Perceived Game Realism: A Test of Three Alternative Models

Abstract

Perceived realism is considered a key concept in explaining the mental processing of media messages and the societal impact of media. Despite its importance, little is known about its conceptualization and dimensional structure, especially with regard to digital games. The aim of this study was to test a six-factor model of perceived game realism comprised of simulational realism, freedom of choice, perceptual pervasiveness, social realism, authenticity, and character involvement and to assess it against an alternative single- and five-factor model. Data were collected from 380 male digital game users who judged the realism of the first-person shooter Half-Life 2 based upon their previous experience with the game. Confirmatory factor analysis was applied to investigate which model fits the data best. The results support the six-factor model over the single- and five-factor solutions. The study contributes to our knowledge of perceived game realism by further developing its conceptualization and measurement.

Introduction

Perceived realism is regarded as a key concept in explaining the mental processing of media messages and possible media effects.^1–3 It has among others been related to aggressive behavior,⁴ the cultivation hypothesis,⁵ enjoyment,⁶ stress reduction,⁷ spatial learning,⁸ and the effectiveness of persuasive communication⁹ and edutainment.¹⁰ Although the construct has received a significant amount of attention as a moderator, not a lot is known about its dimensional structure. While a majority of researchers have operationalized perceived realism as a unidimensional concept,^11–15 others have argued that reality judgments are made along several dimensions and should be investigated accordingly.^3,5,16,17 Furthermore, the findings of studies that include perceived realism are often unstable because of varying operational and conceptual definitions. Within the context of digital game realism, these difficulties have become increasingly prominent. While seminal studies have been conducted in the field of television research,^{5,16–18,19–21} perceived realism has only sparsely been addressed in relation to digital games.^3,22–24 The current study builds upon literature relating to the boundary between the virtual and the real to further our understanding of the dimensional structure and operationalization of perceived game realism. Confirmatory factor analysis (CFA) is used to test which of three alternative models—one unidimensional and two multidimensional—conforms best to the realism judgments of players of the game Half-Life 2 (HL2).²⁵

Perceived Game Realism

Although authors from various disciplines have empirically investigated related constructs, such as presence,²⁶ wishful identification,² and engagement,²⁷ the explicit study of digital game realism has mainly been performed from a theoretical or qualitative perspective.^{3,22,23,28,29} In these studies, the authors opposed to the transfer of concepts from the domain of television studies to that of digital game studies and to the excessive attention that has been directed toward the representational aspects of digital games. More specifically, a plea was made for an extensive framework that integrates rule-based and narratological aspects.

Recently, three studies have synthesized the literature on perceived game realism and addressed the dimensional structure of the concept. Shapiro et al.³ made a theoretical distinction between four elements that are constitutive of the overall reality impression of a game: typicality, character types, character judgments, and emotion. Malliet,²³ using in-depth interviews with gamers, identified five constituents of perceived game realism: factuality, authenticity, character involvement, perceptual pervasiveness, and digital games as a virtual experience. Ribbens and Malliet²⁴ presented a theoretical typology based upon a supplementary literature review and performed a first empirical test of the dimensional structure of perceived game realism using exploratory factor analysis. The analysis revealed six dimensions. First, simulational realism emphasizes the importance of the programmed rules of digital games and the different types of behavior that are possible within these rules.^30,31 In particular, behavior facilitated by the game mechanics that is considered useful in real life may be judged as realistic.³² Second, freedom of choice refers to the feeling of coauthorship that arises during gameplay and highlights the similarities between making choices in a game and making choices in real life.^23,33 A third dimension, character involvement, relates to the sensation of being embodied within the video game world through engagement with an avatar, and consequently, describes the experience of in-game occurrences as comparable to the experience of real-life occurrences.^34,35 Fourth, perceptual pervasiveness identifies the sensory (auditory, visual, and tactile) intensity characterizing many contemporary games as a contributor to the overall reality impression.¹⁶ Fifth, social realism highlights the extent to which the game content and game rules are consistent with the actual world.³⁶ Finally, authenticity refers to a credible depiction of the fictional world.²⁴ Although this component was mainly comprised of items that were related to the plausibility of the depicted events or characters, this result should be interpreted tentatively because of problems with the operationalization of the underlying construct. The original conceptualization of authenticity forwarded by Malliet²³ refers to the players' belief in the game designers' intention and ability to convey an honest or real message. The virtual world of a game should not necessarily have to match the player's expectations about real life to be judged as realistic. Because players believe the designers have created a coherent fictional environment, they judge the environment as being more realistic.

