The observation that only the SCG of slots exhibited a significant association with current tobacco use requires further investigation given that it is presently unclear why this particular SCG alone is linked to adolescent smoking status. Given the availability of empirical evidence suggesting that both slot-machine gambling [88, 89] and tobacco use  are highly addictive activities, it may be the case that individuals with a tendency to engage in addictive behaviours may seek them out. This explanation is particularly plausible given the considerable structural similarities that exist between the SCG of slots and slot-machine games intended HappyLuke for monetary gambling . Additional research is needed, however, to better understand these findings.
Although existing research pertaining to online gambling suggests that poorer academic performance may be associated with participation in gambling among adolescent samples [15, 43, 44], these effects were not observed in the present investigation across all three SCG types. Rather, social casino gamers did not differ significantly from non-gamers in terms of typical grade achieved in school. Consequently, while achievement in school may be relevant to monetary gambling, it does not appear to be an key factor in defining the profiles of social casino gamers across SCG types.
Although every effort was made to design a study that would yield valid findings, the present investigation is not without its limitations. A central limitation of the study is its cross-sectional design. While this design allowed us to obtain large and representative data that inform the various factors associated with social casino gaming, it prohibits us from stating conclusively that the factors examined are causally predictive of SCG use. The results, however, do demonstrate key links between participation in SCGs and personal and environmental indicators relevant to adolescent populations that have largely been unexamined by previous research endeavours. As a result, the present study offers a foundation for subsequent longitudinal analyses examining predictors of adolescent SCG use.
A further limitation of the present investigation is its reliance on self-report data, which may have been affected by response bias. Specifically, participants may have under-reported specific types of behaviours, especially those deemed socially undesirable , although it has been shown that this type of self-management is less problematic for non-pathological gambling behaviours . In an effort to offset potential issues pertinent to self-report assessment in the present study, and to augment honest responding, we collected data through anonymous paper-based questionnaires, and we instructed survey administrators to refrain from moving through the survey location when the measures were being completed by respondents. Although steps were taken to maximize honest responding, the self-report nature of the measures used in the present study did not allow us to collect information about the gambling habits of close others directly from participants’ peers and parents. Future investigations may wish to include other-report measures into their designs in order to obtain primary information from adolescents’ friends and family regarding their own gambling and/or SCG play.