collections is not permitted, except that material may be duplicated by you for your research use or educational purposes in electronic or print form. You must obtain permission for any other use. Electronic or print copies may not be offered, whether for sale or otherwise to anyone who is not an authorised user. Pokémon GO 2016: Exploring Situational Contexts of Critical Incidents in Augmented Reality Kari, Tuomas Kari, T. (2016). Pokémon GO 2016: Exploring Situational Contexts of Critical Incidents in Augmented Reality. Journal of Virtual Worlds Research, 9(3). https://doi.org/10.4101/jvwr.v9i3.7239 2016 Volume 9, Number 3 Edge December 2016 Editor In Chief & Issue Editor Yesha Sivan Tel Aviv University The Coller Institute of Venture Coordinating Editor Tzafnat Shpak Cover image: Marco Verch. Sony: Project Morpheus https://www.flickr.com/photos/30478819@N08/19701245783/ The JVWR is an academic journal. As such, it is dedicated to the open exchange of information. For this reason, JVWR is freely available to individuals and institutions. Copies of this journal or articles in this journal may be distributed for research or educational purposes only free of charge and without permission. However, the JVWR does not grant permission for use of any content in advertisements or advertising supplements or in any manner that would imply an endorsement of any product or service. All uses beyond research or educational purposes require the written permission of the JVWR. Authors who publish in the Journal of Virtual Worlds Research will release their articles under the Creative Commons Attribution No Derivative Works 3.0 United States (cc-by-nd) license. The Journal of Virtual Worlds Research is funded by its sponsors and contributions from readers. http://jvwresearch.org Pokémon GO 2016: Exploring Situational Contexts in AR 1 Edge / Dec. 2016 Journal of Virtual Worlds Research Vol. 9, No. 3 Volume 9, Number 3 Edge December, 2016 Pokémon GO 2016: Exploring Situational Contexts of Critical Incidents in Augmented Reality Tuomas Kari University of Jyvaskyla, Finland Faculty of Information Technology Abstract Pokémon GO, an augmented reality mobile game, captured the attention of millions of people around the world in July 2016. Various sources from around the globe have reported both positive and negative incidents and outcomes related to the game. Some of the incidents have been particularly remarkable for the player, i.e., critical incidents. A critical incident is a single experience, which a person perceives or remembers as unusually positive or negative. Critical incidents typically are highly influential for human behavior, and thus, important to study. Playing augmented reality games can take place in varying situational contexts. Situational context includes information that can be used to characterize the situation of a person, place, or object, and has been shown to be influential in mobile use context. This study investigates in which kinds of situational contexts do critical incidents of augmented reality game Pokémon GO take place. The focus is on four different situational contexts that Pokémon GO can be played in: sociality, interaction state, place, and reasons for playing. The study is based on analysing an online survey sample of 226 responses. The findings pose insights and implications regarding augmented reality applications and games in general, and thus, assist the developers in their efforts to provide the users with meaningful and positive experiences with games and other augmented reality applications. http://jvwresearch.org Pokémon GO 2016: Exploring Situational Contexts in AR 2 Edge / Dec. 2016 Journal of Virtual Worlds Research Vol. 9, No. 3 1. Introduction Pokémon GO (Niantic, Inc., 2016) is a mobile game in which the purpose is to find Pokémon creatures and battle others with the collected Pokémon. The game was launched in July 2016 and was quickly adopted around the world: according to market intelligence firm Newzoo (2016), Pokémon GO was downloaded 550 million times in first 80 days. The game also had more downloads in its first week in the Apple App Store than any other app in history (Polygon, 2016). By November 2016, the total number of downloads had exceeded 600 million (Kotaku, 2016). However, since its peak, the download trend has turned into decline, falling to less than 10 million monthly downloads in November 2016 (BBC, 2016). Also, as is merely natural to mobile games in general, similar downfall has occurred with the player base (SurveyMonkey, 2016). Nevertheless, millions were still playing the game and making in-app purchases four months after the game was published (BBC, 2016). Pokémon GO was estimated to have generated $950 million in revenues during 2016 (Venturebeat, 2017). Pokémon GO uses an augmented reality mechanics (i.e., supplementing the real world with animated characters). The game requires navigating in the real world setting and thus requires the player to be physically active in order to play the game (Baranowski, 2016). These kinds of games requiring physical activity to play the game are referred to as exergames (Kari & Makkonen, 2014). Exergames themselves are widely studied in regard to physical activity but limitedly in mobile context (Kari, 2014). There have already been reports about the positive influence of Pokémon GO to physical activity and increased exercise (e.g. Althoff, White, & Horvitz, 2016; Howe et al., 2016; Nigg, Mateo, & An, 2017; Serino, Cordrey, McLaughlin, & Milanaik, 2016), outdoor activity, cultural and historical awareness, socialization (Serino et al., 2016), and its ability to help those suffering from severe social withdrawal (Tateno, Skokauskas, Kato, Teo, & Guerrero). The game has also been suggested to lead people to replace sedentary indoor screen-time with active outdoor time (LeBlanc & Chaput, 2016; Nigg et al., 2017). Although, question has been raised regarding the sustainability of the increased physical activity (Howe et al., 2016). Despite its many benefits, there have also been reports of negative incidents that have occurred while playing the game, such as trespassing, violence (Serino et al., 2016), and accidents and injuries (Joseph & Armstrong, 2016). The study by Colley et al. (2017) also suggests there are both positive and negative aspects in Pokémon GO. They report, for example, that the game’s design fortifies present geographicallylinked biases, that the game may have instigated a fairly uncommon shift in global human mobility patterns, and that the game has geographically-linked safety risks (Colley et al., 2017). Overall, we have heard reports on both the positive and negative incidents of the game (McCartney, 2016). Studying these kinds of positive and negative effects of the game is of course important, but investigating also other aspects of the game, or through the game, is highly warranted. For example, studying the critical incidents of Pokémon GO and augmented reality mobile games in general is very important when considering the positive and negative reports regarding Pokémon GO. However, this kind of research is still very limited. Therefore, this study aims to fill this gap and provide further insights on the matter, hence increasing the theoretical and practical understanding of this specific phenomenon. The main research question of this study is: In what kinds of situational contexts do critical incidents of augmented reality game Pokémon GO take place? An answer to this question can pose important findings and implications regarding not just the game itself but also augmented reality applications and games in general, as well as mobile exergames. Therefore, this study increases the theoretical understanding on the usage experiences of these kinds of games and also aids the developers in their efforts to provide the users with meaningful and positive experiences with games and other augmented reality applications. The findings can also assist the developers in achieving the positive incidents and in avoiding the http://jvwresearch.org Pokémon GO 2016: Exploring Situational Contexts in AR 3 Edge / Dec. 2016 Journal of Virtual Worlds Research Vol. 9, No. 3 negative ones. This paper consists of six sections. After this introduction, the concepts of situational context and critical incident are presented, followed by methods, results, discussion, and finally, limitations and future research. 2. Background 2.1 Situational Context Situational context includes “information that can be used to characterize the situation of an entity. An entity is a person, place, or object that is considered relevant to the interaction between a user and an application, including the user and applications themselves” (Dey, 2001, p. 5). Situational context is something that varies depending on the product or service under use. For example, the situational contexts between using augmented reality mobile games, or virtual reality console games can be different in many ways. Whereas mobile gaming can take place outdoors in the city or nature, console gaming practically always takes place indoors. It has been suggested that the underlying situational context can influence the behavior after a product or service experience (Mehrabian & Russell, 1974; Xiao & Benbasat, 2011). Prior research has indeed demonstrated that situational context is important in mobile use context (e.g. Hong & Tam, 2006; Liang & Yeh, 2011; Liu & Li, 2011; Mallat, Rossi, Tuunainen, & Öörni, 2009; Salo & Frank, 2015). For instance, Salo and Frank (2015) found that the situational context of critical mobile application incident can influence use continuance and word-of-mouth. This study focuses on four different situational contexts: sociality, interaction state, place, and reasons for playing. Three of the situational contexts (sociality, interaction state, place) were derived from Salo & Frank (2015), who studied situational contexts of critical incidents in mobile application usage. In addition, the context reasons for playing was investigated in this study. Reasons for playing - refers to the primary reasons why the user has taken up the action itself, in this case, playing Pokémon GO. Considering the exergaming nature of Pokémon GO and many other augmented reality mobile games, the reasons for playing were investigated specifically from the point of view that the game was played mainly for fun, exercise, or both. Previous research has shown that with exergames, and thus, with Pokémon GO, the reasons for playing