Despite the insights produced in the above-mentioned studies, there are still several issues that need to be resolved. First, the results of Ribbens and Malliet²⁴ should be interpreted cautiously, as at the time of writing it is the only published study that empirically explored the dimensional structure of perceived game realism. To validate a model, a confirmatory analysis is required.³⁷ Furthermore, Ribbens and Malliet²⁴ have studied perceived game realism at the level of the medium, which resulted in a substantial amount of random variance as a consequence of the aggregation of the respondents' conceptualizations of digital games. While some respondents may have used Super Mario Galaxy³⁸ as a standard to judge the Likert-type questions, others may have used a complex sandbox game, such as Grand Theft Auto.³⁹ The authors therefore suggested, in line with the game studies paradigm, that future research should measure perceived realism at the level of the genre or game. Additionally, a plea was made to refine the items used to operationalize the proposed dimensions in terms of coherence and level of abstraction. The first aim of this study was therefore to refine the operationalizations of Ribbens and Malliet²⁴ and to validate the six-factor model using CFA. The realism judgments in this study were limited to one game to avoid the aggregation of a multitude of evaluation standards in the analysis.

Second, in television research, the different aspects of perceived realism are sometimes, although incorrectly according to some,^5,16–18 subsumed under one dimension.^13,14 Within game research, the distinction between a unidimensional and multidimensional structure of perceived game realism has never been subject to an empirical test. Although several authors advocate for a multidimensional conceptualization,^3,23,24 an empirical assessment is required to support this argument. In this study, we examine whether the six-factor model proposed by Ribbens and Malliet²⁴ conforms more strongly to the realism judgments of players than a unidimensional operationalization of perceived realism that is in line with practices within television research.^11–15

Third, usually a distinction is made between an external (the real world is used as a referent of one's reality judgment) and internal (logic and consistency of the fictional world are used as referents of one's reality judgment) component of perceived realism.^21,36 Although little empirical research has clarified how realism judgments are processed, the use of an internal rather than an external referent to assess the realness of a media experience can result in an increased enjoyment.³⁶ The use of internal realism judgments to disregard the artificiality of a message can be particularly important in the context of digital game play. Malliet²³ pointed out that most gamers are aware of the mathematical algorithms that produce the output on the television or computer screen. Moreover, as perceived realism is correlated with both transportation⁶ and wishful identification,² internal reality judgments can be instrumental in enhancing immersion and fun while playing a digital game. However, the conceptual distinction between an external and internal dimension of perceived realism has been disputed. Popova¹⁸ claims that external and internal reality judgments are not made independently from one another and that, following the limited capacity model of motivated media message processing,⁴⁰ stimuli are not matched to different referents of reality. To date, little empirical research has provided support for either perspective. In this article, we examine whether a dimensional structure that subsumes external (i.e., social realism) and internal realism (i.e., authenticity) judgments under one latent construct provides a better fit to the data than one that treats external and internal reality judgments as separate dimensions.

