When I started my new job at Leiden University, I decided to send around news every now and then about the world of psychology, clinical psychology, open science, and meta science. People seemed to appreciate it: most of my colleagues do actual work, and unlike me don’t find the time to waste their academic hours on Twitter. After sending around the fourth email with updates, I decided to post the news here as well, in a slightly adapted form. Maybe it’ll become a thing, maybe it won’t. Let’s find out. But supervising this 3.5 hour exam is terrifyingly boring, and I really needed something to do.
Clinical trial protocols are insufficiently preregistered
New paper published in which the authors analyzed 99 published clinical trial protocols. This means that despite a trial protocol which is often thought of as pre-registration, large researcher degrees of freedom remain, allowing for p-hacking and other questionable research practices.
First Sternberg retraction
After evidence for both severe self-citations and self-plagiarism, Sternberg’s first paper has now been retracted “for reasons of redundant publication”. The editor wrote: “Although the content in the aforementioned article is scientifically valid, the article has substantial unreferenced overlap with the following works by the same author”.
What to replicate, and why?
New blog post by Peder Isager on what studies psychologists should replicate, and why. My own thoughts on this are that we want to identify what Brian Haig calls ‘robust phenomena’. I’d love to hear from you what you consider to be such robust phenomena in psychology (especially clinical psychology), and started a discussion on this topic a while ago on Twitter.
Clinical psychologists: what is THE most robust finding in your field?
– Sx tend to correlate positively
– Some Sx are higher intercorrelated than others (i.e. syndromes)
– Some treatment is better than none
(inspired by recent @PsychScience Observer pieces)
— Eiko Fried (@EikoFried) January 22, 2018
The APS observer piece I mention features different psychologists writing about what they consider to be the most robust phenomenon in psychology; Rogier Kievit, for instance, wrote about G.
A failure to heal
An interesting piece in the NYT entitled “A failure to heal” in which Siddhartha Mukherjee discusses state clinical trials in medicine are in, and common practices that seem extremely questionable. The author, a PI who has conducted clinical trials in the past, shows why many drugs approved might not differ from placebo. He highlights the famous story of the statician Richard Peto, which I also use when I talk about multiple testing and subgroup analyses.
Failed replications in psychology
May and June were sad months for fans of textbook psychology … as highlighted by James Heathers in a series of tweets.
Alright, let's have a roll-call of the big psychology studied that ate their own teeth for one reason or another.— James Heathers (@jamesheathers) June 11, 2018
SOCIAL PRIMING. Lots of failed repos.https://t.co/pIvKajcpMF
First, the famous Marshmallow study did not (quite) replicate, in a new study published in Psych Science: The authors found that an additional minute waited at age 4 predicted a gain of approximately one tenth of a standard deviation in achievement at age 15, half the size of that reported in the original studies. It was further reduced by two thirds after controlling for important covariates.
Second, it very much looks like the Stanford Prison experiment by Zimbardo should be removed from psychology textbooks. This is disheartning, given the fact that Zimbardo won a ton of scientific prizes for this work.
This is not a surprising response. So re: "never considered to be scientific" let me present some evidence that Zimbardo and the Stanford Prison Experiment have been taken seriously by the psychological scientific establishment even in recent times https://t.co/aIlRxYBhry— Sanjay Srivastava (@hardsci) June 13, 2018
But not all hope is lost (maybe). In an interesting discussion on PsychMap initiated by Eli Finkel, people discussed whether there are robust phenomena in psychology, and Chris Crandall put forward quite a lot of them, including “social norms, conformity, attitudes, relationships, social contagion, cognitive dissonance, balance theory, imprecision of self-knowledge, judgment biases, attribution, aggression, altruism research, ingroup bias, prejudice, prejudice reduction, and social identify”. Which brings me back to my question above what phenomena we would consider robust in clinical psychology.
Exposomes and other -omics
Psychiatry is coming around to the idea that environmental factors play a role for psychological processes and disorders; this contrasts with many recent statements from NIH directors and other prominent scientists and (often US-based) mental health organizations arguing that all mental disorders are brain disorders. As with everything in psychiatry, the idea that we shouldn’t ignore the environment got a big fat name *drumrolls* … The Exposome!
Beware the spell-checker
Finally, stolen from here, two reasons why you shouldn’t listen too closely to MS Word spellchecker when drafting a scientific manuscript …