Tags:
create new tag
, view all tags, tagging instructions
Return to Online Journal Club Discussion

(PLEASE NOTE: On the bottom of this page, you will find the topic for discussion and the name of the contributor.)

Cruder than Gigerenzer describes

Email Notification of Changes: Click here and add TITLE of the topic to the body of the email.

Please add comments and then click on the "Add comment" button.

 

OnlineJournalDiscussionForm edit

Title_Discussion_Topic Cruder than Gigerenzer describes
Name_Topic_Initiator Peter Bacchetti
Online_Journal_Club_Meeting Meeting 3
Description - Problem to be explored I don’t think the id wants probabilities; it wants Yes or No. In the actual practice that I see day to day, the fallacy is not confusing P(D|H) with P(H|D), but rather interpreting P>0.05 as proving the null hypothesis and P<0.05 as disproving the null hypothesis and indicating that the observed effect is real (and accurate). This interpretation is so convenient and desirable that it is very hard to resist, even for statisticians who are well aware of how wrong it is. This may reflect a strong preference that people have, sometimes called dichotomania, for Yes/No rather than continuous or probabilistic thinking.

I would argue that the roots of this problem lie squarely in the Neyman-Pearson “hypothesis testing” framework. Gigerenzer and Fisher (the later version) got it right that automatic decision making is not the usual goal in scientific research. We publish results to convey information and, sometimes but not always, to try to convince others of a decision that we believe is warranted by the evidence. Simply making an automatic decision does not require publication, and it is no accident that the standard examples are from industrial quality control rather than science.

Estimates, confidence intervals, and actual P-values are often reported, but they rarely figure into the interpretation of a study’s results as they should; interpretation instead usually focuses just on whether or not P<alpha (almost always, 0.05).
See Also

Disclaimer The views expressed within CTSpedia are those of the author and must not be taken to represent policy or guidance on the behalf of any organization or institution with which the author is affiliated.
Topic revision: r2 - 02 Dec 2013 - 11:41:23 - MaryBanach
 

Copyright &© by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding CTSPedia? Send feedback