Most theories of perceived game realism have been developed by reference to games that feature complex virtual worlds, such as Grand Theft Auto²³ or Everquest.³ We therefore expect that the six-factor solution may apply to games from genres, such as shooter games, role playing games, and adventure games. This study will test the six-factor solution with HL2, a First-Person-Shooter (FPS) that takes place in a science fiction scenery and includes a mixture of settings, props, and game rules that do not necessarily have a real-life counterpart. HL2 was chosen for three reasons. First, shooter games are highly popular and perceived realism is often suggested as a moderator in the negative effects of violent game play.^4,5 Second, FPS are mainly targeted at male, hardcore gamers. Research has indicated that a subset of these players can be successfully targeted via online game forums.⁴¹ Third, the mixture of real life and fantasy elements makes HL2 an interesting case to examine the distinction between the internal and external component of perceived realism.

In sum, within this article, we apply CFA to address three research questions with regard to the dimensional structure of perceived realism in HL2.

(1) Does the six-factor model proposed by Ribbens and Malliet²⁴ provide a better fit to the data than an alternative one-factor model of perceived realism?

(2) Can the six-factor model of Ribbens and Malliet²⁴ be confirmed when it is tested in relation to one specific game rather than to games in general?

(3) Does the six-factor model proposed by Ribbens and Malliet²⁴ provide a better fit to the data than an alternative five-factor model that subsumes external and internal realism judgments under one latent variable?

Methods

Research has indicated that online research methodologies are useful to examine the psychosocial aspects of digital game play.⁴¹ As players of FPS are often male, hardcore gamers, and familiar with the Internet,^42,43 a web survey distributed via Flemish online game forums was considered as an appropriate method. For each of the forums, a new thread was opened that explained the goal of the study. Anonymity was assured and a hyperlink to a standardized questionnaire on a secure server was provided. No inclusion criteria were stated with the exception of having played HL2.

Realism measurements

Each of the six dimensions that came forward in the findings of Ribbens and Malliet²⁴ was measured using a five-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). Since one goal was to improve the measurements of the latent variable constructs, up to eight items were used. The measures used in Ribbens and Malliet²⁴ provided the basis for the measures used in this study. However, adaptations were made to 1. make the items consistent with regard to the level of abstraction and 2. to improve coherency between the items. Most notably, the items for measuring authenticity were modified to coherently reflect the conceptualization of Malliet.²³

Analysis

We first conducted exploratory principal components analyses with an oblique rotation (Direct Oblimin) per preconceived dimension to reduce the number of items. The factor solution of each preconceived dimension yielded one distinct and reliable factor. Depending on the factor loadings, four or five items were retained per dimension. Deleted items can be found in appendix.

Subsequently, the one-, five-, and six-factor models were tested using CFA (AMOS19). The chi-squared-to-degrees-of-freedom ratio (χ²/df), the comparative fit index (CFI), the Tucker–Lewis coefficient (TLI), the root mean square error of approximation (RMSEA), and the Akaike's information criterion (AIC) were used to assess model fit. Recommended cutoff points for these indices are 0.90⁴⁴ or 0.95⁴⁵ for CFI and TLI while RMSEA values as high as 0.06⁴⁵ and χ²/df values as high as 3 and preferably 2 indicate good fit.^45,46 Finally, AIC denotes superior linkages with a lower value.³⁷

Results

Descriptive statistics

Five hundred sixty one respondents opened the link to the survey. Respondents who left more than four values blank were excluded from the analysis, which resulted in a sample of 382. As only two women were present in the sample, they were excluded from the analysis. Remaining missing data with regard to the items of perceived realism (N=0.004 percent) were replaced with sample means. The mean age of the respondents was 22.7 (SD=5.3) years. About 59.4 percent of the respondents were enrolled in education: 23.4 percent in high school and 76.6 percent in a postsecondary education. Thirty-four percent of the sample was employed, contrary to 6.6 percent of unemployed respondents. Forty-eight percent of these nonstudents had a diploma of higher education, 44 percent finished high school and 8 percent did not obtain a high school degree. In line with other studies that used an online methodology to recruit players of FPS,^43,44 most respondents were avid players who played games 13.5 hours a week on average (SD=13.7). The vast majority of the sample (81.6 percent) had finished HL2 at least once.

