Data Colada
@datacolada.bsky.social
Announcing posts for our blog. This account is seldom monitored.
Answering the most popular question in a PhD methods course datacolada.org/133
[133] Heterofriendly: The Intuition for Why You Always Need Robust Standard Errors - Data Colada
When I taught my first PhD-level methods course, I invited students to submit questions about any topic in statistics or methodology. Six out of 10 students asked about the same topic: robust & cluste...
datacolada.org
A more user friendly t-test regression variable description frequency plots, and more. datacolada.org/132
[130] Researchbox: Even easier to use than before. More transparently permanent. datacolada.org/130
[130] ResearchBox: Even Easier to Use and More Transparently Permanent than Before - Data Colada
Over the past 10 years or so, posting data, code, and materials for published papers has gone from eccentric to mundane. There are a few platforms that enable sharing research files, including Researc...
datacolada.org
If you saw this 3% of valuations within $1 of mean 58% of valuations within $1 of median Which would be more representative of the data: the mean or the median? datacolada.org/125?ab=1
[125] "Complexity" 2: Don't be mean to the median - Data Colada
In Colada[124] I summarized a co-authored critique (with Banki, Walatka and Wu) of a recent AER paper that proposed risk preferences reflect 'complexity' rather than preferences a-la Prospect Theory. ...
datacolada.org
Measurement error, Nobel Prize winning research, friendly disagreements, and more.... datacolada.org/124
[124] "Complexity": 75% of participants missed comprehension questions in AER paper critiquing Prospect Theory - Data Colada
Kahneman and Tversky’s (1979) “Prospect Theory” article is the most cited paper in the history of economics, and it won Kahneman the Nobel Prize in 2002. Among other things, it predicts that people ar...
datacolada.org
Delving into a disagreement with prominent political scientists on a fundamental data-analysis question: what do you want to learn from a regression interaction? datacolada.org/123
[123] Dear Political Scientists: The binning estimator violates ceteris paribus - Data Colada
This post delves into a disagreement I have with three prominent political scientists, Jens Hainmueller, Jonathan Mummolo, and Yiqing Xu (HMX), on a fundamental methodological question: how to analyze...
datacolada.org
Did a police training improve officers' behavior during arrests? A forthcoming QJE paper with a flexible pre-registration and key p-values of p=.048 & p=.051 says yes. datacolada.org/122
[122] Arresting Flexibility: A QJE field experiment on police behavior with about 40 outcome variables - Data Colada
A forthcoming paper in the Quarterly Journal of Economics (QJE), "A Cognitive View of Policing" (htm), reports results from a field experiment showing that teaching police officers to "consider differ...
datacolada.org
A popular tool in political science for dealing with nonlinearities in invalidated by... ...nonlinearities. datacolada.org/121
[121] Dear Political Scientists: Don't Bin, GAM Instead - Data Colada
There is a 2019 paper, in the journal Political Analysis (htm), with over 1000 Google cites, titled "How Much Should We Trust Estimates from Multiplicative Interaction Models? Simple Tools to Improve ...
datacolada.org
Want to know how many people show an effect in a between-subjects study? Consider this off-label use of the Kolmogorov-Smirnov test datacolada.org/120
[120] Off-Label Smirnov: How Many Subjects Show an Effect in Between-Subjects Experiments? - Data Colada
There is a classic statistical test known as the Kolmogorov-Smirnov (KS) test (Wikipedia). This post is about an off-label use of the KS-test that I don’t think people know about (not even Kolmogorov ...
datacolada.org
Hidden Confound in Pre-Registrations Critique datacolada.org/119 Post ends with link to submit suggestions for AsPredicted & OSF
[119] A Hidden Confound in a Psych Methods Pre‑registrations Critique - Data Colada
A forthcoming paper in Psych Methods (.pdf) had a set of coders evaluate 300 pre-registrations in terms of how informative they were about several study attributes (e.g., hypotheses, analysis, DVs). T...
datacolada.org
Colada[115] shows the prevalence of pre-registrations in papers published in 2022, with lessons to improve in 2023+ datacolada.org/115
