Image:Redraft.doc

From CyberOne Wiki

• Image
• File history
• Links

Partial rewrite of section 3a with comments

overview comments as someone with a science background (BS bio, taken some stats classes) the arguments re their small sample sizes especially seemed quite inadequate. small sample sizes (well, in this case generally it wasn't the sample that was small but rather those within the sample that had a given characteristic) may result in low p values (http://en.wikipedia.org/wiki/P-value) and hence statistical significance. it's the p value that matters, not the sample size. (though it's harder to get a tiny p value with a small sample size.) also, saying, well if it had been 8 it wouldn't have been statistically significant doesn't really make sense at all. You can argue that the p value is borderline (very close to 0.05 or whatever the chosen alpha level was http://en.wikipedia.org/wiki/Significance_level) but you cannot argue a counterfactual. there were 9, not 8! at a certain point it is arbitrary to choose an alpha value but 0.05 is pretty much standard. Jendawson 11:26, 4 November 2008 (EST)

File history

Click on a date/time to view the file as it appeared at that time.

• Search for duplicate files
• Edit this file using an external application

  See the setup instructions for more information.

Links