Testing the hypothesized models

The assumptions for univariate and multivariate normality were met.^37,47 The fit indices of the single-, five-, and six-factor model are presented in Table 1.

Table 1.

Fit Indices of the Confirmatory Factor Analyses Testing the One-, Five-, and Six-factor Models

Criteria	Threshold	One-factor model	Five-factor model	Six-factor model
χ ²		2484.73	887.73	611.82
χ²/df	<3.00	7.67	2.83	1.98
CFI	>0.90	0.559	0.883	0.938
TLI	>0.90	0.522	0.869	0.930
RMSEA	<0.06	0.132	0.069	0.051
AIC		2646.73	1069.73	803.82

χ², chi square statistic; χ²/df, chi-squared-to-degrees-of-freedom ration; CFI, comparative fit index; TLI, Tucker–Lewis index of non-normed fit index; RMSEA, root-mean-squared error of approximation; AIC, Akaike's information criterion.

Model comparison tests showed that the six-factor model consistently yielded a more acceptable fit than the one- and five-factor solution. The AIC value was also considerably lower for the six-factor model than for the single- and five-factor model.

Standardized estimates of factor loadings of the six-factor solution are presented in Table 2. All factor loadings were above 0.40 and the range of loadings was between 0.57 and 0.90. The Cronbach alpha estimates of internal reliability of all subscales were satisfactory.

Table 2.

Standardized Estimates of Factor Loadings of the Six-factor Solution

No.	Items	sim. real.	freed.	perc. perv.	soc. real.	auth.	char. Inv.
1	By playing HL2 I can learn how certain problems are resolved in the real world.	0.85
2	By playing HL2 I can learn something about the real world.	0.83
3	By playing HL2 I feel better prepared to handle certain life-like situations.	0.79
4	By playing HL2 I can learn how to control certain situations.	0.77
5	By playing HL2 I can learn how to behave in the real world.	0.74
6	While playing HL2, I feel I'm creating my own story.		0.80
7	While playing HL2, I feel that I determine the course of the game.		0.78
8	In HL2, I feel that I determine the outcome of the battle.		0.71
9	While playing HL2, I feel I can determine which path to follow in the video game world.		0.65
10	While playing HL2, I feel that I'm pulling the strings.		0.63
11	In HL2, everything looks impressive.			0.90
12	The picture quality of HL2 is incredible.			0.84
13	In HL2, everything is depicted in great detail.			0.84
14	HL2 has a poor picture quality (R)			0.66
15	The events in HL2 bear similarities with events in real life.				0.70
16	Characters in HL2 act like persons in real life.				0.67
17	In HL2, aspects from reality are woven into the game.				0.66
18	The characters in HL2 bear similarities with people in real life.				0.64
19	The locations in HL2 look similar to places in real life.				0.64
20	The developers of HL2 spent great care to make sure that they built a credible world.					0.77
21	It's obvious that the developers of HL2 had an eye for detail.					0.73
22	The developers of HL2 spent a lot of time and effort to make HL2 feel natural.					0.62
23	The developers of HL2 examined which objects (e.g., weapons) fit within the context of the game.					0.57
24	While playing HL2, it feels as if I'm present in the video game world.						0.69
25	While playing HL2, the video game world feels real.						0.69
26	I feel that the character in HL2 is an extension of myself.						0.68
27	While playing HL2, I experience the events like I experience events in real life.						0.67
	Chronbach alpha	0.89	0.84	0.88	0.80	0.76	0.78

sim real., simulational realism; freed., freedom; perc. perv., perceptual pervasiveness; soc. real., social realism; auth., authenticity; char. inv., character involvement. Empty cells represent the factor loadings that are fixed to 0. All other factor loadings are significant at p<0.001.

HL2, Half-Life 2.

