Hi Everyone.

I hope you're doing as well as possible in these turbulent times.

In my decades of working on data projects in the field, one of the most common internal sources of a lack of equity I've seen has been the way the project methodology is chosen. Frequently when I've seen community members and local teams and data contributors most disappointed or disillusioned at the end of a data project it's because the questions the data will be used to answer and the methodology used to answer those questions have been chosen (or dictated) by outside "experts." 

A very well run "gold standard" project can tell us that the average income in the community rose by $100. But it will hide the fact that this is because some people's income rose by $1000 and many people's incomes fell by $100.

This week on Data Amnesty we covered the topic "What the heck is an RCT, and why would I use one anyway?" The link to the video is here. It's the beginning of a series of tools that will help you take back the power of designing research questions and choosing methodologies.

RCTs, or Randomized Controlled Trials are important to know about from an equity perspective in data projects. This is because RCTs are often claimed to be "the gold standard of evidence." And they are not. And they are often much worse at producing evidence that emphasizes equitable outcomes and statistical models that include critical variables or indicators necessary to embed an equity lens into data. 

RCTs are the gold standard at answering one very specific research question. And are often unable to answer many other questions that are frequently more important to communities.

More tools to come soon.

Love,
Heather