Reposted by Drew BaileyAndy de Barros @andydebarros.bsky.social · 25/03/2026New paper: Morocco's Pioneer Middle Schools—a government-led whole-school reform—improved socioemotional skills, tripled year-on-year learning, reduced dropout ~1/3. "Beyond Basics" w/ Campos Quintero, El Amrani Mida, Glewwe, Kumar, and Lépine de-barros.com/publication/... @cega-uc.bsky.social 0177
Reposted by Drew BaileyGuido Imbens @imbens.bsky.social · 21/03/2026Don’t Do Difference In Differences (DDDID), cheers, guido 37714
Reposted by Drew BaileyMegan Stevenson @meganstevenson.bsky.social · 11/03/2026Excited to share a new paper with @jfischman, just accepted at JEL. We argue that empirical research tends to be biased and overconfident due to a weakness in the dominant econometric framework: insufficient attention paid to humans “in the loop” with the research process. 1/ 14914
Drew Bailey @drewhalbailey.bsky.social · 18/02/2026In my experience, striving to carve nature at its joints using Likert scales is a common phase that curious and creative students go through. Maybe this reading should go at the end of every course unit on clustering? 010
Reposted by Drew BaileyAnnenberg Institute @annenberginstitute.bsky.social · 18/02/2026📢 #EdWorkingPapers: Tyler Watts, @emmarosehart.bsky.social, & @drewhalbailey.bsky.social analyze 87 RCTs & find that fadeout is common across most programs. Intervention characteristics explain only a small share of differences in persistence. 📄 bit.ly/4aGrXY4 054
Reposted by Drew BaileyEmma Hart @emmarosehart.bsky.social · 12/02/2026Why do educational intervention impacts fade? Isn't catch-up a good thing? Are sleeper effects real? Does fadeout mean failure? @drewhalbailey.bsky.social, Tyler Watts, and I address these questions & more in an EdNext piece & 4 new working papers! www.educationnext.org/why-do-most-...educationnext.orgWhy Do Most Education Interventions Fade Out Over Time?There is evidence both to explain and complicate the “fadeout effect” 1124
Drew Bailey @drewhalbailey.bsky.social · 25/01/2026Replying to @bwroberts.bsky.social on cross-lagged panel models open.substack.com/pub/drewhalb...open.substack.comReplying to Brent on cross lagged panel modelsLast week, Brent Roberts blogged on “The Inconceivability of the CLPM and RI-CLPM”. 1112
Drew Bailey @drewhalbailey.bsky.social · 21/01/2026Person at talk translating all model descriptions into formal equations is the academic talk version of the person filling out the scorecard at a baseball game. 030
Reposted by Drew BaileyMichael Clemens @mclem.org · 04/12/2025This fetid landmark, this historical stain on humanity is primarily due to the stunningly reckless obliteration of America's foreign assistance agency earlier this year. Led by the richest man on earth. In secret, on a weekend. With zero analysis or discussion of its catastrophic impacts.wsj.comFor First Time in Decades, Child Deaths Will Rise This YearAlmost a quarter of a million more children around the world are projected to die in 2025 than in 2024. 41391748
Reposted by Drew BaileyRuben C. Arslan @ruben.the100.ci · 03/12/2025Fadeout of cognitive training remains one of the more replicable findings in psychology in this preregistered study of 300 preschool children. Well-done study with a 4 year follow up. The language gains either faded or the control group caught up. journals.sagepub.com/doi/full/10.... 24619
Drew Bailey @drewhalbailey.bsky.social · 17/11/2025I think the “why most published research findings are false” paper suggests this as one of the heuristics for identifying “false” findings. 120
Drew Bailey @drewhalbailey.bsky.social · 03/10/2025So hard! When there is no cross-lagged effect in this data generating model, RI-CLPM estimates one, but when there *is* one effect, the ARTS model doesn't! Not sure the paper bears much on whether cross-lagged effects are rare, but def on our ability to use these models without external info. 000
Reposted by Drew BaileyRich Lucas @richlucas.bsky.social · 19/09/2025Interested in models used to estimate lagged effects in panel data? We (@rebiweidmann.bsky.social, Hyewon Yang) have a new paper looking at patterns of stability and their implications for bias and model choice: osf.io/preprints/ps... [1/x]osf.ioOSF 12512
