Video game playing (“gaming”) has long been a popular activity, especially among children and adolescents (Ivory, Reference Ivory2015). According to a Canadian survey, 88–95% of children aged 6 and 12 years play video games, with girls playing 9 h and boys 12 h per week on average (Entertainment Software Association of Canada, 2020). Another survey conducted in the United States indicates that the average daily gaming time for children aged 4–8 years is 40 min, and this duration increases with age (Rideout & Robb, Reference Rideout and Robb2020). However, high levels of gaming among children have prompted concerns because video games are intentionally designed to be as engaging and rewarding as possible (Hodent, Reference Hodent2017), with some even incorporating gamblification features (Király et al., Reference Király, Zhang, Demetrovics and Browne2022). As an example, a growing trend in video game design, introduced in 2003, known as “loot boxes,” significantly enhances player engagement and spending by employing randomized reward mechanisms akin to gambling (Lemmens, Reference Lemmens2022; Spicer et al., Reference Spicer, Nicklin, Uther, Lloyd, Lloyd and Close2022; Zendle et al., Reference Zendle, Meyer, Cairns, Waters and Ballou2020). This may then pose a significant risk for the developing reward systems of young players (Spicer et al., Reference Spicer, Nicklin, Uther, Lloyd, Lloyd and Close2022; Zendle et al., Reference Zendle, Meyer and Over2019).
According to studies conducted with adults, the intensity of striatal dopamine released during gaming is comparable to the effect of substances such as alcohol and stimulants (Boileau et al., Reference Boileau, Assaad, Pihl, Benkelfat, Leyton, Diksic and Dagher2003, Reference Boileau, Dagher, Leyton, Welfeld, Booij, Diksic and Benkelfat2007; Cox et al., Reference Cox, Benkelfat, Dagher, Delaney, Durand, McKenzie and Leyton2009; Koepp et al., Reference Koepp, Gunn, Lawrence, Cunningham, Dagher, Brooks and Grasby1998). Consistent with the activation of reward-related pathways, high engagement in gaming can lead players to experience cravings for gaming (Antons et al., Reference Antons, Liebherr, Brand and Brandtner2023). Supporting this, there is also evidence that gaming is associated with neuroplastic changes in brain regions involved in reward processing (Kühn et al., Reference Kühn, Romanowski, Schilling, Lorenz, Mörsen, Seiferth and Gallinat2011; Palaus et al., Reference Palaus, Muñoz-Marrón, Viejo-Sobera and Redolar-Ripoll2017). Moreover, in addition to external rewards, it has been proposed that gaming can provide a sense of competence, autonomy, and social connectedness, which are key elements of intrinsic motivation (Przybylski et al., Reference Przybylski, Rigby and Ryan2010). Coupled with compelling mechanisms designed to maintain maximum engagement (e.g. Király et al., Reference Király, Zhang, Demetrovics and Browne2022), these elements make gaming highly appealing and a difficult habit to regulate, especially for children, whose developing self-regulatory abilities may heighten their vulnerability to its allure and addictive qualities.
Given its high level of stimulation and engagement, more frequent gaming by children may lead to decreased enjoyment and motivation for other, more mundane activities (Christodoulou et al., Reference Christodoulou, Majmundar, Chou and Pentz2020; Klimczyk, Reference Klimczyk2023), such as academic learning. According to a large-scale cross-sectional study of adolescents, gaming for more than 2 h/day was significantly associated with reduced time devoted to homework and a higher incidence of school absenteeism (Hellström et al., Reference Hellström, Nilsson, Leppert and Åslund2012). Similarly, another large-scale cross-sectional study found that children and adolescents who game spent less time reading and studying compared to non-gamers (Cummings & Vandewater, Reference Cummings and Vandewater2007). This study estimated that for each hour boys spent gaming on weekdays, they spent 2 min less reading (30% decrease), and every hour girls played video games during weekdays was associated with 13 min less doing homework (34% decrease). Yet, despite evidence associating gaming with lower academic engagement, the possible underlying mechanisms – such as diminished academic motivation – remain insufficiently investigated.
