Research 101

Two years ago, I tutored an undergraduate subject called Mind, Brain, and Behaviour. It is an introductory subject, 1st year, and features (among other things) really really basic research methods. I remember thoroughly enjoying the teaching back in 2014; going over the basic concepts again was fun, the students were great, and the support for the tutors absolutely excellent.

This semester, I am teaching MBB again. And although I am still loving it – nothing about the contents, students, nor support has changed – this time it is not quite the same. One reason for this is that a lot has happened in psychology over the past two years. So, to keep up with contemporary issues in our beloved science, in one of the research methods tutorials we will be discussing the Replicability Project, and Gilbert et al.’s response (and the response to the response, etc). I am mildly apprehensive about this tutorial, since I don’t know whether I am able to summarize the “state of the field” in any straightforward, introductory, way. But  I am also looking forward to it – and maybe in discussing it with the students, I will work  out what I think?

There are also two other reasons why 2016 ≠ 2014. These things are related to the above, but more “personal” (and also more minor). They are called… Bayes, and R.

I say “personal”, because throughout the last 6 years of my degree, Bayesian approaches to… well, to truth, and knowledge, and statistics, have popped up now and again as an alternative to frequentist approaches. And every time, I have been fascinated, inspired… and then I have gone right back to null hypothesis significance testing. (But every time, I do also learn a little bit more about Bayes – most recently, by following some exercises on Arbital. ) And R has likewise haunted me for several years – as I become more and more frustrated with SPSS, R becomes a more and more attractive alternative, and the switching cost becomes something I am more and more willing to pay.

So, as I stand there in front of the class, in 2016 instead of 2014, I have an extra two years of prompting by this pair of perennial companions at my back. The accumulated experience manifests as two little voices – one questioning me with “what about Bayes? what about Bayes?” when I talk about distributions and significance testing; a second niggling at me to “mention R! mention R” when the discussion turns to data analysis.

Luckily (?), we’re not really getting into the maths this semester, nor are we going anywhere near SPSS. Luckily (?), I have this blog, so I can offer off-the-syllabus and unsolicited advice here instead.

…

Unluckily, I just spent far too long browsing some random blogs, and now I need sleep.

So my advice shall be short:

1. Learn about Bayes, because this and this.

2. Learn about R, so you’re not stuck with SPSS later.