Our goal is to provide a discussion forum for those interested in using simulation- and randomization-based inference as a large component of their introductory statistics courses.  We will have postings from developers of several curricula, with their insights as to why and how to use these methods. See Overview for more information.

Follow the links below to access these discussion topics:

We look forward to your comments.

Please email Jill VanderStoep <> or Todd Swanson <> if you have any problems or suggestions for future posts.

This material is based upon work supported by the National Science Foundation under Grant Number DUE-1323210.  Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

7 thoughts on “Introduction

  1. Vicki-Lynn Holmes

    When teaching my students ANOVA (Multiple Means), I’ve only been using the MAD statistic. I’m not certain WHEN to best use the three options (or which option is better): MAD vs Max-Min vs F-statistic.

    What do you all use and why?


    1. Beth Chance

      I don’t know if there is a “best” though you could talk about power of the different approaches, Max-Min is not very powerful, MAD doesn’t reflect sample size information. I do like to start with MAD and Max-Min as the transition to multiple groups because I think students find them to be more intuitive statistics (and in fact often suggest them on their own). Once they realize all we need to do with multiple groups is find a good statistic and everything proceeds as before, then we talk about the F, its logic, and that it has the nice mathematical model.

      1. Nathan Tintle

        I agree with Beth. I usually have students brainstorm their own options first. They usually come up with Max-Min and MAD (more or less) on their own. I can then talk about information loss from Max-Min, and no math model for MAD, but that there is a model for the F-statistic. Discussion around Chi-square proceeds in much the same way.

  2. George Cobb

    As Beth and Nathan point out, MAD and max-min offer a good place to start because they appeal to intuition and students may suggest them on their own. The F-statistic? Not.

    All the same, there are important reasons to consider teaching F in addition. If distributions are normal and SDs are equal, F is best, not just for comparing means, but also for fitting equations to data (regression). F is the statistic most often used in published research.

    However, you have to work to make F seem like a reasonable choice. (If our goal is to compare means, why should we rely on the ratio of two different measures of variability?) Making the case takes time and effort. As I see it, whether to teach F is a judgment call based on who your students are.

  3. Husen Darmawan

    I just like the helpful information you supply to your articles. I will bookmark your weblog and test again here frequently. I am somewhat sure I will learn many new stuff proper here! Best of luck for the following


Leave a Reply

Your email address will not be published. Required fields are marked *