The school-age years represent a crucial time for children to cement key academic skills that will pave the way for their educational attainment and healthy development (Allison, et al., Reference Allison and Attisha2019; Peng & Kievit, Reference Peng and Kievit2020; Williams et al., Reference Williams, Clarkson, Hastings, Watkins, McTague and Hughes2022). Thus, it remains crucial to investigate how gaming, a common leisure activity among children, may impact academic motivation at this age. Understanding the directionality of such an association is equally crucial, as past research on the associations between gaming and educational outcomes has been predominantly cross-sectional (Adelantado-Renau et al., Reference Adelantado-Renau, Moliner-Urdiales, Cavero-Redondo, Beltran-Valls, Martínez-Vizcaíno and Álvarez-Bueno2019). Overcoming this limitation is relevant because video games may either influence academic motivation or act as an alternative source of fulfillment for students who experience low levels of enjoyment at school. Indeed, past correlational research has shown that higher levels of obsessive gaming were correlated with lower levels of overall life satisfaction and basic psychological need satisfaction (Przybylski et al., Reference Przybylski, Weinstein, Ryan and Rigby2009).
In light of the limitations in the literature, in the current study, we test bidirectional associations between child gaming and academic motivation at ages 7, 8, and 10 years using a random-intercept cross-lagged panel model (RI-CLPM) (Hamaker et al., Reference Hamaker, Kuiper and Grasman2015). An important strength of this statistical model is that it models within-person changes. In other words, each person serves as their own baseline control, which accounts for stable, existing individual differences in social, psychological, and academic experiences (e.g. socioeconomic differences). Since gaming and academic motivation differ between boys and girls (Hastings et al., Reference Hastings, Karas, Winsler, Way, Madigan and Tyler2009; Leonhardt & Overå, Reference Leonhardt and Overå2021; Niskier et al., Reference Niskier, Snaychuk, Kim, da Silva, de Souza Vitalle and Tavares2024), we estimate associations separately for boys and girls.
Methods
Participants and setting
Participants were part of the Quebec Longitudinal Study of Child Development (QLSCD, 1998–2023), a population-based cohort intended to study child biopsychosocial development and academic adjustment. For the QLSCD, birth registers were used to create a stratified random sample of families with infants born between 1997 and 1998 in the province of Quebec (Canada). Inclusion criteria included pregnancy lasting between 24 and 42 weeks and mothers being proficient in either French or English. In total, 2,120 participating families were followed longitudinally and subsequently assessed yearly or biennially. The current study includes data from the QLSCD collected in 2005, 2006, and 2008, when children were at ages 7 (n = 1,537), 8 (n = 1,526), and 10 (n = 1,402) years. Participating children with information on gaming and academic motivation, at least at one time point, were retained in the current study. Parents of participating children provided informed consent, and children also provided assent at age 10 years. The Health Quebec ethics committee approved the QLSCD protocol. Additional details on the QLSCD are described elsewhere (Orri et al., Reference Orri, Boivin, Chen, Ahun, Geoffroy, Ouellet-Morin and Côté2021).
Measures
Gaming (video game playing time)
At ages 7, 8, and 10, parents reported child gaming responding to the following question: “On average, how much time does your child spend each day playing computer or video games?.” Responses included: “None,” “Less than one hour,” “From 1 up to 3,” “From 3 up to 5,” “From 5 up to 7,” “More than 7 hours.” These categories were converted as follows: 0, 0.5, 2, 4, 6, and 8 h/day. This question was developed for the QLSCD and can be accessed at https://www.jesuisjeserai.stat.gouv.qc.ca/informations_chercheurs/outils_collecte/outils_collecte_an.html.
Academic motivation
At ages 7, 8, and 10, children were asked by a trained interviewer about their interest in reading, writing, and mathematics with the following questions – 1: “I like reading/writing/math,” 2: “Reading/writing/math interest me a lot,” 3: “I read/write/do math even when I am not obliged to do so.” Their responses were reported on a 5-point Likert scale ranging from “ Always no ” to “ Always yes.” Child responses to these nine questions were then used to estimate a standardized global intrinsic motivation score ranging from 0 to 10. Our measure was derived from the intrinsic motivation subscale of the Elementary School Motivation Scale, which is based on the self-determination theory of motivation (Guay et al., Reference Guay, Chanal, Ratelle, Marsh, Larose and Boivin2010; Ryan, Reference Ryan2023). Internal reliability was adequate, α = .72–.80. Furthermore, this measure has been previously validated and is a robust predictor of academic engagement and achievement (Garon-Carrier et al., Reference Garon-Carrier, Boivin, Guay, Kovas, Dionne, Lemelin and Tremblay2016).
