Many thanks for the replies and suggestions thus far. I am sorry for not clarifying that I illustrate the limitations of a 5% significance level to students by sharing the notion of clinical significance and providing a topical example. However, this does not eliminate the value that would be gained from having a resource which points the student to how the notion of statistical significance evolved, similarly w.r.t. to that of confidence interval.
Additionally, there is a lot of literature around, I am sure, such as 'Statistics with Confidence' which say why the confidence interval is the preferable choice when interpreting tests of significance. However, it would also be of much interest to see a historical account (suitable for non-statisticians) of how the confidence interval evolved.
With a view to being more helpful, contexts in much this support material would prove most useful would be in the interpretation of a t-test, say and in the use of Forest Plots to test for the significance of an effect.
Hope this clarifies any misunderstandings
Please keep sending these suggestions!
Yours most gratefully
Margaret
---------------------------------
Yahoo! Messenger NEW - crystal clear PC to PCcalling worldwide with voicemail
|