can be hedonic (e.g. fun), utilitarian (e.g. exercise) (Osorio, Moffat, & Sykes, 2012), or both fun and exercise simultaneously (Berkovsky, Coombe, Freyne, Bhandari, & Baghaei, 2010). Sociality - seems to play an important role in Pokémon GO (Serino et al., 2016; Tateno et al., 2016). In this study, sociality refers to whether the game is played alone individually or together with others. Prior research has shown that the pleasure of using a system can come from either the use itself or from interaction with others. Hence, sociality can be influential regarding the user experience (Junglas, Goel, Abraham, & Ives, 2013). Interaction state - refers to the state of mind in which the user conceives the activity (Apter, 1989) of playing Pokémon GO. For interaction state, this study applies Apter’s (1989) reversal theory. It presents a pair of meta-motivational systems, “telic” and “paratelic”, which steer human behavior. These meta-motivational systems are presented as two different states: the “telic” taskoriented state and the "paratelic” activity-oriented state. They indicate whether one is motivated by achievement and future goals or the enjoyment of process in the moment. The task-oriented state is more serious and the activity-oriented state more playful. When in the task-oriented state, a person has some important goal or task in mind, whereas in the activity-oriented state, the person has the behavior itself in mind (Apter, 1989). An example of playing Pokémon GO in a task-oriented state is when a player’s mind is set to beating an opponent or to achieve something. An example of an activity-oriented state is spontaneous playing with just the enjoyment in mind. http://jvwresearch.org Pokémon GO 2016: Exploring Situational Contexts in AR 4 Edge / Dec. 2016 Journal of Virtual Worlds Research Vol. 9, No. 3 Place - describes the real world setting in which the game is played. Pokémon GO and other augmented reality mobile games can be used in various places (Althoff et al., 2016; American Heart Association, 2016; Serino et al, 2016). In this study, place has been categorized into home, nature, public setting with permitted access, public setting with forbidden access, car, and other places. The selected situational contexts are extensive but obviously not all-inclusive. Other influential situational contexts, for example, the emotional state (Lee, Kim, & Kim, 2005), have also been shown to influence the usage of technology. And according to Chávez, Ide, and Kirste (1999), information about context can be practically unlimited. In this study, the focus was on those contexts that are typically present when playing Pokémon GO. Therefore, these contexts are considered to include the most central ones involved with the actual gaming experience, and hence, the ones most essential to understand. An example (derived from the data of this study) of a contextual setting while playing Pokémon GO is: playing Pokémon GO for fun in nature together with others in an activity-oriented state. 2.2 Critical Incident A critical incident is defined as a single experience, which a person perceives or remembers as unusually positive or negative (Edvardsson & Roos, 2001). Critical incidents, typically, are particularly influential for human behavior (Flanagan, 1954). For example, a single critically negative incident may overrule a set of average positive incidents and lead to discontinuance or other unwanted behavior with the service or product in question (Cenfetelli, 2004). Critical incidents have a significant role in creating customer relationships and establishing user perceptions towards the products, services, and providing companies (Edvardsson & Strandvik, 2000; Payne, Storbacka, & Frow, 2008). Thus, examining critical incidents can have important implications for both research and practice. Previous research has examined critical incidents of mobile applications (e.g. Salo & Frank, 2015) and services (e.g. Gummerus & Pihlström, 2011; Salo, Olsson, Makkonen, Hautamäki, & Frank, 2013), but to author’s best knowledge, not of mobile augmented reality games. 3. Methods This study followed a quantitative approach. The data was collected between middle of July and early November 2016 by using an online survey. The survey was English language and meant for international audience. The survey was created with the LimeSurvey 2.05+ software. Before the survey was launched, it was pre-tested with ten early adopters of Pokémon GO, based on which few small modifications were made. Invitation to participate in the survey was distributed through social media (Facebook, Twitter, and Reddit) and different discussion forums with varying topics (e.g. gaming, wellness, lifestyle, cooking, culture, young elderly, etc.) in order to reach a wide range of players. Obviously, this method has the limitation of only being able to reach those who are following the mentioned social media channels or the discussion forums used to distribute the survey. Also, certain populations are less likely to have Internet access and to respond to online questionnaires. However, as Pokémon GO requires data connection in order to play the game, the lack of Internet access most probably does not rule out any target respondents. Only those respondents