----------
> From: Michael Plog <[log in to unmask]>
> Reply-To: American Evaluation Association Discussion List
> <[log in to unmask]>
> Date: Sat, 9 Apr 2005 07:53:21 -0500
> To: [log in to unmask]
> Subject: Evaluation Humor
>
> The real meaning of some scientific jargon, attributed (without verification)
> to Dyrk Schingman, Oregon State University:
>
> ³It has long been known²... I didn¹t look up the original reference.
> ³A definite trend is evident²... These data are practically meaningless.
> ³While it has not been possible to provide definite answers to the
> questions²... An unsuccessful experiment, but I still hope to get it
> published.
> ³Three of the samples were chosen for detailed study²... The other results
> didn¹t make any sense.
> ³Typical results are shown²... This is the prettiest graph.
> ³These results will be in a subsequent report²... I might get around to this
> sometime, if pushed/funded.
> ³The most reliable results are obtained by Jones²... He was my graduate
> student; his grade depended on this.
> ³In my experience²... Once
> ³In case after case²... Twice
> ³In a series of cases²... Thrice
> ³It is believed that²... I think.
> ³It is generally believed that²... A couple of other guys think so too.
> ³Correct within an order of magnitude²... Wrong.
> ³According to statistical analysis²... Rumor has it.
> ³A statistically oriented projection of the significance of these findings²...
> A wild guess.
> ³A careful analysis of obtainable data²... Three pages of notes were
> obliterated when I knocked over a glass of beer.
> ³It is clear that much additional work will be required before a complete
> understanding of this phenomenon occurs²... I don¹t understand it.
> ³After additional study by my colleagues²... They don¹t understand it either.
> ³Thanks are due to Joe Blotz for assistance with the experiment and to Andrea
> Schaeffer for valuable discussions²... Mr. Blotz did the work and Ms. Shaeffer
> explained to me what it meant.
> ³A highly significant area for exploratory study²... A totally useless topic
> selected by my committee.
> ³It is hoped that this study will stimulate further investigation in this
> field²... I quit.
>
> *************************************
>
> Did you hear about the politician who promised that, if he was elected, he'd
> make certain that everybody would get an above average income?
>
> *************************************
>
> The Ten Commandments of Statistical Inference
>
> 1.Thou shalt not hunt statistical inference with a shotgun.
> 2.Thou shalt not enter the valley of the methods of inference without an
> experimental design.
> 3.Thou shalt not make statistical inference in the absence of a model.
> 4.Thou shalt honor the assumptions of thy model.
> 5.Thy shalt not adulterate thy model to obtain significant results.
> 6.Thy shalt not covet thy colleagues' data.
> 7.Thy shalt not bear false witness against thy control group.
> 8.Thou shalt not worship the 0.05 significance level.
> 9.Thy shalt not apply large sample approximation in vain.
> 10.Thou shalt not infer causal relationships from statistical significance.
>
>
> *************************************
>
> Torture the data long enough and they will confess to anything.
>
> *************************************
>
> Statistics in the hands of an evaluator are like a lamppost
> to a drunk--they're used more for support than illumination.
>
> *************************************
>
> ------------------------------------------------------------------
> EVALTALK - American Evaluation Association (AEA) Discussion List. See also
> the website: http://www.eval.org
> To unsubscribe from EVALTALK, send e-mail to [log in to unmask]
> with only the following in the body: UNSUBSCRIBE EVALTALK
> To get a summary of commands, send e-mail to [log in to unmask]
> with only the following in the body: INFO REFCARD
> To use the archives, go to this web site:
> http://bama.ua.edu/archives/evaltalk.html
> For other problems, contact a list owner at [log in to unmask] or
> [log in to unmask]
>
|