Online video gaming has grown into one of the most popular leisure activities of our time. In recent years, this has led to greater use of electronic gadgets such as computers, cell phones, and gaming consoles, as well as more time spent online, especially by adolescents and young adults [1]. Having fun, seeking challenges and excitement, and dealing with emotions or escaping from reality have been mentioned as the main reasons for playing online games [2].

Around the globe, the number of online players has risen from 700 million in 2013 [3] to 2.2 billion in 2017 [4], and according to the Global Online Gaming Market report, the Asia-Pacific region has the highest number of online players [5]. In the Middle East as well there is a large and growing market, with 20 million people in Türkiye regularly playing online games, as an example [6].

Recently, excessive gaming has been noted as a public health concern in Iran [7,8]. The Computer and Video Games Foundation of the country has reported that 28 million Iranians have played video games in 2018, showing a 6-fold increase compared to the 2013 survey. Of these gamers, 48% reported playing online [9]. Considering the global and the national increase in the prevalence of online gaming (especially among adolescents and young adults), there is a need for better understanding and formulating different aspects of this phenomenon and its potential correlates. In this context, gaming motivations seem to be poorly understood and under-investigated.

Games are played with different motivations by different groups and subgroups [2]. For example, men seem likelier to play games to feel successful or to make friends, whereas women may play to pass time.10 A validated taxonomy of motives for gaming should be developed, and a sound tool for measuring them is required. This tool could help establish associations between motivations and the behavioral outcomes [11].

Various theories of motivation have been used to conceptualize online gaming behavior, with each theory forming the basis of a different instrument such as the Player Experience of Need Satisfaction (PENS) [12] and Motives for Online Gaming Questionnaires (MOGQ) [2]. The Gaming Motivations Scale (GAMS) was constructed within the theoretical framework of Self-Determination Theory (SDT) [13], which focuses on player motives and personality within social contexts and distinguishes between motives in terms of an individual’s level of perceived autonomy and control [14]. SDT is founded on the premise of individuals possessing fundamental psychological needs for competence, autonomy, and relating to others, whose satisfaction facilitates autonomous motivation whereas their nonsatisfaction leads to a sense of pressure to behave in particular ways (controlled motivation) or lack of motivation [15].

Within SDT, behaviors depend on six types of motivations; namely intrinsic motivation, integrated regulation, identified regulation, introjected regulation, external regulation, and amotivation. The GAMS was developed and validated to measure the extent to which participants’ gaming behavior relies on each of the SDT’s six motivation domains. These domains range from internal motivation, where a person plays the game to improve his skills or enjoy exploring the game environment, to external motivation which involves the game to achieve specific, instrumental goals, such as scoring points [13].

However, the GAMS has not been extensively evaluated beyond the original study. Also, the GAMS was not originally evaluated for online and offline games separately. The distinction is important for investigation of gaming motivations, because there might be a fundamental difference between offline and online games [16] and there are few other tools assessing online gaming motives specifically [17].

For this purpose, this study aimed to translate the GAMS into Persian, and then evaluate its psychometric properties for the assessment of motivations of online gaming in an Iranian population.

GAMS is an 18-item questionnaire (GAMS-18), designed to identify individuals’ motives for playing video games, measuring six domains of motivation using three questions for each domain with a 7-point Likert scale. This six domains are namely intrinsic motivation (questions 1, 7, and 13), integrated regulation (questions 2, 8, and 14), identified regulation (questions 3, 9, and 15), introjected regulation (questions 4, 10, and 16), external regulation (questions 5, 11, and 17), and amotivation (questions 6, 12, and 18).

The study methodology consisted of the following three stages: 1) translation and determination of content validity; 2) determination of the external and internal consistency reliability; and 3) determination of the structural validity.

First, the questionnaire was translated using the forward-backward method based on the World Health Organization instructions [18].

After translating the questionnaire, instructions were modified to assess online gaming rather than any type of video gaming. Then, the opinions of three groups were used to assess content validity. These were 1) seven domain experts with relevant expertise in the field, 2) seven lay experts, and 3) a psychometrist. Three aspects of content validity, namely relevance, clarity, and comprehensiveness were assessed. The relevance and clarity indices were calculated on two levels (for individual items as well as the entire instrument) using the Universal Agreement (UA) approach and the mean approach. The minimum acceptable value for the content validity index (CVI) for both individual items and the entire questionnaire was considered to be 0.80. Also, the content validity ratio (CVR) was assessed based on Ayre and Scally’s revised method [19].

The reliability of the questionnaire was investigated by the self-administration of GAMS to 70 students from the Azad University of Tehran and retesting two weeks after the initial assessment. Intraclass correlation coefficient (ICC) was used to assess the test-retest reliability. The internal consistency reliability of the questionnaire was calculated using Cronbach’s alpha.

To assess validity, data from an online survey with several questionnaires about problematic gaming and gaming motivation were used [20,21]. A total of 791 individuals with a history of online gaming in the prior year had completed this questionnaire. To assess the convergent validity, we used the PENS instrument. Confirmatory factor analysis of GAMS was performed using AMOS software (version 24; IBM SPSS., Chicago, IL, USA) and as a confirmatory step, chi-square statistic and model fit indices were assessed also we used of modification indices (MI) for model fitting. Online gaming disorder were diagnosed using the Ten-Item Internet Gaming Disorder Test (IGDT-10). The IGDT-10 has been developed to operationalize the nine proposed IGD criteria in Diagnostic and Statistical Manual of Mental Disorders, 5th edition. According to the IGDT-10, five or more out of the nine criteria is required for the screening of internet gaming disorder [22]. The validity and reliability of the IGDT-10 has been confirmed in multiple different languages, such as English and Farsi [21].