who had experience of playing Pokémon GO were asked the questions concerning critical incidents and their situational context. To collect the contextual information about the critical incidents, an exact wording of previous critical incident research (e.g. Bitner, Booms, & Tetreault, 1990; Meuter, Ostrom, Roundtree, & Bitner, 2000) was used: “Think of a time when you had an outstandingly positive or negative experience [with Pokémon GO]”, followed by a series of questions; first, about whether the incident was positive or negative, and then more detailed questions regarding the critical incident and its situational context. The descriptive questions http://jvwresearch.org Pokémon GO 2016: Exploring Situational Contexts in AR 5 Edge / Dec. 2016 Journal of Virtual Worlds Research Vol. 9, No. 3 regarding the situational context were all structured multiple-choice questions, and those can be found in the appendix. The survey also included other questions for the purpose of another study. The collected data was quantitatively analyzed with the IBM SPSS Statistics 24 software. The ‘Cannot say’ responses to the situational context questions were excluded from the analysis of the particular context, e.g. if a respondent was unable to state the interaction state of the incident, that response was excluded from the analysis of the interaction state. The study also investigated the differences between the situational context responses and the positivity/negativity of the incident. The statistical significance and the strength of the dependencies between the responses and positivity/negativity of the incident were analyzed through contingency tables (crosstabs), the Pearson’s χ2 tests of independence, and the Cramér’s V coefficients. In some instances, the common condition for the validity of χ2 test of "No more than 20% of the expected counts are less than 5 and all individual expected counts are 1 or greater" (Yates, Moore, & McCabe, 1999, p. 734) was not met. Consequently, as proposed in the widely used guidelines by Agresti (2002), the results of Pearson’s χ2 tests were advanced with (Monte Carlo) exact tests. The Monte Carlo tests were based on a 10 000 sampled tables and 99 % confidence level. This procedure is considered reliable and independent of the dimension, allocation, distribution, and the balance of the analyzed data (Mehta & Patel, 2012). The level of significance was set at p<0.05 level. 4. Results In total, the survey received 226 complete responses, which were used for the analysis. The sample turned out to be female-dominant, as 66.4 % of the respondents were female. The age of the respondents ranged from 12 to 64 with mean age being 28.8 years old. More detailed description of the sample is presented in Table 1. Table 1: Description of the sample (N = 226) n % Gender Male 72 31.9 Female 150 66.4 Other 4 1.8 Age -20 years 26 11.5 21-30 years 136 60.2 31-40 years 35 15.5 41- years 29 12.8 Employment status Full-time student 80 36.9 Employee 110 50.7 Unemployed 11 5.1 Other 16 7.4 N/A 9 - Household type One person household 67 31.0 One family household without kids 85 39.4 One family household with kids 61 28.2 Other 3 1.4 N/A 10 - http://jvwresearch.org Pokémon GO 2016: Exploring Situational Contexts in AR 6 Edge / Dec. 2016 Journal of Virtual Worlds Research Vol. 9, No. 3 Out of all the reported 226 critical incidents, 176 (77.9 %) were positive and 50 (22.1 %) were negative. The distribution of all the critical incidents as well as the positive and negative incidents between different situational contexts is presented in Table 2. Table 2: Distribution of critical incidents between situational contexts n % Positive % Negative % Reasons for playing (n = 222) Fun 119 53.6 51.7 60.9 Only exercise 0 0 0 0 Both fun and exercise 103 46.4 48.3 39.1 N/A 4 - - - Sociality (n = 220) Alone 66 30.0 23.7 53.2 Together with others 154 70.0 76.3 46.8 N/A 6 - - - Interaction state (n = 214) Activity-oriented 165 77.1 77.6 75.0 Task-oriented 49 22.9 22.4 25.0 N/A 12 - - - Place (n = 224) Home 15 6.7 5.7 10.4 Nature 43 19.2 18.8 20.8 Public setting (permitted) 148 66.1 68.8 56.3 Public setting (forbidden) 3 1.3 0.6 4.2 Car 9 4.0 4.0 4.2 Other 6 2.7 2.3 4.2 N/A 2 - - - Table 3 summarizes the results of the Pearson’s χ 2 tests of independence, Monte Carlo exact tests, and Cramér’s V, which were used to examine the statistical significance and strength of dependencies between the responses and the positivity/negativity of the incident. To clarify, these tests indicate whether the positivity/negativity of the incident was related to the response and how strong was the relationship. With sociality and place, the common condition for the validity of χ2 test was not met, and the (Monte Carlo) exact test result was referred to. Table 3: Positivity/Negativity dependencies on situational contexts N χ 2 df p p (Monte Carlo) V Reasons for playing 222 1.232 1 0.267 0.321 0.074 Sociality 220 15.329 2 <0.001 0.001 (exact) 0.264 Interaction state 214 0.139 1 0.710 0.834 0.025 Place 224 6.395 5 0.270 0.257 (exact) 0.169 χ2 = Pearson’s χ2; df = Degrees of freedom; p = P values for Positivity/Negativity dependencies; P(Monte Carlo) = P values with Monte Carlo exact tests; V = Cramér’s V. Level of significance was p<0.05