953
edits
Changes
Jump to navigation
Jump to search
Line 36:
Line 36: + +
Talk:Harvard abortion study (view source)
Revision as of 20:06, April 24, 2007
, 20:06, April 24, 2007no edit summary
: Great analysis, but I do quibble with your statement about women who did not have any children. The ranges are so wide on the 95% CI that no conclusions can be drawn about them. The numbers are statistically insignificant, another sign of how the 100,000+ women in this study were simply too young for studying breast cancer.--[[User:Aschlafly|Aschlafly]] 15:59, 24 April 2007 (EDT)
: Great analysis, but I do quibble with your statement about women who did not have any children. The ranges are so wide on the 95% CI that no conclusions can be drawn about them. The numbers are statistically insignificant, another sign of how the 100,000+ women in this study were simply too young for studying breast cancer.--[[User:Aschlafly|Aschlafly]] 15:59, 24 April 2007 (EDT)
::Andy - It was statistically not significant, though barely - p = .06 - if I understand correctly what you were referring to. The criterion for statistical significance is arbitrary and doesn't always reflect whether something is meaningful. On the other hand, statistical significance is dependent on the sample size, and with such a huge sample any effect that doesn't reach significance is bound to be small and likely not meaningful. So I can't argue with your quibble. I already have a copy of the article but I appreciate the offer. [[User:Murray|Murray]] 16:06, 24 April 2007 (EDT)
== Opinions in article ==
== Opinions in article ==