Surveying All Adults in a Household: the Potential for Reducing Bias in Prevalence Estimates and the Opportunity to Study Households with More than One Problem Gambler

Traditionally gambling prevalence studies have randomly sampled one adult per household. As a result, prevalence studies have not examined the potential for more than one problem gambler to be living in the same household. The 1998 Nova Scotia Video Lottery Survey screened all adults in the household for video lottery gambling behaviours and then interviewed all regular VL gamblers to identify problem gamblers. An analysis of bias due to sampling one adult per household found that prevalence studies could be overestimating prevalence by 23%, and that another regular gambler living with a problem gambler was likely (54%) to also be a problem gambler.