Sounds like an awesome symposium.
How much of the discussion will be based upon what is reproducibility?
In my Data Science class, I intentionally have my students run a model, be it linear or logistic regression, cart models, random forests, etc, with the same data set, but have them use different random seeds to create the training and testing data sets. The results generally have some commonalities. But, often include or remove terms from a previous model. They might all include 'X1' and 'X5' but will disagree on everything else. All the data I use was created by me, so we know what is 'in the model'. Out of 100+ random seeds for the models, no one has got everything and only a few times do the models agree with each other completely, meaning they show the exact same variables are important.
------------------------------
Andrew Ekstrom
Statistician, Chemist, HPC Abuser;-)
------------------------------
Original Message:
Sent: 05-10-2022 16:22
From: Monnie McGee
Subject: Symposium on Scientific Reproducibility and Statistical Significance
The Dedman College Interdisciplinary Institute and the Department of Statistical Science at Southern Methodist University are proud to announce a Symposium on Scientific Reproducibility and Statistical Significance. The symposium will take place on Friday, June 3, from 9:30 - 3:30 on the SMU campus in Dallas, Texas. There is also a virtual attendance option.
The aim of the symposium on scientific inquiry and statistical significance is four-fold
1. To disseminate the stance of the American Statistical Association on the appropriate use of results from null hypothesis significance testing for analysis of studies from social science, science, engineering, and humanities.
2. To offer alternatives to such testing
3. To discuss changes to publication policies that would benefit both individual scientists and science writ large.
4. To discuss methods of educating current and future scientists in appropriate methods for gathering and analyzing data.
The link below is a registration link. Please let us know whether you can attend.
https://www.eventbrite.com/e/scientific-reproducibility-and-statistical-significance-symposium-tickets-327521494607
------------------------------
Monnie McGee
Associate Professor
Department of Statistical Science
Southern Methodist University
Dallas,TX United States
------------------------------