Studies suggest that youth who are exposed to their first gambling experience at an earlier age are at increased risk of developing problems. However, studies reporting age of onset of gambling exposure as a risk factor for gambling problems are cross-sectional by design and the relationship between both variables are sometimes inferred over extending periods of time. Methodologically speaking, it could induce a recall bias, a fact already documented in numerous areas of research related to high-risk conducts in adolescence. Thus, the goal of this study was twofold: to longitudinally describe, among adolescents, the level of discrepancy between reports of age of initiation to gambling activities, and to assess if the discrepancies could be associated with a certain number of known predictors of gambling participation. Additionally, recall period effect (e.g. time passed between answering the same question) was also assessed. Data were collected from a large longitudinal study on gambling among youth and four measurement times at 1-year interval were used, with only young people who have been introduced to gambling retained in the analyses (n =297; 63.3% boys; mean age = 15.25 years). Results revealed significant inconsistencies about age of initiation to gambling activity between measurement times. Moreover, results also revealed that age (e.g. being older) and time passed are significantly associated to the level of inconsistencies of self-reported age of initiation of gambling activity. Theoretical and practical implications of these findings are discussed.
Keywords: Adolescence; Early exposure; Gambling age of onset; Gambling problems; Reliability.