Statistical analysis
To estimate the direction of associations between gaming and academic motivation in school-aged children, we use a RI-CLPM (Hamaker et al., Reference Hamaker, Kuiper and Grasman2015). The RI-CLPM is a structural equation model used to determine the temporal sequencing of associations between two variables that are measured repeatedly across time. The RI-CLPM has the advantage of separating stable between-person effects (e.g. individual differences in academic motivation due to stable factors and socioeconomic status) from dynamic within-person effects (e.g. how children’s deviations from their own baseline of video game consumption in 1 year affect their academic motivation 1 year later relative to their unique baseline). In a RI-CLPM, the cross-lagged and autoregressive paths measure only within-person fluctuations. More information on the RI-CLPM is available elsewhere (Baribeau et al., Reference Baribeau, Vigod, Ma, Vaillancourt, Szatmari and Pullenayegum2022; Hamaker et al., Reference Hamaker, Kuiper and Grasman2015; Mulder & Hamaker, Reference Mulder and Hamaker2021). Recommended cutoff points to interpret the effect sizes of cross-lagged paths are <.07 (small), .07–.11 (medium), and ≥.12 (large) (Orth et al., Reference Orth, Meier, Bühler, Dapp, Krauss, Messerli and Robins2022).
Before specifying our model, based on our a priori hypothesis that the results would be different between boys and girls, we first checked if the variances and covariances of gaming and academic motivation variables were sex-invariant. We compared the model fit of a matrix freely estimated with a matrix constrained for equality across sexes, using multigroup analysis. A significant χ 2 implies that the hypothesis of equal variance/covariance structures for boys and girls can be rejected, which would provide additional empirical support to stratify our model by the sex of the children. Subsequently, we compared the fit of a model constraining cross-lagged regressions for equality over time with the fit of a freely estimated model (Mulder & Hamaker, Reference Mulder and Hamaker2021). The model was defined after testing the invariance between sex and cross-lagged paths over time. In terms of assumptions, our sample size was sufficient (n = 1,631 and n > 20 per observable variable), and we observed no extreme collinearities (all correlations r < .80). However, child gaming variables were not normally distributed (skewness > |2| and kurtosis > |7|). For this reason, we estimated our model using the Huber–White robust variant of the maximum likelihood (ML) estimator, which accommodates non-normal distributions. We estimated the final model by including repeated measures of gaming time and academic motivation at ages 7, 8, and 10 years. See Figure 1 for a representation of our model. The R code for the model specification is provided in the Supplementary Materials. The full model formula for the RI-CLPM is specified elsewhere (Hamaker et al., Reference Hamaker, Kuiper and Grasman2015).
Figure 1. Random-intercept cross-lagged panel model of academic motivation and video game playing between ages 7 and 10 . Each shape represents a variable. Circles are latent variables, and rectangles are observable variables. Straight arrows represent regressions, and curved arrows represent covariances. Asterisks indicate significant associations (p < .05). Indicated in the picture are the standardized estimates of the cross-lagged within-person effects and the between-person associations. Factor loadings of random intercepts were constrained to 1.00. Mot, academic motivation; RI, random-intercept latent variable; Observed, observed variables at data collection; VG, video game playing levels; Y, age in years. Data compiled from the final master file of the Québec Longitudinal Study of Child Development (1998–2023), ©Gouvernement du Québec, Institut de la statistique du Québec, Canada.
Missing data
Missing data from ages 7 to 10 years ranged from 19.3 to 22.6% for gaming and 9.5 to 19.2% for academic motivation. Across all time points, the proportion of full respondents (participants with data at all time points) was 62.4% for gaming and 73.6% for motivation, the proportion of partial respondents was 32.9% for gaming and 24.2% for motivation, and the proportion of nonrespondents was 4.7% for gaming and 2.3% for motivation. Little’s test revealed that the data are not missing completely at random (χ 2 = 277, df = 148, p < .001). Given the characteristics of the missing data, and in line with Newman (Reference Newman2014), we opted to use full information ML to address missing data.
