Date Published: January 23, 2019
Publisher: Public Library of Science
Author(s): Piers D. L. Howe, Adriana Vargas-Sáenz, Carol A. Hulbert, Jennifer M. Boldero, Atte Oksanen.
In 2016, the gambling habits of a sample of 3361 adults in the state of Victoria, Australia, were surveyed. It was found that a number of factors that were highly correlated with self-reported gambling frequency and gambling problems were not significant predictors of gambling frequency and problem gambling. The major predictors of gambling frequency were the degree to which family members and peers were perceived to gamble, self-reported approval of gambling, the frequency of discussing gambling offline, and the participant’s Canadian Problem Gambling Severity Index (PGSI) score. Age was a significant predictor of gambling frequency for certain types of gambling (e.g. buying lottery tickets). Approximately 91% of the explainable variance in the participant’s PGSI score could be explained by just five predictors: Positive Urgency; Frequency of playing poker machines at pubs, hotels or sporting clubs; Participation in online discussions of betting on gaming tables at casinos; Frequency of gambling on the internet, and Overestimating the chances of winning. Based on these findings, suggestions are made as to how gambling-related harm can be reduced.
Gambling is a common pastime and is found in all countries in the world, although in some, such as the United Arab Emirates, Brunei and Cambodia, it is either illegal or highly restricted. For most individuals, their gambling does not cause problems but for some individuals their gambling harms either themselves, their family or their friends. This harm can be financial, emotional or social. Worldwide, the standardized past year rate of problem gambling varies from a low of 0.12%-0.5% to a high of 5.8%-7.6%, depending on the specific country and the exact scale used to assess it [1, 2]. In general, the rates of problem gambling are lowest in Europe, intermediate in North America and Australia and highest in Asia. The present study focussed on the state of Victoria, Australia, as a condition of its funding.
Participants were sourced from an online survey panel, operated by The Online Research Unit (ORU). The ORU is an Australian research company and is certified by the International Organization for Standardization (ISO 20252 and ISO 26362). The ORU maintains a panel of volunteers who have agreed to participate in online surveys. A mix of incentives including vouchers and charitable donations of small value is provided to participants via the ORU in return for participating in these surveys. Crucially, the participants in our study had an ongoing relationship with the ORU and understood that their responses would be completely anonymous. This was important as this would have encouraged them to disclose any gambling problems they may have had, as gambling problems are stigmatized  and it is known that people are more likely to reveal sensitive information when guaranteed that their responses will be anonymous, as opposed to merely being confidential . The ORU invited members of their survey panel to participate in the online survey via email. To avoid biasing the recruitment, this email did not specify the nature of the survey (i.e., it did not mention that it was related to gambling). In recruiting participants, the ORU matched for age, sex, and location (metropolitan vs regional) relative to the general Victorian population as determined by the demographic data supplied by the 2011 Australian Bureau of Statistics (ABS) survey. (The data for the 2016 census had not yet been released when the survey was conducted). This ensured that the sample was as representative as possible. However, because these participants were drawn from a study panel whose members were self-selected, this sample is not necessarily representative of the general population. The exact breakdown of the sample relative to age, gender and location is detailed in the supplementary information. The analysis was conducted using IBM SPSS version 22 .
Before running the large-scale survey, the survey was piloted on 53 university students. This pilot confirmed that the online survey functioned as expected (i.e., it had no bugs), could be completed in an appropriate length of time, and was comprehensible. Further details regarding this pilot are available in the supplementary information.
Three thousand nine hundred and six Victorians were contacted by the ORU in June or July 2016. Of these, 3361 agreed to participate (86%). Participants ranged in age from 18 to 88 years (Mage = 46.7, SD = 16.7), 48% were male and 71% lived in the metropolitan area of Melbourne. The majority of participants reported that they were born in Australia (77%), were in a relationship (62%), and spoke English at home (94%). The median response time for this survey was 12.9 minutes.
The aim of this study was to determine which factors are the most important predictors of gambling frequency and problem gambling. It was found that the major predictors of gambling frequency were the degree to which family members and peers were perceived to gamble, self-reported approval of gambling, participation in offline discussions of gambling, and PGSI score. In addition, age was an important predictor of gambling frequency for some forms of gambling (e.g., lottery tickets). Because the degree to which others are perceived to gamble was one of the strongest regression predictors of gambling frequency, the study also investigated the accuracy of the perceptions of the degree to which others gamble and approve of gambling. Consistent with Larimer and Neighbors , it was found that, relative to self-reports, individuals overestimated how much others gamble and overestimated how much they approved of gambling. This suggests that campaigns aimed at reducing gambling would do well to focus on correcting these discrepancies. Additionally, such campaigns should use personalised norms, as such norms appear to be particularly effective at affecting gambling behaviour .
Although previous research has shown that a large number of factors are correlated with gambling frequency, it was unclear from that research to what extent those factors could predict gambling frequency or problem gambling. As such, it was unclear what the primary drivers of gambling are, so it was unclear on which factors future research should focus. The main finding of the current study was that only some of the factors that are correlated with gambling frequency actually predict either gambling frequency or problem gambling, beyond that which can be predicted by other factors. Future work should focus on these major predictors as they are likely to be the most important factors driving gambling frequency and problem gambling.