We used the Pearson correlation coefficient to assess linear correlations between GAMS factors and with PENS factors. Mean score differences of GAMS factors between sex, age group, and online gaming disorder was assessed with independent t-test and one-way analysis of variance tests. Multiple comparisons of mean scores were conducted based on Bonferroni post-hoc tests. A p<0.05 conducted as statistically significant differences.

All analyses were performed using IBM SPSS Statistics for Windows (version 25; IBM Corp., Armonk, NY, USA) and AMOS (version 24; IBM SPSS). The research protocol was approved by the Ethics Committee of Tehran University of Medical Sciences in Iran (no. IR.TUMS.VCR.REC.1395.800).

Previous research on gaming and its potential consequences has seldom investigated motivation as an important aspect of online video gaming [2]. A few instruments have been developed in recent years, such as the MOGQ, which has also been translated and validated in Iran [23]. GAMS is believed to be a comprehensive instrument which can appropriately assess different aspects of an individual’s motivations based on the SDT model. The validity and reliability of the original English version of the questionnaire has been shown to be acceptable [13]. Based on the SDT model, the questionnaire comprises six motivation domains, measuring different aspects of motivation for gaming along the internal-external spectrum. The goal of this study was to assess the psychometric properties of the Persian version of the GAMS.

In this study, the Persian version of the GAMS showed very good comprehensiveness and good content validity. To the best of our knowledge, the test-retest reliability of the GAMS has not been assessed before. Our study showed a good test-retest reliability of the domains and the entire questionnaire, such that the entire questionnaire had a test-retest reliability of 0.89 and the domain with the lowest reliability, intrinsic motivation, had a reliability of 0.76. The results also show a good internal consistency reliability of 0.90, with the domain with the lowest internal consistency reliability again being intrinsic motivation.

Correlations between the various factors of the questionnaire varied widely in strength, from 0.347 (between amotivation and external regulation) to 0.830 (between internal consistency and integrated regulation), with the generally high correlations between internal consistency and other factors suggesting a conceptual similarity between this factor and all others. The factor with the lowest correlation with the rest was amotivation, which is to be expected given its conceptual uniqueness in GAMS; while other domains assess the degree of various kinds of motivation, this domain reflects lack of motivation. Notably, GAMS domains generally showed low correlations with PENS. The highest correlation between a GAMS and PENS questionnaire domains the highest correlation was 0.698 between intrinsic motivation and autonomy. The amotivation factor in GAMS had the lowest correlation with PENS factors generally. This suggests that the GAMS conceptualizes and measures gaming motivations differently than PENS and MOGQ [23].

In the assessment of construct validity, the first item which was in the intrinsic motivation domain had the lowest factor loading at 0.56, whereas the seventh item (For the pleasure of trying/experiencing new game options), also in the intrinsic motivation domain, had the lowest factor loading in the original questionnaire at 0.57. Lafrenière et al. [13] also found a low factor loading value (0.58) for question 11 (for the prestige of being a good player). Peracchia et al. [24] found item 18 (Honestly, I don’t know; I have the impression that I’m wasting my time), in the amotivation domain, to have the lowest loading. Questions 8 in the integrated regulation domain and 17 in the external regulation domain had the highest factor loadings (0.82), while in the original questionnaire question 10 (Because I must play to feel good about myself) in the introjected regulation domain had the highest loading (0.96) [13] and Peracchia et al. [24] found question 17 to have the highest factor loading (0.93) in the external regulation domain. These differences may reflect genuine cultural divergences between studied populations: for example, another recent study suggests that fantasy and escape motives may be the most important motivations for pathological gaming among Iranians distinctly, though further research is needed to investigate this hypothesis since Iranian gamers may have differences in terms of preferred gaming platforms and genres and their age and sex profile, and Iranian men and women have different playing patterns as well [20]. Based on the results of MI tests, we modified the original model with the addition of error paths between items 1 and 7 items 3 and 9, which improved the fit indices of the model. Lafrenière et al. [13] reported that the addition of an error covariance path between introjected regulation and amotivation domains could improve the model, but justified it with the hypothesis that the domains are related in a linear sequence. After the modification, our model had a comparable performance based on fit indices to other studies using GAMS.

As Lafrenière et al. [13] suggest should be done, correlations between GAMS and two major online gaming motivation questionnaires, the 3-domain PENS, were also investigated in this study, showing low correlations overall. The highest correlations were observed between the autonomy and competence domains of PENS and the intrinsic motivation GAMS domain, and the lowest was between the competence PENS domain and the amotivation domain of GAMS. The low correlations suggest important conceptual and psychometric mismatches between GAMS and the other questionnaires.

No GAMS domain score significantly differed between men and women, but there was an age difference with those in the 18-20 years age group having the highest average scores across all domains compared to the 21-23 years and the 24-50 years age group. Also, while the average motivation scores in all domains were higher among individuals with gaming disorder, the difference in motivation scores between those with and without gaming disorder was not significant. This is in contrast to another study, in which heavy gamers had significantly different motivation scores than light gamers in all GAMS domains except for the amotivation domain [24]. The unexpected nonsignificance of differences between those with and without gaming disorder might be due to the small sample size, especially in the gaming disorder group. Further research is required to investigate the role of GAMS gaming motivation domains in problematic and disordered gaming.

Lastly, although the GAMS is a short questionnaire and is easy to comprehend and administer, it also has some limitations. The self-report and subjective nature of the questionnaire is a limitation and as Lafrenière et al. [13] suggested, future studies should involve other sources of information, like family members and friends, and include objective measures.