Results
Descriptive statistics
Of the 1,631 children, 48.1% (n = 785) were boys and 51.9% (n = 846) were girls. Data on gaming time or academic motivation were available for 1,514 children (boys = 721) at age 7 years, 1,509 children (boys = 719) at age 8 years, and 1,392 children (boys = 670) at age 10 years. At all time points, boys spent significantly more time playing video games than girls (mean difference = .413 h/day, 95% confidence interval [CI] = .347 to .479), and girls had significantly higher academic motivation compared to boys (mean difference = .398, 95% CI = .242 to .553). Gaming increased significantly from ages 7 to 10 years (mean difference = .214 h/day, 95% CI = .071 to .250) for both sexes. Conversely, academic motivation significantly decreased from age 7 to 10 years (mean difference = −.287, 95% CI = −.414 to −.159) for boys and girls. Descriptive statistics are summarized in Table 1.
Table 1. Descriptive statistics for child gaming and academic motivation
Note: Descriptive statistics of hours per day of video game use and academic motivation of boys and girls between ages 7 and 8 years. IQR, interquartile range; Mot, academic motivation in writing, reading, and mathematics ranging from 0 to 10; SD, standard deviation; VG, video game playing in hours/day. Welch’s t-test was used to test differences between boys’ and girls’ means. Data compiled from the final master file of the Québec Longitudinal Study of Child Development (1998–2023), ©Gouvernement du Québec, Institut de la statistique du Québec, Canada.
Invariance testing
The fit comparison of the freely estimated and the constrained variance/covariance matrix of boys and girls revealed a significant difference between models (Δ χ 2 = 213.46, Δdf = 15, p < .001), suggesting that relaxing constraints between boys and girls significantly improves the model. This result corroborates our decision to stratify our analysis by sex. We then compared a time-invariant model with a freely estimated model over time, which yielded no significant differences (Δ χ 2 = 6.3, Δdf = 8, p = .61), suggesting that the lagged effects are time-invariant (Mulder & Hamaker, Reference Mulder and Hamaker2021). Therefore, we decided to constrain the lagged effects over time in our model.
Random-intercept cross-lagged panel model
Our RI-CLPM generated excellent global fit indices (root mean square error of approximation [RMSEA] = 0.000, standardized root mean square residual [SRMR] = 0.018, Robust comparative fit index [CFI] = 1.000, and χ 2 = 7.712, df = 10, p = .657) and local fit indices (all standardized residuals ≤ |1.96|; see Supplementary Materials). A total of 52 free parameters were estimated in the model, yielding a ratio of 5.2 relative to the degrees of freedom. Our model is represented in Figure 1 along with the results of cross-lagged effects. The full results of our model are described in Table 2, including the autoregressive effects, cross-lagged effects, and between-person associations. Most relevant to our hypothesis are the cross-lagged effects. For boys, within-person level increases in gaming predicted lower levels of academic motivation from ages 7 to 8 (β = −.117, 95% CI = −.226 to −.008), and from ages 8 to 10 (β = −.105, 95% CI = −.197 to −.014). This suggests a medium to large effect of gaming on academic motivation for boys between the ages of 7 and 10 (Orth et al., Reference Orth, Meier, Bühler, Dapp, Krauss, Messerli and Robins2022). Conversely, academic motivation did not predict boys’ later levels of gaming from ages 7 to 8 years (β = −.074, 95% CI = −.192 to .044), nor from age 8 to 10 (β = −.075, 95% CI = −.190 to .041).
Table 2. RI-CLPM longitudinal relationship between child gaming and academic motivation for boys and girls separately
Note: Results from the random-intercept cross-lagged panel model (RI-CLPM) divided by sex, including cross-lagged paths, autoregressive paths, and between-person associations, respectively. B represents the unstandardized estimate of each regression path. β represents the standardized estimate of the regression, followed by the 95% confidence interval of β. P-values < 0.05 were considered significant (marked with *). Mot, academic motivation in writing, reading, and mathematics; VG, video game playing. Data compiled from the final master file of the Québec Longitudinal Study of Child Development (1998–2023), ©Gouvernement du Québec, Institut de la statistique du Québec, Canada.
