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:
- How I spend the first day of class?
- Why simulation-based?
- How do I utilize technology when teaching with simulation-based inference methods?
- Should we teach bootstrapping or not in the introductory course?
- How do you incorporate student projects in simulation-based introductory statistics course?
- How do you use real data
- The hardest thing about getting started with the simulation-based curricula
- How do you incorporate simulation-based methods in your high school classroom/AP Statistics class
- Why continue to teach normal-based methods of inference, and how to help students make the connection between normal-based and simulation-based methods?
- Implications of teaching simulation-based methods for undergraduate statistics curricula
- Convincing/training others to teach SBI
- Using simulation based methods around the world
- Assessing students’ understanding of SBI
- Assessing effectiveness of an SBI curriculum
- Teaching simulation-based inference in large classrooms
We look forward to your comments.
Please email Jill VanderStoep <email@example.com> or Todd Swanson <firstname.lastname@example.org> 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.
I’m very glad to see the establishment of this blog and the related list-serve: I’m sure we have much to discuss, and many ideas to share. Thanks!
I am excited to learn from this! Thanks for taking the time to put this together.
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?
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.
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.
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.
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
Hi, I will be teaching our school’s first AP Stats and have adopted the ISI text. I am interested in the ISI text chapter correlation to the College Board AP Stats CED units. Where may I find this information? Thank you!
Thank you for the question! I have updated the “index page” (https://www.isi-stats.com/APindex.html) to include a link to the instructor’s guide for the AP edition and renamed the “syllabus” link to pacing guide. Let us know if these documents don’t answer your question?!
Also adding a “topic alignment” document…
I like the format of your Topic Outline alignment document (Introduction to Statistical Investigations: AP® edition Chapter and Section References), but the Units do not match the current for College Board’s AP® Statistics Course and Exam Description (effective 2020) as below. Please double check and confirm, thank you.
CB 1: Exploring One-Variable Data
CB 2: Exploring Two-Variable Data
CB 3: Collecting Data
CB 4: Probability, Random Variables, and Probability Distributions
CB 5: Sampling Distributions
CB 6: Inference for Categorical Data: Proportions
CB 7: Inference for Quantitative Data: Means
CB 8: Inference for Categorical Data: Chi-Square
CB 9: Inference for Quantitative Data: Slopes
Updated 8/10: http://www.isi-stats.com/isi-ap/TopicOutline.pdf