18.08.2013 Views

gambling in alberta - Research Services - University of Lethbridge

gambling in alberta - Research Services - University of Lethbridge

gambling in alberta - Research Services - University of Lethbridge

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Prevalence<br />

Problem Gambl<strong>in</strong>g and Related Indices<br />

The impact <strong>of</strong> legalized <strong>gambl<strong>in</strong>g</strong> problem <strong>gambl<strong>in</strong>g</strong> is a particularly important question. The<br />

first th<strong>in</strong>g to recognize is that significant rates <strong>of</strong> problem <strong>gambl<strong>in</strong>g</strong> were shown to exist <strong>in</strong> the<br />

United States <strong>in</strong> the 1970s even before the widespread <strong>in</strong>troduction <strong>of</strong> legal <strong>gambl<strong>in</strong>g</strong> (Kallick et<br />

al., 1979). The same is almost certa<strong>in</strong>ly true <strong>of</strong> Alberta, although there are no prevalence<br />

studies prior to 1993 to document how significant a problem it was.<br />

What is available are 12 population prevalence surveys <strong>of</strong> <strong>gambl<strong>in</strong>g</strong> and problem <strong>gambl<strong>in</strong>g</strong><br />

conducted between 1993 and 2009. The details <strong>of</strong> these studies are conta<strong>in</strong>ed <strong>in</strong> Appendix B.<br />

Methodological differences make comparisons between surveys difficult. This issue was briefly<br />

mentioned earlier <strong>in</strong> this report with respect to compar<strong>in</strong>g participation rates for each form <strong>of</strong><br />

<strong>gambl<strong>in</strong>g</strong> from one year to the next. However, these considerations are much more impactful<br />

for problem <strong>gambl<strong>in</strong>g</strong> prevalence compared to <strong>gambl<strong>in</strong>g</strong> prevalence, as there is very little<br />

variation (or debate) about how to ask about past year participation <strong>in</strong> <strong>gambl<strong>in</strong>g</strong> activities,<br />

participation questions are much less ‘sensitive’ than questions about problem <strong>gambl<strong>in</strong>g</strong>, and<br />

the higher prevalence rates <strong>of</strong> <strong>gambl<strong>in</strong>g</strong> participation mean that methodological noise has a<br />

proportionally smaller impact.<br />

Indeed, Williams & Volberg (2009, 2010) have documented that problem <strong>gambl<strong>in</strong>g</strong> prevalence<br />

rates can vary by a magnitude <strong>of</strong> 5 depend<strong>in</strong>g on which assessment <strong>in</strong>strument is used, how the<br />

survey is described to potential participants (i.e., ‘<strong>gambl<strong>in</strong>g</strong> survey’ versus a description that<br />

does not clearly <strong>in</strong>dicate it is a <strong>gambl<strong>in</strong>g</strong> survey), how it is adm<strong>in</strong>istered (face-to-face versus<br />

telephone), and the threshold used to ask questions about problem <strong>gambl<strong>in</strong>g</strong> (i.e., any past<br />

year <strong>gambl<strong>in</strong>g</strong> versus <strong>gambl<strong>in</strong>g</strong> at least once a month). Further illustration <strong>of</strong> the importance <strong>of</strong><br />

methodology is the difference <strong>in</strong> problem <strong>gambl<strong>in</strong>g</strong> prevalence rates obta<strong>in</strong>ed <strong>in</strong> traditional<br />

telephone surveys versus onl<strong>in</strong>e panel surveys. Alberta onl<strong>in</strong>e panel surveys have been<br />

conducted <strong>in</strong> 2006, 2007, 2008, and 2009 (with the present study conduct<strong>in</strong>g them <strong>in</strong> 2008 and<br />

2009). In all cases, the prevalence <strong>of</strong> problem <strong>gambl<strong>in</strong>g</strong> was approximately twice as high as the<br />

rates obta<strong>in</strong>ed via telephone surveys (it is beyond the scope <strong>of</strong> the present paper to discuss this<br />

<strong>in</strong> detail, but the evidence compiled by Dr. Robert Williams <strong>in</strong>dicates that the telephone surveys<br />

produce the more accurate rates) (Williams & Volberg, 2010).<br />

Thus, while it is potentially problematic to make comparisons between surveys, there are some<br />

th<strong>in</strong>gs that make this less difficult <strong>in</strong> the present situation. The fact that the methodological<br />

procedures were very similar <strong>in</strong> many <strong>of</strong> these studies 71 (i.e., telephone adm<strong>in</strong>istration, survey<br />

be<strong>in</strong>g described as a ‘<strong>gambl<strong>in</strong>g</strong> survey’, use <strong>of</strong> any past year <strong>gambl<strong>in</strong>g</strong> be<strong>in</strong>g only criterion for<br />

be<strong>in</strong>g asked problem <strong>gambl<strong>in</strong>g</strong> questions). The exceptions were a) the 2002 Statistics Canada<br />

survey that was adm<strong>in</strong>istered face-to-face and was described as a survey assess<strong>in</strong>g ‘well-be<strong>in</strong>g<br />

71 This is also related to the fact that Dr. Harold Wynne conducted the 1993, 1998, and 2001 studies and Dr. Robert<br />

Williams conducted the 2007, 2008, and 2009 studies.<br />

163

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!