For girls, no cross-lagged effects were significant. Video game playing did not predict subsequent academic motivation from ages 7 to 8 years (β = .012, 95% CI = −.091 to .115), and from ages 8 to 10 years (β = .009, 95% CI = −.069 to .087). Similarly, academic motivation did not predict gaming from ages 7 to 8 years (β = .017, 95% CI = −.096 to .130) and from ages 8 to 10 years (β = .015, 95% CI = −.082 to .111). For both boys and girls, we found no significant between-person associations between gaming and academic motivation.
Sensitivity analyses
We also estimated a RI-CLPM using complete cases only to assess the potential bias of our missing data approach. This model generated excellent fit indices as well (RMSEA = 0.000, SRMR = 0.019, Robust CFI = 1.000, and χ 2 = 8.124, df = 10, p = .617). A similar pattern of results was observed, suggesting little to no bias. There were no significant within-person cross-lagged effects for girls. For boys, within-person increases in gaming predicted lower levels of academic motivation from ages 7 to 8 years (β = −.131, 95% CI = −.231 to −.030), and from ages 8 to 10 years (β = −.103, 95% CI = −.179 to −.028). Cross-lagged paths from motivation to subsequent gaming were all nonsignificant. Full results of the model using complete cases are available in the Supplementary Materials.
One limitation of the RI-CLPM is that it does not provide information on between-person effects longitudinally. To assess how between-person effects affect our model longitudinally, we performed a traditional cross-lagged panel model (CLPM) to compare results with our RI-CLPM. The CLPM presented good fit indices (RMSEA = 0.05, SRMR = 0.041, Robust CFI = .929, and χ 2 = 48.5, df = 16, p = .000). In terms of our main hypothesis, the results lead to the same conclusions: gaming predicted lower levels of academic motivation from ages 7 to 8 years (β = −.074, 95% CI = −.138 to −.009), and from ages 8 to 10 years (β = −.067, 95% CI = −.125 to −.010) among boys but not girls. Full results of the CLPM are available in the Supplementary Materials. Because CLPMs are nested in RI-CLPMs, we also performed a difference χ 2-test statistic to compare the fit of the two models. The comparison revealed a significantly better fit in the RI-CLPM (ΔAkaike Information Criterion = 36, ΔBayesian Information Criterion = 3, Δ χ 2 = 41.2, Δdf = 6, p = .000).
Finally, we calculated intraclass correlation (ICC) for both gaming and academic motivation to disentangle the proportion of variance explained by between-person differences from within-person fluctuations. ICC values can be interpreted as the proportion of variance that is explained by between-person rather than within-person variation. Gaming ICC3,1 was 0.307 (95% CI = .276 to .338) and motivation ICC3,1 was 0.333 (95% CI = .302 to .364), both indicating high longitudinal within-person variability, which accounts for more than 60% of the variance, and further suggesting that a within-person change model was warranted.
Discussion
To our knowledge, the present study is the first to provide longitudinal evidence that higher levels of child gaming are associated with reduced academic motivation among school-aged boys. The majority of previous studies on the effects of gaming on academic outcomes have used cross-sectional designs or have been unable to account for the direction of association (Adelantado-Renau et al., Reference Adelantado-Renau, Moliner-Urdiales, Cavero-Redondo, Beltran-Valls, Martínez-Vizcaíno and Álvarez-Bueno2019; Mundy et al., Reference Mundy, Canterford, Hoq, Olds, Moreno-Betancur, Sawyer and Patton2020; Sanders et al., Reference Sanders, Parker, del Pozo-Cruz, Noetel and Lonsdale2019; Sharif et al., Reference Sharif, Wills and Sargent2010). Our results help clarify the directionality and temporality of these associations by showing that higher levels of gaming in boys predict lower levels of academic motivation, and not vice versa. Our observed effect sizes ranged from medium to large, which indicates meaningful or clinically significant effects of gaming on boys’ academic motivation during middle childhood (Orth et al., Reference Orth, Meier, Bühler, Dapp, Krauss, Messerli and Robins2022). Finally, associations were observed for boys, but not for girls.
Our results expand previous research suggesting that gaming among children and adolescents is linked to attentional difficulties, reduced academic engagement, and lower academic achievement (Adelantado-Renau et al., Reference Adelantado-Renau, Moliner-Urdiales, Cavero-Redondo, Beltran-Valls, Martínez-Vizcaíno and Álvarez-Bueno2019; Hellström et al., Reference Hellström, Nilsson, Leppert and Åslund2012; Tiraboschi et al., Reference Tiraboschi, West, Boers, Bohbot and Fitzpatrick2022). Directional associations from gaming to academic motivation were only observed in boys in our study. The sex-specific nature of our findings may be due to boys’ tendency to spend more time playing video games than girls (Leonhardt & Overå, Reference Leonhardt and Overå2021). As such, increased gaming time among boys might encroach upon time otherwise allocated to academic endeavors (Cummings & Vandewater, Reference Cummings and Vandewater2007). In addition, boys and girls also differ in the types of video games they play, with boys tending to play more competitive multiplayer and violent video games (Hastings et al., Reference Hastings, Karas, Winsler, Way, Madigan and Tyler2009). As such, the types of video games played by boys may be more detrimental to their development and eventual academic motivation (Eshuis et al., Reference Eshuis, Pozzebon, Allen and Kannis-Dymand2023; Vargas-Iglesias, Reference Vargas-Iglesias2020; West et al., Reference West, Konishi, Diarra, Benady-Chorney, Drisdelle, Dahmani and Bohbot2018).
The observed association between child gaming and academic motivation highlights the need for further research on the underlying mechanisms that may explain this relationship. One possibility is that video game reward schedules could alter child reward sensitivity, making them less responsive to incentives offered in the school context (Drummond & Sauer, Reference Drummond and Sauer2018; Hodent, Reference Hodent2017; Li et al., Reference Li, Mills and Nower2019). According to one longitudinal study, the association between child screen use (including gaming and TV) at age 10 years and school performance at age 14 years is mediated by child sensation-seeking behavior (Sharif et al., Reference Sharif, Wills and Sargent2010). This suggests that child gaming may impact academic performance by diminishing motivation for more effortful and less novel activities. Another possibility is that associations between gaming and academic motivation are accounted for by increased symptoms of hyperactivity and inattention, which could result from increased gaming and contribute to lower academic motivation (Masi et al., Reference Masi, Abadie, Herba, Emond, Gingras and Amor2021; Paulus et al., Reference Paulus, Sinzig, Mayer, Weber and von Gontard2018; Smith et al., Reference Smith, Langberg, Cusick, Green and Becker2020; Tiraboschi et al., Reference Tiraboschi, West, Boers, Bohbot and Fitzpatrick2022). Finally, as gaming duration increases, the amount of time children spend studying likely decreases (Cummings & Vandewater, Reference Cummings and Vandewater2007), which could undermine their academic skills and, consequently, their academic motivation (Hartanto et al., Reference Hartanto, Toh and Yang2018; Schaffner et al., Reference Schaffner, Philipp and Schiefele2016).
In terms of practical implications and recommendations, our results suggest that parents be sensitized about gaming and its consequences during middle childhood. In addition, practitioners can be encouraged to offer information to parents about the potential consequences of gaming for academic and behavioral outcomes, particularly for boys. Furthermore, practitioners can encourage families to develop a personalized family media use plan, paying particular attention to gaming during middle childhood (HealthyChildren.org, 2024).
In terms of strengths, our study is the first to demonstrate that video game playing predicts lower levels of academic motivation in school-aged boys, and not vice versa. As such, an important strength of our study is our ability to disentangle the direction of this association. Furthermore, specifically estimated associations at the within-person level provide a robust statistical control for stable individual (i.e. baseline motivation and academic competence) and environmental (i.e. socioeconomic status and school environment) factors. Finally, our study is strengthened by using a large population-based sample, which strengthens the generalizability of our findings.
