Research designs and the required statistical analyses are getting more and more complicated. It is becoming increasingly common for researchers to consult with statisticians to analyze their data. This can create an issue where the lead researcher on a project does not know how to run the analysis they describe in their papers and presentations. In this episode we discuss whether the lead researcher has an obligation to know how to run, interpret, and describe the analyses in their research projects. Smith raises concerns and potential issues with relying on someone to run the analyses without being able to double check for accuracy. Wingrove and Holden argue that we should trust the statisticians to do their job.

---

Marginally Significant is hosted by:

Andrew Smith @andrewrsmith

Twila Wingrove @twilawingrove

Andrew Monroe @monroeandrew

Chris Holden @profcjholden

You can contact Marginally Significant on Twitter (@marginallysig), through email (marginallysig@gmail.com), or on the web (marginallysignificant.fireside.fm/contact).