Here is the Ask Me Anything Talking Data Equity video link
We had an excellent talk about p-values, group size, null vs nil hypothesis and all the rest! Watch the video here:
So how do we address the issue of small group sizes? As in your example on pay equity, often the group that is experiencing inequities is the smaller group.
Hey @Stacey this is indeed the fundamental question. How to deal with the situation of small group sizes is going, as always, to depend on the specific of the question youâre trying to answer with the data.
However, bottom line in terms of p-values and small group sizes is that we really cannot use p-value as the only criteria to decide if a difference between two groups âis realâ or âmattersâ. Itâs just not possible to do it this way if weâre trying to embed equity in the work.
As you say, a lot of the time, the smaller group is experiencing inequities and to then put an even higher âburden of proofâ on this small group for us to thing that the difference matters is really doubling down on the problem our data project is likely to address.
One alternative is to use the actual size of the difference (sometimes called the effect size) as the criterion.
Data Amnesty! The September 9th episode of Talking Data Equity was a data amnesty version. This is when Heather hosts and you can anonymously submit questions you feel are âsillyâ or âembarrassingâ or âtoo beginning.â (They are not )
Here is the recording
Talking Data Equity with Kristin Powell, Deputy Director of Black Futures Lab on their Black Census Project.
Hey there! Weâve posted many of the resources from this talk in the âResourcesâ section of the forum. Here are some of the direct links and you can find all of them (and add more!) over in that section
Hi all! Is it possible to reorganize this forum (and others) so that more recent posts are at the TOP of the thread and we scroll down for older posts? I think thatâs much more user friendly. Thank you!
Hey @thea_bea we canât change the reply order on this topic but youâre totally right that it was annoying to scroll all the way to the bottom manually, so Iâve added a jump-to-last-post link at the top! Thank you so much for the suggestion!
Hi All! You can find many different guides, examples, etc on Bayesian analysis over in our âResourcesâ section of this forum. Here are a few links:
Thank you for adding that @Hans !!