Post hoc analyses

Although the six-factor model yielded a satisfactory fit, we report the post hoc analyses to help future researchers improve the conceptualization and operationalization of perceived game realism. The indicators for revisions included nonsignificant regression weights, low squared multiple correlations, standardized residuals larger than 2.58, and modification indices.^37,48

All regression weights reached significance (p<0.001), thereby indicating that the latent variables predict a significant portion of all items' variance. Likewise, squared multiple correlations reached acceptable values. However, the examination of the standardized residual covariance matrix revealed 11 residual values for which the covariance was systematically under- or overestimated. Three items accounted for most residuals in excess of 2.58: item 14, item 21, and item 26. After deleting these items stepwise, item 18 still had large residuals (−2.70) with item 7. As the parameter representing the error covariance between item 18 and item 16 had the highest modification index value (24.373) and the highest expected parameter change (0.207), item 18 was deleted. This item can be considered redundant as there is a high degree of overlap in content with item 16. The final model showed a very good fit of the data and yielded a chi-square value of 333.23 with 215 degrees of freedom, p<0.001, χ²/df=1.55, CFI=0.97, TLI=0.97, RMSEA=0.038, and AIC=501.22.

Conclusions

CFA indicated that the dimensional structure of perceived realism in HL2 could best be represented by a six-factor model comprised of simulational realism, freedom, perceptual pervasiveness, social realism, authenticity, and character involvement. Based upon this result, the following conclusions can be drawn. First, the six-factor model of Ribbens and Malliet²⁴ is confirmed, at least for the realism judgments on HL2 of players in this sample. The model foregrounds that, besides the audiovisual characteristics, the rule-based characteristics of an electronic game are constructive of a game's overall reality impression. Second, the findings support the distinction between external and internal realism judgments and, consequently, the idea that the players of HL2 can make an internal realism judgment that is distinct from a judgment that compares the video game universe with a real-life referent. Finally, perceived game realism is a complex multidimensional construct. Researchers should therefore avoid discussing or measuring perceived game realism in general, but try to delineate their understanding of the concept.

Discussion

The perceptions that players have toward virtual worlds have frequently been studied from a philosophical or theoretical point of view, but are also important in explaining the social impact of digital games. This study contributes to the existing body of literature by increasing our understanding of the conceptualization and the measurement of perceived game realism, a concept that has been linked to diverse media outcomes, such as aggression, enjoyment, and enhanced learning. An understanding of the dimensions and an improved measurement of perceived game realism could help researchers single out possible effects of digital game play, and help developers of serious games direct their attention to those aspects of game play that are most influential in producing learning effects.

Despite recent efforts, the development and measurement of perceived game realism are still in need of further validation and refinement. The findings of this study support the trends found in Ribbens and Malliet,²⁴ but the design makes it difficult to generalize the data. This study, for instance, found a distinction between an external and internal realism judgment, but it is unknown whether this result will persist when tested using other games or when realism judgments are measured immediately after exposure. Likewise, although online research methodologies are common practice in digital game research, we must acknowledge the self-selected nature of the sample. Future research should use different data sets gathered from the same and different populations to validate the current model. The author is currently involved in a research project in which 1. the dimensional structure of perceived game realism is further tested using different shooter games and different populations and 2. the psychometric qualities of the scale are further investigated. Additional research should examine whether the framework is robust across game genres. As far as the analysis is concerned, CFA can only indicate which of the proposed models fits the data best, but does not rule out the possibility of other conceptualizations. One might argue for instance, that aspects of emotional realism are underrepresented in the six-factor model.³ Finally, the unidimensional model in this study did not equal the unidimensional conceptualizations used in television research exactly. More specifically, they are alike in that they confuse different aspects of perceived realism, but the dimensions in this study are conceptualized to include a more diverse range of aspects of perceived realism. This may be reflected in the poor model fit of the unidimensional conceptualization. Nevertheless, evidence is growing that perceived realism should be treated as a complex, multidimensional construct.

