Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Biotech / Medical : SARS and Avian Flu -- Ignore unavailable to you. Want to Upgrade?

To: Wyätt Gwyön who wrote (747)	5/7/2003 9:21:53 AM
From: Biomaven	Respond to of 4232

assuming a parametric g distribution can you explain what this is? it gives different figures for parametric g and non-parametric g distributions. which are more believable? Basically a parametric statistic makes an assumption about how the data is distributed, and then, given that assumption, fits one or more parameters to the particular data. A non-parametric statistic makes no assumption about what the underlying data looks like. In general, a parametric statistical analysis will give you more power - that is you can do a better analysis with the same amount of data than you can with a non-parametric statistical analysis. But whether your conclusions are correct depends on whether the underlying data fits the distribution you have assumed. To be more concrete, if the distribution of say deaths over time in fact had two peaks - say a bunch died three days after onset and then there was another peak after two weeks, then that data would not fit a gamma distribution well. But the graphs they show don't seem like that, indicating that perhaps the parametric approach is reasonable here. Caveat: It's a long time since I last dealt with statistics like this, and I would be more than happy to be corrected by any resident statisticians. Peter