Markham Francis, Doran Bruce, Young Martin
Fenner School of Environment and Society, The Australian National University, Canberra, ACT, Australia.
Fenner School of Environment and Society, The Australian National University, Canberra, ACT, Australia.
Soc Sci Med. 2016 Aug;162:106-14. doi: 10.1016/j.socscimed.2016.06.008. Epub 2016 Jun 8.
An emerging body of research has documented an association between problem gambling and domestic violence in a range of study populations and locations. Yet little research has analysed this relationship at ecological scales. This study investigates the proposition that gambling accessibility and the incidence of domestic violence might be linked. The association between police-recorded domestic violence and electronic gaming machine accessibility is described at the postcode level. Police recorded family incidents per 10,000 and domestic-violence related physical assault offenses per 10,000 were used as outcome variables. Electronic gaming machine accessibility was measured as electronic gaming machines per 10,000 and gambling venues per 100,000. Bayesian spatio-temporal mixed-effects models were used to estimate the associations between gambling accessibility and domestic violence, using annual postcode-level data in Victoria, Australia between 2005 and 2014, adjusting for a range of covariates. Significant associations of policy-relevant magnitudes were found between all domestic violence and EGM accessibility variables. Postcodes with no electronic gaming machines were associated with 20% (95% credibility interval [C.I.]: 15%, 24%) fewer family incidents per 10,000 and 30% (95% C.I.: 24%, 35%) fewer domestic-violence assaults per 10,000, when compared with postcodes with 75 electronic gaming machine per 10,000. The causal relations underlying these associations are unclear. Quasi-experimental research is required to determine if reducing gambling accessibility is likely to reduce the incidence of domestic violence.
越来越多的研究表明，在一系列研究人群和地点中，问题赌博与家庭暴力之间存在关联。然而，很少有研究在生态尺度上分析这种关系。本研究调查了赌博可及性与家庭暴力发生率可能相关的这一命题。在邮政编码层面描述了警方记录的家庭暴力与电子游戏机可及性之间的关联。每10000人的警方记录家庭事件数和每10000人的与家庭暴力相关的人身攻击犯罪数被用作结果变量。电子游戏机可及性的衡量标准是每10000人的电子游戏机数量和每100000人的赌博场所数量。使用贝叶斯时空混合效应模型，利用澳大利亚维多利亚州2005年至2014年的年度邮政编码层面数据，对一系列协变量进行调整，以估计赌博可及性与家庭暴力之间的关联。在所有家庭暴力和电子游戏机可及性变量之间发现了具有政策相关性的显著关联。与每10000人中有75台电子游戏机的邮政编码地区相比，没有电子游戏机的邮政编码地区每10000人的家庭事件数少20%（95%可信区间[C.I.]：15%，24%），每10000人的家庭暴力攻击数少30%（95% C.I.：24%，35%）。这些关联背后的因果关系尚不清楚。需要进行准实验研究来确定减少赌博可及性是否可能降低家庭暴力的发生率。
