🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

I agree totally. Ditto for the 500+ other ways to say the same thing. Here is the beginning of Hankins' list. mchankins.wordpress.com/2013/04/21/sti…

Prism 9 will only put the automatic "stars on graph" after you ran ANOVA and requested multiple comparison tests. Louise Topping compared each mean to every other mean. With 4 columns (groups), that is 6 follow-up tests. You asked about 7, but the screen video goes by fast.

Good questions! Prism 9 only puts asterisks on the graph after you chose ANOVA and multiple comparison tests. Yes, multiple comparisons are accounted for, and actual (multiplicity adjusted) p-values are shown in the results table and can be placed on the graph instead of **.

Please elaborate on your suggestion. When you choose ANOVA, GraphPad Software Prism asks you to specify which (if any) comparisons to make, and only selected comparisons will appear on the graph (and you can choose to label all comparisons or only those with P < 0.05).

Correct! If you request fewer preselected comparisons (based on experimental design and questions, not based on seeing the data), the multiple comparison testing will have more power. I've passed along the suggestion to add wording to this effect on the dialog. Thanks!

My Intuitive Biostatistics is now online for all Prism users. Update Prism to 9.1 and link to "Prism Academy" from the Welcome dialog. Set up an account for the Academy, which is part of your Prism license, and you'll have access to the book, videos, and more.

Eisner has written a great review/editorial about psuedoreplication, pointing out that triplicate values from triplicate animals is not the same as n=9. The solution is to run analyses that take into account nesting. rupress.org/jgp/article-pd…

Vickers and colleagues have published a great set of guidelines for making graphs and tables. The guidelines were written for researchers in urology, but are general enough to apply to most fields, basic and clinical. auajournals.org/doi/pdf/10.109…

The perfect paper for April 1. Since low p-values don't mean very much, Hussey explains how to generate them more quickly. open.lnu.se/index.php/meta…

"If the five years since the ASA's statement have shown anything, it is that the inferential virus of NHST (null hypothesis significance testing) is not going away any time soon."

Assuming large(ish) sample sizes.

I’m glad you like my book! But I don’t use testing for “equivalence” and “noninferiority” interchangeably. I try to distinguish carefully. But they are both in the chapter title so I can see how someone might misinterpret.

Lisa Motulsky crossed 8.2 MILE. Time 01:34:56, pace 11:35 min/mile, est. finish 11:46 am track.rtrt.me/e/SVE-BIGSUR-2… bsim

Lisa Motulsky crossed 12 MILE. Time 02:06:46, pace 11:21 min/mile, est. finish 11:31 am track.rtrt.me/e/SVE-BIGSUR-2… bsim

Lisa Motulsky crossed 21.2 MILE. Time 03:51:12, pace 11:22 min/mile, est. finish 11:40 am track.rtrt.me/e/SVE-BIGSUR-2… bsim

Lisa Motulsky crossed 24 MILE. Time 04:22:20, pace 11:08 min/mile, est. finish 11:40 am track.rtrt.me/e/SVE-BIGSUR-2… bsim

Lisa Motulsky completed the Big Sur MARATHON! Time 04:47:35 (UNOFFICIAL) track.rtrt.me/e/SVE-BIGSUR-2… bsim

We recently published a review of lognormal distributions. It is long, but ends with a concise set of recommendations and a list of common misconceptions. And a poem! pharmrev.aspetjournals.org/action/showPdf…

Here is the list of misconceptions about lognormal distributions. pharmrev.aspetjournals.org/action/showPdf…