Our study is not without limitations. Our observational design prevents us from discarding some potential time-varying confounders, such as child or parent mental health (Gubbels et al., Reference Gubbels, van der Put and Assink2019). Therefore, no causal conclusions should be drawn, as a third variable associated with both gaming and academic motivation could explain the observed association. Randomized controlled experiments are needed to confirm a causal effect of child gaming on academic motivation. Another limitation is that the use of parent reports of child gaming may be subject to recall and social desirability bias. Nonetheless, our incorporation of child-reported measures of academic motivation helps reduce the possibility of shared measurement bias. Our results are also limited by our focus on the amount of time children spent playing video games, without consideration of video game content or gameplay features. Furthermore, in the present study, we measured intrinsic motivation specifically. As such, future studies could examine associations between child gaming and additional dimensions of motivation and academic outcomes, such as school engagement, connectedness, and academic achievement. A final limitation is that our data were collected between 2005 and 2008. As such, replications are warranted with more recent longitudinal data.
Future studies should examine possible moderators of the observed association. In particular, the characteristics of games and child gameplaying warrant particular attention. For instance, frequent engagement in multiplayer gaming may lead to more adverse outcomes than more casual single-player gaming (Przybylski & Mishkin, Reference Przybylski and Mishkin2016). Similarly, video games featuring gamblification elements (e.g. loot boxes) or employing more intense reward schedules may exert a stronger disruptive effect on academic motivation (Antons et al., Reference Antons, Liebherr, Brand and Brandtner2023). Patterns of gameplay are also important to consider. For example, gaming on weekdays may have different implications for academic motivation than gaming primarily on weekends (Hartanto et al., Reference Hartanto, Toh and Yang2018). Finally, studies could also employ person-centered approaches to help identify subgroups of children who may be more vulnerable to the negative consequences of video game playing on academic outcomes. Indeed, according to one cross-sectional study of 12-year-olds, heavy gamers displayed lower academic engagement when compared to non-gamers, but no associations with academic engagement were found for those who played <3 h/day (Przybylski & Mishkin, Reference Przybylski and Mishkin2016). As such, person-centered approaches may be useful for identifying children who may be more vulnerable to the negative consequences of video game playing.
Conclusion
Child gaming is widespread among children and adolescents. Our study provides compelling evidence that more time devoted to this common leisure activity is associated with lower academic motivation during the school-aged years. Notably, the influence of gaming on academic motivation appears more pronounced in boys, potentially reflecting gender differences in gaming interests and habits. Together, these findings indicate that parents, healthcare professionals, and educators should be encouraged to monitor child gaming, particularly for boys in the early years of elementary school.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/S0033291725101153.
Acknowledgments
The authors would like to thank their statistician and colleague Annie Lemieux for providing insights about data analysis. The authors would like to thank the Institut de la statistique du Québec for access to the data used in this study that were compiled from the final master file of the Québec Longitudinal Study of Child Development (1998–2023), ©Gouvernement du Québec, Institut de la statistique du Québec. The Québec Longitudinal Study of Child Development was supported by funding from the ministère de la Santé et des Services sociaux, le ministère de la Famille, le ministère de l’Éducation et de l’Enseignement supérieur, the Lucie and André Chagnon Foundation, the Institut de recherche Robert-Sauvé en santé et en sécurité du travail, the Research Centre of the Sainte-Justine University Hospital, the ministère du Travail, de l’Emploi et de la Solidarité sociale, and the Institut de la statistique du Québec. Gabriel Arantes Tiraboschi, Caroline Fitzpatrick, and Gabrielle Garon-Carrier had full access to the data used in this study. Other authors did not have clearance to access the data.
Funding statement
This research was funded by the Fonds de recherche du Québec – Société et Culture (FRQSC), scholarship number #331484. Aside from funding, the FRQSC had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Gabrielle Garon-Carrier holds a Canada Research Chair funded by the Social Sciences and Humanities Research Council (grant#950–232804). Caroline Fitzpatrick holds a Canada Research Chair funded by the Social Sciences and Humanities Research Council (grant#2021-00009).
Competing interests
The authors declare none.
Ethical standard
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.
Open access