Footnotes

Acknowledgments

The author would like to thank Prof. dr. Steven Malliet, Damien Larkin and the anonymous reviewers for their valuable comments.

Author Disclosure Statement

The author has no conflict of interest.

Appendix

References

Busselle

. Television exposure, perceived realism, and exemplar accessibility in the social judgment process. Media Psychology, 2001; 3:43–67.

Konijn

, Bijvank

, Bushman

. I wish I were a warrior: the role of wishful identification in the effects of violent video games on aggression in adolescent boys. Developmental Psychology, 2007; 43:1038–1044.

Shapiro

, Peña-Herborn

, Hancock

. 2006. Realism, imagination, and narrative video games. Vorderer

, Bryant

. Playing video games: motives, responses & consequences. Mahwah, NJ: Erlbaum, 275–289.

Krcmar

, Farrar

, McGloin

. The effects of video game realism on attention, retention and aggressive outcomes. Computers in Human Behavior, 2011; 27:432–439.

Potter

. Perceived reality and the cultivation hypothesis. Journal of Broadcasting and Electronic Media, 1986; 30:159–174.

Bilandzic

, Busselle

. Enjoyment of films as a function of narrative experience, perceived realism and transportability. Communications. The European Journal of Communication Research, 2011; 36:29–50.

de Kort

YAW

, IJsselsteijn

. Reality check: the role of realism in stress reduction using media technology. CyberPsychology and Behavior, 2006; 9:230–233.

Meijer

, Geudeke

, van den Broek

. Navigating through virtual environments: visual realism improves spatial cognition. CyberPsychology and Behavior, 2009; 12:517–521.

Andsager

, Austin

, Pinkleton

. Questioning the value of realism: Young adult's processing of messages in alcohol-related public service announcements and advertising. Journal of Communication, 2001; 51:121–142.

10.

Hamam

, Eid

, El Saddik

. Effect of kinesthetic and tactile haptic feedback on the quality of experience of edutainment applications. Multimedia Tools and Applications, 2012[Epub ahead of print]10.1007/s110442-012-0990-7.

11.

Huesmann

, Moise-Titus

, Podolski

et al. Longitudinal relations between children's exposure to TV violence and their aggressive and violent behavior in young adulthood: 1977–1992. Developmental Psychology, 2003; 29:201–221.

12.

Haridakis

. Viewer characteristics, exposure to television violence, and aggression. Media Psychology, 2002; 4:323–353.

13.

Peter

, Valkenburg

. Adolescents' exposure to sexually explicit online material and recreational attitudes toward sex. Journal of Communication, 2006; 56:639–660.

14.

Rivadeneyra

, Ward

. From Ally McBeal to Sábado Gigante: Contributions of television viewing to the gender role attitudes of Latino adolescents. Journal of Adolescent Research, 2005; 20:453–475.

15.

Segrin

, Nabi

. Does television viewing cultivate unrealistic expectations about marriage? Journal of Communication, 2002; 52:247–263.

16.

Hall

. Reading realism: audiences' evaluations of the reality of media texts. Journal of Communication, 2003; 53:624–641.

17.

Hawkins

. The dimensional structure of children's perceptions of television reality. Communication Research, 1977; 4:299–320.

18.

Popova

. 2010. Perceived reality of media messages: Concept explication and testing. Conference paper: Annual Meeting of the International Communication Association, Singapore, 2010.

19.

Busselle

, Greenberg

. The nature of television realism judgments: a reevaluation of their conceptualization and measurement. Mass Communication and Society, 2000; 3:249–268.

20.

Potter

. Perceived reality in television effects research. Journal of Broadcasting and Electronic Media, 1988; 32:23–41.

21.

Shapiro

, Chock

. Psychological processes in perceiving reality. Media Psychology, 2003; 5:163–198.

22.

Galloway

. Social realism in gaming. Game Studies 2004. 4. www.gamestudies.org/0401/galloway/. 2012 Feb. 27.

