Brief psychology news 11/2019

November 2019 news from Clinical Psychology, Quantitative Psychology (Methods, Measurement), Meta Psychology, Open Science, and Data Visualization. For prior news, see the rubric Psychology News on this blog. If you want to get these via email, you can subscribe on the right side of the blog (on mobile: scroll all the way down). No ads, and I never ever send emails for any other reason.

Clinical Psychology

1. Cristea & Naudet on research waste in psychological treatments: “We describe hype and insufficient consideration of what is known in defining research priorities, persistent risk
   of bias, particularly due to selective outcome reporting, for psychotherapy trials across mental disorders, intellectual and financial biases, […] publication bias, largely inexistent adoption of data sharing, issues of multiplicity and fragmentation of data and findings, and insufficient adoption of reporting guidelines…”
2. Kennis et al. show in a meta-analysis and review of prospective biomarker studies that “there is a lack of evidence for leading biological theories for onset and maintenance of depression”
3. Gardner & Kleinman on “Medicine and the Mind — The Consequences of Psychiatry’s Identity Crisis”. Money quote: “Biologic psychiatry has thus far failed to produce a comprehensive theoretical model of any major psychiatric disorder, any tests that can be used in a clinic to diagnose clearly defined major psychiatric disorders, or any guiding principle for somatic treatments to replace the empirical use of medications. Biologic knowledge is foundational to good psychiatry, but we believe that misapprehension of its limitations is stunting the field [… and] benefit[s] from a falsely simplified and deterministic formulation of mental illness and its treatment”
4. Bai et al. published systematic review and meta-analysis of randomised controlled trials of anti-inflammatory agents for the treatment of major depressive disorder; the effect sizes seem quite strong (which makes me a bit skeptical), but looks like a thorough paper. Comments appreciated, I know too little about the literature to vet the paper properly
5. Meta-analysis in JAMA Psychiatry appears to show that the best “biomarkers” for mental illness are actually behavioral markers … (link to Twitter, cannot open paper from here because Jama Psychiatry blocks the IP of the conference hotel I’m at)
6. Fritz et al. have a new paper on which I got to be a co-author, on network models of resilience in adolescents with and without childhood adversity. Jessica first wrote a preregistered systematic review on resilience factors, and then went on to include the factors that came out of this review in the empirical network study; this is one of the strongest theoretical foundations for including variables in network models in the literature so far
7. Choi et al. show in n~8,000 that physical activity protects against new episodes of depression, and that its benefits apply even in the face of genetic risk for depression.

Quantitative Psychology

1. Preprint: Marsman on the idiographic Ising Model
2. Blog post by Parsons showing that pre-procressing of data dramatically moderates the reliability of the Stroop task #MeasurementSchmeasurement
3. Not psychology, but measurement work shows that the breath-test alcohol analyzers have considerable measurement problems
4. Fantastic special issue in Psychological Methods on “Multi-Study Methods for Building a Cumulative Psychological Science”
5. McNeish et al. fit dynamic measurement models to verbal test score at age ~ 20, and found that they predicted observed scores 50 years later 2.5 times better than using traditional longitudinal IRT
6. I wrote an R tutorial on the community detection method ‘clique percolation’ which allows nodes to be assigned to multiple communities simultaneously (a bit like EFA/CFA with cross-loadings

Meta Psychology

1. Frith plausibly argues that slower science may lead to faster and more robust insights
2. Given the low inter-rater reliability of both scientific manuscripts and grant proposals, researchers and funders have long considered lotteries for grants. It’s happening now
3. Greer & Samuel on “Becoming a Principal Investigator: Designing and Navigating Your Academic Adventure”
4. Jaremka et al. on “Common Academic Experiences No One Talks About: Repeated Rejection, Impostor Syndrome, and Burnout”
5. Barnett et al. examined close to 1 million confidence intervals in PubMed abstracts and found that a disproportionate number were just below the magical 0.05 threshold
6. Smaldino on the importance of good theory in science
7. Preprint: Yarkoni on the Generalizability Crisis in Psychology; he argues that “the inferential statistics we report in psychology papers are so disconnected from the hypotheses they’re meant to test that they may as well be made up”

Open science

1. Preprint: van Sambeck & Lakens show that whether reviewers sign a review or not is related to their recommendation
2. Preprint: van den Akker et al. provide a template and tutorial (including example preregistration) for preregistering secondary data analyses; the tutorial focuses on how to handle prior knowledge of the data but also involves general preregistration advice
3. Preprint: Szollosi et al. argue “Preregistration is redundant, at best”; I like their focus on theory, and learned a lot from these specific authors over the last years, for example the fantastic work of Iris van Rooij on formalizing theory. However, I’m not sure I agree with the main message (i.e. the title). The main benefit of preregistration I see is that it provides some level of transparency. Without transparency, we cannot vet validity evidence. Jessica Flake and I make this point in greater detail in our preprint on how transparency is the first crucial (but not sufficient) step in improving measurement practices in psychology. I thought Paul Smaldino and EJ Wagenmakers had some constructive feedback as well
4. Wiernik put together a bunch of MS Word templates in specific journal styles, helping you to make your preprints real pretty :)
5. Werner on a preregistration mishap, and how it isn’t the end of the world

Data visualization

1. R tutorial on how to make a street map of your favorite city using ggplot2

Paper of the month

The Psych Science Accelerator put online the preprint of their newest study, a collaborative effort of authors who collected data in 130 locations and 48 countries (n~11,500). This may well be the future of how robust empirical psychological research looks like. And there is absolutely no reason we couldn’t strive to do such important multi-lab work in clinical psychology

You can find me on Twitter. And if you cannot access full texts of papers I link to in this blog, I describe a way around the problem here.