Another "troll" from Phillip!
Murray Jorgensen
At 08:30 26/03/2002 -0800, Phillip Good wrote:
>Buying statistical software will no more make you a statistician, than
>purchasing « Brain Surgery Made Easy » at the grocery store check-out will
>turn you into a neurosurgeon.<?xml:namespace prefix = o ns =
>"urn:schemas-microsoft-com:office:office" />
>
>Consider the following common errors,
>
>o collecting data so that the observations are dependent
>
>o failing to randomly assign treatments to subjects
>
>o attributing statistical significance to adhoc-posthoc comparisons
>
>o ignoring the assumptions underlying statistical tests, for example,
>using analysis of variance techniques on non-normal data
>
>o confusing correlation and regression with cause and effect
>
>o implicitly ordering p-values (e.g., “p=0.03 is ‘more’ significant than
>p=0.05”)
>
>o relying on stepwise regression
>
>o relying on any regression or data mining technique without validation
>
>o citing standard errors for data drawn from non-symmetric distributions
Dr Murray Jorgensen on leave from:
Mathematics and Statistics Department of Statistics
University of Victoria University of Waikato
PO BOX 3045 STN CSC Hamilton, New Zealand
Victoria, B.C. Email: [log in to unmask]
Canada V8W 3P4 Phone: (250) 721-7460 wk (250) 477 2581 home
|