Reading Between the Numbers

Fox News ^ | 3/24/04 | Wendy McElroy

[RebelBanker]

Last week, sociologist Dr. Ellie Lee cast doubt upon the increasingly popular theory that postnatal depression affects as many as one in five new mothers, telling the BBC that the research underlying that claim is "wrong."

How can we trust research?

[RebelBanker]

Great article on HOW to look at statistics.

[altura]

Absolutely!

I always say the only statistics I trust are the ones I make up myself.

For example ... there is a huge panic thing now over HRT (hormone replacement therapy).

Supposedly it increases a woman's risk for Something Bad. But if you look at the actual figures and assuming you accept them, the risk is almost statistically zero.

Another warning: writers often lift statistics from other sources taking them on face value ... if traced back to the original studies ... guess what? There are often NO original studies.

[AZFolks]

Fantastic article. One other point to examine, especially in political polls. When a pollster calls, he asks if you are Republican, Democrat, black, brown, white etc. After the results are collected, the pollster "corrects" the poll to represent the distribution of each group in the population. Therefore, if a pollster says there are 38% of Americans who are Democrats he will get a "better" poll number for Democrats than one who assumes 36% are Democrats. This is how the New York Times alters its polls to get the result it wants. Dick Morris explains this, with examples, in his book "Off With Their Heads."