23.

Malliet

. An exploration of adolescents' perceptions of videogame realism. Learning, Media and Technology, 2006; 31:377–394.

24.

Ribbens

, Malliet

. Perceived digital game realism: a quantitative exploration of its structure. Presence: Teleoperators and Virtual Environments, 2010; 19:585–600.

25.

Valve Corporation. Half-Life 2 [Game] Washington: Valve Corporation, 2004.

26.

Witmer

, Singer

. Measuring presence in virtual environments: a presence questionnaire. Presence, 1998; 7:225–240.

27.

Brockmyer

, Fox

, Curtiss

et al. The development of the game engagement questionnaire: a measure of engagement in video game-playing. Journal of Experimental Social Psychology, 2009; 45:624–364.

28.

Bell

. Realism and subjectivity in First-Person Shooter videogames. Georgetown Peer-reviewed Journal of Communication, Culture and Technology. 2003; 3. www.gnovisjournal.org/2003/04/16/realism-subjectivity-in-first-person-shooter-games/. 2012 Jan. 14.

29.

Juul

. 2005. Half real: video games between real rules and fictional worlds. Cambridge, MA: MIT Press.

30.

Eskelinen

. The gaming situation. Game Studies, 2001; 1. www.gamestudies.org/0101/eskelinen/. 2012 Dec. 28.

31.

Frasca

. 2003. Simulation versus narrative: introduction to ludology. Wolf

MJP

, Perron

. The video game theory reader. London: Routledge, 151–158.

32.

Jansz

. The emotional appeal of violent video games. Communication Theory, 2005; 15:219–241.

33.

Kücklich

. Perspectives of computer game philology. Game Studies, 2003; 3. www.gamestudies.org/0301/kucklich/. 2012 Feb. 27

34.

Lee

. Presence, explicated. Communication Theory, 2004; 14:27–50.

35.

Calleja

. Digital game involvement. Games and Culture, 2007; 2:236–260.

36.

Busselle

, Bilandzic

. Fictionality and perceived realism in experiencing stories: a model of narrative comprehension and engagement. Communication Theory, 2008; 18:255–280.

37.

Byrne

. 2010. Structural equation modeling with AMOS: basic concepts, applications, and programming. New York: Routledge.

38.

Nintendo. Super Mario Galaxy [Game] Kyoto: Nintendo, 2007.

39.

Rockstar North. Grand Theft Auto IV [Game] New York: Rockstar Games, 2008.

40.

Lang

. The limited capacity model of mediated message processing. Journal of Communication, 2000; 50:46–70.

41.

Wood

RTA

, Griffiths

, Eatough

. Online data collection from video game players: Methodological issues. CyberPsychology & Behavior, 2004; 7:511–518.

42.

Jansz

, Tanis

. Appeal of playing online First Person Shooter games. CyberPsychology and Behavior, 2007; 10:133–136.

43.

Ghuman

, Griffiths

. A cross-genre study of online gaming: Player demographics, motivation for play, and social interactions among players. International Journal of Cyber Behavior, Psychology and Learning, 2012; 2:13–29.

44.

Bollen

. 1989. Structural equations with latent variables. New York: Wiley.

45.

, Bentner

. Cutoff criteria for fit indices in covariance structure analysis: conventional criteria versus new alternatives. Structural Equation Modeling, 1999; 6:1–55.

46.

Schreiber

, Nora

, Stage

et al. Reporting structural equation modeling and confirmatory factor analysis: A review. The Journal of Educational Research, 2006; 99:323–337.

47.

Muthén

, Kaplan

. A comparison of some methodologies for the factor analysis of non-normal Likert variables: A note on the size of the model. British Journal of Mathematical and Statistical Psychology, 1992; 45:19–30.

48.

Jöreskog

, Sörbom

. 1988. LISREL 7: A guide to the program and applications. Chicago: SPSS.