Browse By Day
Browse By Time
Browse By Panel
Browse By Session Type
Browse By Topic Area
Goals: The current longitudinal study tested reciprocal relationships between video game play and depressive symptoms in a national sample of youth and young adults from the Add Health (N = 9,421, Mage at T1 = 16.15 years, SD = 1.64, 55% female), over the span of 11 years (Waves 2, 3, and 4, ages 16 to 27).
Hypotheses: Based on the findings in existing literature it was hypothesized that (H1) gaming in excess would predict developmental changes in depressive symptoms by the next time point; that the effect of gaming on depressive symptoms would be curvilinear, namely a null slope at lower amounts of play time, but an increasingly positive relationship at higher reported amounts of play time; that (H2) gaming habits would be transient; that (H3) gamers categorized as excessive (spending +2 SD above the mean play time) at T1 (age 16) would report consistently higher mean levels of depressive symptoms than casual gamers and non-gamers, but also a higher rate of change (slope) in depressive symptoms over time; and that (H4) for males the relationship between gaming and depressive symptoms would be stronger than females. The study also tested whether prior depressive symptoms affected developmental changes in gaming, and if so, to what extent (RQ1)?
Methods: Intensity of video game use was measured as hours of playing video or computer games weekly. Depressive symptoms were measured by nine items of the Center for Epidemiologic Studies Depression Scale (α = 80 to .81, Radloff, 1977). A cross-lagged structural equation model was estimated with depressive symptoms and play time across 3 time points, with each earlier construct predicting the developmental changes in later constructs to test the RQ1 and H1 (see Figure 1). Age, race, SES, and family structure were included as control variables. To test for curvilinear relationship of gaming with depressive symptoms, a quadratic term of play time at T1 and T2 was included as a predictor of T2 and T3 depressive symptoms, respectively. Stability coefficients from the model and raw weekly play times across the 3 time points were examined to test H2. Sex differences (H4) were investigated by testing the cross-lagged model in a multi-group framework. A multi-group latent growth model tested the difference in depression trajectories among non-gamers, casual gamers, and excessive gamers (H3).
Results: Findings indicated that excessive gaming, but not gaming in general, predicted increases in depressive symptoms over time (H1, Figure. 1), and that excessive gaming was largely transient over time (H2, Figure. 1). Only a very small minority (0.1%) of the sample was considered excessive at all 3 assessments; but (3) no pairwise differences in intercept or slope of depression trajectories were found among non-gamers, casual gamers, and excessive gamers (H3, Table 1). Depressive symptoms in adolescence lead to a reduction of time spent playing games by T3 (age 21; RQ1, Figure 1). Multi-group model tests by sex provided additional evidence that the longitudinal relationships from excessive gaming to depressive symptoms were supported for male, but not for female youth (H4).