Drew Bailey @drewhalbailey.bsky.social · 09/09/2025I really like this paper dealing with the problem of “mischievous” responding in longitudinal panel data, by @joecimpian.bsky.social journals.sagepub.com/doi/full/10....journals.sagepub.comSage Journals: Discover world-class researchSubscription and open access journals from Sage, the world's leading independent academic publisher. 120
Drew Bailey @drewhalbailey.bsky.social · 25/08/2025Dag makhani: Causal inference and Indian cuisine 020
Drew Bailey @drewhalbailey.bsky.social · 19/08/2025Like, the effect of dropping a bouncing ball on the velocity of the ball over time is a weird oscillating function? 010
Drew Bailey @drewhalbailey.bsky.social · 18/08/2025About 2/3 of the posts on this platform linking to the recent NYT article on null findings from Baby’s First Years have this reaction. You can search the headline and verify yourself! www.nytimes.com/2025/07/28/u...nytimes.comStudy May Undercut Idea That Cash Payments to Poor Families Help Child Development 030
Drew Bailey @drewhalbailey.bsky.social · 09/08/2025I used Paige’s first book in a class students with a wide range of previous exposure to and attitudes about behavior genetics, and they all seem to find it very interesting. Will probably try this one too! 000
Drew Bailey @drewhalbailey.bsky.social · 25/07/2025Random Intercepts and Slopes in Longitudinal Models: When Are They "Good" and "Bad" Controls? or Illusory Traits 2: Revenge of the Slopes Led by Siling Guo, with Nicolas Hübner, Steffen Zitzmann, Martin Hecht, and Kou Murayama. Comments welcome! osf.io/preprints/ps...osf.ioOSF 0135
Reposted by Drew BaileyJulia M. Rohrer @dingdingpeng.the100.ci · 25/06/2025New blog post! Let's say you've measured two variables repeatedly and want to investigate how one affects the other over time. Here are some recommendations for how to do that well. www.the100.ci/2025/06/25/r...the100.ciReviewer notes: So you’re interested in “lagged effects.”In some fields, researchers who end up with time series of two variables of interest (X and Y) like to analyze (reciprocal) lagged effects between them. Does X affect Y at a later point in time, and d... 1421060
Drew Bailey @drewhalbailey.bsky.social · 23/06/2025Although field-specific authorship norms probably mostly just reflect the values of people in the field, I also think they can affect those values too. This seems like a good example! (I have some guesses about unintended consequences of tiny authorship teams too, btw.) 010
Reposted by Drew BaileyKevin M. King @kevinmking.bsky.social · 20/06/20256) LCGAs never replicate across datasets or in the same dataset. They usually just produce the salsa pattern (Hi/med/low) or the cats cradle (Hi/low/increasing/decreasing). This has misled entire fields (see all of George Bonnano's work on resilience, for example). psycnet.apa.org/fulltext/201...psycnet.apa.orgAPA PsycNet 6294
Reposted by Drew BaileyRuben C. Arslan @ruben.the100.ci · 13/06/2025Our paper "A fragmented field" has just been accepted at AMPPS. We find it's not just you, psychology is really getting more confusing (construct and measure fragmentation is rising). We updated the preprint with the (substantial) revision, please check it out. osf.io/preprints/ps... 115461
Drew Bailey @drewhalbailey.bsky.social · 11/06/2025But I really hope we get 10 more years of strong studies now on the effects of large increases in access on outcomes for "always takers" and especially for elite students. There are lots of good reasons to expect these effects should differ. (2/2) 010
Drew Bailey @drewhalbailey.bsky.social · 11/06/2025I have seen lots of higher ed talks and papers in the last 10 years convincingly demonstrating that just making some cutoff (getting into a more selective college or major, not taking remedial classes) helps the marginal student. Great to see an emerging consensus. (1/2) 100
Drew Bailey @drewhalbailey.bsky.social · 11/06/2025For every cause, x, there is some group of people (often disproportionately people who study x) who think the effects of x are way bigger than they are. Therefore, I think we are doomed to read (or worse, make) "Yeah, but the effect of x is small" takes forever. 000
Reposted by Drew BaileyPaul Bogdan @pbogdan.bsky.social · 09/04/2025I investigated how often papers' significant (p < .05) results are fragile (.01 ≤ p < .05) p-values. An excess of such p-values suggests low odds of replicability. From 2004-2024, the rates of fragile p-values have gone down precipitously across every psychology discipline (!) 141
