Apple, the Apple logo, and Mac are trademarks of Apple Computer, Inc., registered in the U.S. and other countries. The Made on a Mac Badge is a trademark of Apple Computer, Inc., used with permission.
Rick Mercer nails the Liberal Adaptable Party of Canada, with the help of a cast of dozens perfoming a dynamite song-and-dance routine.
Pass your budget. Take all week.
No one can understand our leader speak.
. . .
We’ll pass your bills. Have no fear.
If anyone needs us we’re over here.
h/t: Bourque
Print This Post
Analysis released today from Statistics Canada’s 2004 General Social Survey (GSS) finds that only 1.5% of Canadians aged 18 and over identify themselves as homosexual or bisexual. This would appear to corroborate an earlier Statistics Canada estimate, based on the 2003 Canadian Community Health Survey, that 1.7% of Canadians aged 18-59 are homosexual or bisexual. When one takes into account sampling errors, the two estimates are insignificantly different. (Note also that that today’s estimate covers a somewhat larger age group.)
According to today’s GSS analysis, Canadian homosexuals and bisexuals report higher levels of violent victimisation, yet are no more fearful of violence than are heterosexuals. A summary appears in The Daily.
Gays, lesbians and bisexuals reported experiencing higher rates of victimization for violent crimes in 2004, including sexual assault, robbery and physical assault, than heterosexuals, according to a new study.
. . .
Despite experiencing higher rates of violence, gays, lesbians and bisexuals did not express higher levels of fear than their heterosexual counterparts.Overall, more than 9 out of 10 gay, lesbian and bisexual Canadians indicated that they were "somewhat" or "very" satisfied with their personal safety, a proportion which was similar to heterosexuals.
The Daily summary, unfortunately, does not convey the many caveats and cautions outlined in the full report (pdf). Some 24,000 Canadians responded to the GSS, so the number of actual respondents who identified as homosexual or bisexual was very small—only a few hundred, I would think. The small sample size resulted in relatively large margins of error, and this is reflected in the report. Consider the numbers footnoted in this excerpt.
Gays, lesbians and bisexuals reported experiencing higher rates of violent victimization . . . including sexual assault, robbery and physical assault, than did their heterosexual counterparts.
The rate for gays and lesbians was almost 2.5 times higher than the rate for heterosexuals (2425 violent incidents per 1,000 population) while the rate for bisexuals was 4 times higher than the rate for heterosexuals (4155 versus 99 violent incidents per 1,000 population).
Footnote 5’s caution concerning sampling errors recurs frequently in the pages of the report.
5. Use with caution, coefficient of variation is high (16.6% to 33.3%).
Large sampling errors associated with small sample size are, I assume, the reason why separate estimates for male and female homosexuals or bisexuals are nowhere to be found, despite the fact that disaggregation by sex is routinely reported in analyses of general population surveys.
Statistics Canada refers throughout its report to “gays and lesbians” (never to “homosexuals”), but never shows separate estimates for gays and lesbians—only for both groups combined. The GSS was unable to find enough gays and lesbians, and bisexuals, to permit gender breakdowns.
The next two paragraphs contain a slightly technical discussion of statistical arcana.
Statistics Canada reports sampling error using a statistic called the “coefficient of variation” (CV), which is equal to the sampling error of the estimate divided by the estimate. Thus, to say that homosexuals reported 242 violent incidents per 1000 homosexuals with a CV of between 16.6% and 33.3% is to say that the sampling error of the estimate is between 1/6 and 1/3 of the estimate of 242. Put that together with the rule of thumb that the 95% confidence interval is equal to the estimate plus and minus twice the sampling error and you can see for yourself how imprecise many of the today’s estimates really are.
I must point out that the confidence intervals thus estimated will not be precisely correct because sample observations are individually weighted in accordance with their demographic characteristics. But the exercise will still generate instructive approximations.
Finally, I note that a published article by Paul Cameron is referenced in the report and included in the bibliography. Pro-homosexual activists have vilified Dr Cameron as a junk scientist (see, e.g., this comment), so it is very interesting to see that Statistics Canada considers at least some of his research credible.
More information on the General Social Survey, including a link to the complete questionnaire, is posted here.
Sources:
Statistics Canada, 2008. "Study: Sexual orientation and victimization". The Daily, 28 February. Statistics Canada catalogue no. 11-001-XIE. http://www.statcan.ca/Daily/English/080228/d080228c.htm (accessed 28 February 2008).
Statistics Canada, 2008. “Sexual orientation and victimization, 2004”. Canadian Centre for Justice Statistics Profile Series. Statistics Canada catalogue no. 85F0033M, no. 16. http://www.statcan.ca/english/research/85F0033MIE/85F0033MIE2008016.pdf (accessed 28 February 2008).
Previous related posts:
Print This Post
Posts
