Hi I'm with others here that students should learn NHST and would just add that most statistical procedures have varying shortcomings. With ESs, for example, there is the danger of concluding that a small ES means that the effect is trivial ... not true. And I haven't looked at recent developments in use of CI, but wonder how it accommodates things like adjusting for number of comparisons in post hoc testing? Or more complex results (e.g., partitioning interaction in mixed factor anova)?
And without following Michael P's wise advice to actually read the article he abstracted, I wonder in such studies of unwanted consequences of reporting NHST results (i.e., sig vs nonsig) whether they just did the conclusion or reported p values, as now recommended by APA (and common sense). Do people misinterpret the difference between an effect with p = .053 versus an effect with p = .048? Finally, intro stats cannot be the be-all and end-all of statistical learning for students going further in psychology (or other fields that use statistics ... lots on data mining right now in the business world). We're laying a foundation and part of that foundation should be a proper understanding of NHST, along with other concepts (e.g., strength of relationships). Take care Jim James M. Clark Professor of Psychology 204-786-9757 204-774-4134 Fax [email protected] >>> Michael Palij <[email protected]> 20-Feb-12 10:13 AM >>> Okay, I beg your indulgence and participation in an unscientific poll where you can either post your response to the TiPS list (for discussion) or email your response directly to me. I am finishing a book review on an undergraduate statistics textbook that (a) attempts to eliminate all null hypothesis signitifcance testing (NHST) in favor of focusing on effect sizes (ES), confidence intervals (CI), and (old fashioned) meta-analysis and (b) encourages research on "statistical cognition" which, according to the author, shows that teaching NHST causes greater confusion in students than an ES/CI approach . Given that limited description, I'm going to make this into a 2-alternative forced choice question: Would you use such a textbook as the main textbook in the first/introductory statistics in psychology course? [ [ Yes [ ] No Comments? If you care to, you might comment on whether current intro stat textbooks do an adequate job of covering issues such as effect sizes and confidence intervals (these days I use some version of Gravetter and Wallnau which, in my opinion, do an adequate job introducing the topics which I assume lay the foundation for a more advanced undergraduate course in statistical methods). Thanks in advance for your cooperation. -Mike Palij New York University [email protected] P.S. And, no, this not about procrastinating on finishing the book review. Well, mostly it's not. ;-) --- You are currently subscribed to tips as: [email protected]. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=13251.645f86b5cec4da0a56ffea7a891720c9&n=T&l=tips&o=16147 or send a blank email to leave-16147-13251.645f86b5cec4da0a56ffea7a89172...@fsulist.frostburg.edu --- You are currently subscribed to tips as: [email protected]. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=13090.68da6e6e5325aa33287ff385b70df5d5&n=T&l=tips&o=16172 or send a blank email to leave-16172-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu
