Topic 9 - Presenting the Data

Label your axes! (XKCD)

Readings

Open Textbook Chapter 11: Presenting your research

Supplementary reading:

Learning statistics with jamovi

A free book by Danielle Navarro - a really excellent statistical primer (here as a resource, you are NOT expected to read the whole thing!).

Things to do

Work on your assignment!

jamovi user guide

Free video course on jamovi

Forum

Topic - P-hack your way to success!

This is a bonus Discussion Forum based on a fun activity we did at last year’s Intensive School. I think you’ll enjoy it!

We’ve talked quite a bit about p-hacking in the course of this unit, but what exactly is it? This Shiny app demo was written by one of my favourite Shiny developers, Felix Schonbrot, and I think it really demonstrates how easy it is to produce results from thin air.

Shiny p-hacker

There’s a great blog post to go with it (note: it’s supposed to be ironic!):

Introducing p-hacker

Have a play with the app - it’s best if you set the true effect size to 0 and choose a small n for each group (maybe try 6 different DVs with a group size of 15 - 20). Name the effect whatever you like - say, “Mindfulness Intervention” or something.

Did you get a significant effect in one of your DVs straight away (p < .05)? Maybe it was close?

From my original slides:

  • If that doesn’t terrify you enough, have a look at your “best” result and see if you can make it significant by removing an outlier or two
  • If that doesn’t work, try adding another 5 participants at a time
  • Also try “controlling” for age, gender, SES.
  • How many of you have something “significant” now?
  • You can always try a ”new” experiment (warning: this can be a little addictive…)

If you like, you could save the images of the “effects” you got and attach them to your post.

What does this teach you about how you should approach your research?

Previous