Drew Bailey @drewhalbailey.bsky.social · 15/05/2025Hope to see at least one of these in each APS policy brief from now on! 010
Reposted by Drew BaileyCEC Division for Research (CEC-DR) @cec-dr.bsky.social · 12/05/2025IN MEMORY OF LYNN FUCHS The field of special education lost a visionary and beloved leader with the passing of Lynn Fuchs on May 7, 2025. Her absence leaves a profound void—not only in our scholarly community, but in the hearts of all who had the privilege of knowing her. [ click reading below ]conta.ccIn Memory of Lynn FuchsEmail from CEC Division for Research May 12, 2025 In Memory of Lynn Fuchs The field of special education lost a visionary and beloved leader with the passing of Lynn Fuchs on May 7, 2025. Her absenc 0219
Reposted by Drew BaileyJulia M. Rohrer @dingdingpeng.the100.ci · 11/05/2025Thanks to everybody who chimed in! I arrived at the conclusion that (1) there's a lot of interesting stuff about interactions and (2) the figure I was looking for does not exist. So, I made it myself! Here's a simple illustration of how to control for confounding in interactions:> 651124273
Drew Bailey @drewhalbailey.bsky.social · 08/05/2025(Not saying the public is right necessarily; you can get programs that pass a cost-benefit test with much smaller effects on test scores than laypeople want. But it is a problem for policymakers that the public wants them policy to deliver unrealistically sized effects.) 010
Drew Bailey @drewhalbailey.bsky.social · 08/05/2025If you ask people what kinds of effects they’d need to decide to implement something new, they’re much bigger than realistically sized effects in ed policy. We’ve decided collectively to pretend this isn’t a problem and then get surprised at the backlash when it comes. 120
Drew Bailey @drewhalbailey.bsky.social · 03/05/2025Is there a name for the fallacy that, because things are different from each other, one cannot compare them? (If not, I propose the “apples and oranges fallacy”) @stefanschubert.bsky.social 020
Reposted by Drew BaileyJulia M. Rohrer @dingdingpeng.the100.ci · 29/04/2025Starting to feel like "don't look at the coefficients, just calculate whatever metric is relevant to your research question" is a highly underappreciated stats hack and also I may have to get myself a marginaleffects T-shirt. 66610
Drew Bailey @drewhalbailey.bsky.social · 22/04/2025And you can think of the RI-CLPM as doing something like this too, using repeated measures of the same x over time. 110
Drew Bailey @drewhalbailey.bsky.social · 22/04/2025Not eloquently. But in the appendix of this paper, we show that a "multivariate intercept" model that does this (constraining all loadings to equality) reproduces patterns of causal impacts of some RCTs better than OLS (see Table S4 + Fig S1): pmc.ncbi.nlm.nih.gov/articles/PMC...pmc.ncbi.nlm.nih.govTriangulating on Developmental Models with a Combination of Experimental and Non-Experimental EstimatesPlausible competing developmental models show similar or identical structural equation modeling (SEM) model fit indices, despite making very different causal predictions. One way to help address this ... 140
Drew Bailey @drewhalbailey.bsky.social · 22/04/2025Do one for when people realize the extracted factor might be more useful as a *control* for estimating the effects of interest than as the key predictor of interest. 120
Drew Bailey @drewhalbailey.bsky.social · 18/04/2025You like good music and are in North Carolina: are you into Wednesday? 100
Drew Bailey @drewhalbailey.bsky.social · 03/04/2025Ah got it, thanks. In this case, I guess I agree the link between theory and these statistics is often squishy! 010
Drew Bailey @drewhalbailey.bsky.social · 02/04/2025I think that's the way some people talk about types of validity and reliability. But I view (threats to) validity typologies as compatible with estimands: threats to validity are ways that mapping between estimates and estimands can go wrong! 120
Reposted by Drew BaileyKlint Kanopka @klint.bsky.social · 28/03/2025Incredibly excited to have this finally come out! Model evaluation should be about comparisons, so we have a metric that puts comparisons in predictive performance on a common scale. I can’t make a thread about this better than @crahal.com, so I’ll let him take it away. 042
