Background :

There seems to considerable debate regarding the output of psychological research lacking

a) appropriate, rigourous statistical treatment of data b) replicability

As a second year undergrad psych student, it's one of my biggest fears that my future research might be marred by my lack of knowledge in the above two.

What I am asking for in this post:

I would like reading suggestions in the form of journals, blogs, textbooks or even specific paper/articles that may lead me to other resources. My aim is to develop a sound background in stats/methodology to avoid making embarrassing errors. I feel like it's important to mention my current mathematical level so that suggestions may be easier to make: I'm comfortable with univariate calculus and basics of differential equations, introductory probability theory and high school (done in South Asia) algebra and geometry.


I have read this post on psychology.stackexchange regarding replicability and found it to be helpful.

Please let me know if I must edit this post in order to receive fruitful answers.


1 Answer 1


There are various loosely-defined 'camps' among reform-minded methodologists, so you probably want to try to follow a representative of each.

Andrew Gelman is a prominent landmark in this area, and runs a really nice frequently-updated blog at statmodeling.stat.columbia.edu with frequent paper critiques.

The JASP crew also have a blog, https://www.bayesianspectacles.org/ which is a bit on the aggressive side, but JASP is great and it'll point you to things like the recent 'redefine statistical significance' paper, which will give you plenty to talk about if you run into a methods guru on their coffee break.

You might also enjoy http://daniellakens.blogspot.com/ and the author Daniel Lakens has a Coursera MOOC on this sort of thing. To be sure, there are a ton of good stats MOOCS, just this one happens to be by a psychologist with similar concerns to you :-)

This is a short list but I think it cuts across the camps I'm aware of (I'd love to know more about the camps in methodology myself, maybe some other StackExchange folks have some pointers!) My impression is that at minimum you should try to read someone who solves their problems with parameter estimation and cares about S and M errors (Gelman, Kruschke), someone who solves their problems with model selection and cares about Bayes Factors (EJ Wagenmakers 2018 "Bayes factor design analysis" maybe?), and someone who is happy to be frequentist so long as you're super careful about the interpretation (Lakens, I think? Or, like, most of statistics? Efron's awesome feats of bootstrapping are a great antidote to Wagenmakers' extreme takedowns of everything frequentist).

In my particular corner of psychology, Stan is taking over the world, if that's interesting to you you might get some good mileage out of http://elevanth.org/blog/ I reckon upwards of 20 people recommended the 'rethinking' textbook to me, and they were right! It's pretty great.

  • 1
    $\begingroup$ Should really have mentioned Mayo's "Statistical inference as severe testing" here, but in my defense it had only just come out. Also, definitely check out Navarro 2019 "Between the devil and the deep blue sea" (Computational Brain & Behavior) for some commentary on the relationship between the scientific question and the statistical toolkit. $\endgroup$ Nov 27, 2019 at 18:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.