Stephen Wild
@stephenjwild
Followers
2K
Following
42K
Media
405
Statuses
6K
I try to put straight lines through things but usually fail. Try to be Bayesian when I can. Views my own. RT/like != endorsement. Graduate of YouTube University
Joined February 2018
Apparently everyone has a favorite stats textbook. Spill it, peeps.
69
37
232
@xuyiqing Interesting paper! Thanks for posting. IIUC, your main concern with the GAM approach is that it targets the wrong estimand. If so, I feel that your criticism of the approach is kind of unfair, given that it's easy to target CME w/ GAM. See this notebook:
0
1
8
Happy "Large boulder the size of a small boulder" Day to all who celebrate
Large boulder the size of a small boulder is completely blocking east-bound lane Highway 145 mm78 at Silverpick Rd. Please use caution and watch for emergency vehicles in the area.
0
1
12
See this paper for more, including some ways to address it https://t.co/L8kE8O9ju7
Check out the latest working paper by Prof. @dlmillimet, entitled "Fixed Effects and Causal Inference" available through @iza_bonn. #PonyUp
https://t.co/hAopWhxDQH
1
1
7
Time-Invariant Variables’ Time-Varying Effects: Misinterpretations of the Fixed-Effects Model in Sociological Research https://t.co/ySBKEUOgGA
1
15
82
Looks good. Added to the reading list
Lecture Notes on High Dimensional Linear Regression
0
0
5
🎉Preprint! EMA for mental health: How often should symptoms be measured? Using the Nyquist-Shannon Theorem from signal processing, we propose a theory-driven framework for optimal sampling. Collab with @EikoFried, @mHealth_Lab_KIT , @Andreas__Jansen
0
15
58
Ended up thinking about South Park episodes for no reason today, and Casa Bonita is still one of the best.
1
0
3
Econometrician introducing her new diff-in-diffs method to empirical economists.
7
31
340
Cohen on why sometimes r = .1 may not be small.
Okay, psychologists. When did you collectively decide that an effect size of r = .1 is the cutoff for meaningful effect sizes? Source: https://t.co/k45tbZkwMn
1
1
24
Okay, psychologists. When did you collectively decide that an effect size of r = .1 is the cutoff for meaningful effect sizes? Source: https://t.co/k45tbZkwMn
3
0
15
Chasing Shadows: How Implausible Assumptions Skew Our Understanding of Causal Estimands. Stijn Vansteelandt & Kelly Van Lancker. Statistics in Biopharmaceutical Research.
tandfonline.com
The ICH E9 (R1) addendum on estimands, coupled with recent advancements in causal inference, has prompted a shift toward using model-free treatment effect estimands that are more closely aligned wi...
1
4
16
New blog post on The 100% CI: Writing about technical topics in an accessible manner (by Julia Rohrer)
1
7
30
Looks like this might be worth a careful read.
From LATE to ATE: A Bayesian approach. Isaac M. Opper. Journal of Econometrics.
0
0
1
And another post on causal inference and social media and its effect on mental health. DAGs make a prominent appearance. As always, feedback welcome (especially if you think I am wrong). https://t.co/DfHstla7PV
sjwild.github.io
Stephen Wild’s personal blog about statistics, data science, and the like.
0
8
68
DAGs can be used to understand the mechanics of linear interaction analysis. Easy to see the zero correlation between the first-order and interaction terms after centering (though this is not the reason for centering). See more here: https://t.co/zfRnw50PRL
1
6
44
This looks good. DAGs for interaction terms. https://t.co/bFLyToWW1X
bpspsychub.onlinelibrary.wiley.com
Interaction analysis using linear regression is widely employed in psychology and related fields, yet it often induces confusion among applied researchers and students. This paper aims to address...
1
7
61
Notebook: (Decision, Classification, Regression, Prediction) Trees in Statistics and Machine Learning https://t.co/hbtVSScfm0
0
1
6