Russell T. Warne ๐Ÿ‡บ๐Ÿ‡ธ๐Ÿ‡จ๐Ÿ‡ฑ Profile picture
New book IN THE KNOW: DEBUNKING 35 MYTHS ABOUT HUMAN INTELLIGENCE available for pre-order: https://t.co/5BsiHht4tZ @UVU #psychology professor, dad, arts critic
Agustin Profile picture 1 added to My Authors
24 Mar
๐Ÿšจ๐Ÿšจ๐ŸšจNEW ARTICLE!๐Ÿšจ๐Ÿšจ๐Ÿšจ
I coauthored an article in @jegcec1 with @JaZBur where we surveyed teachers and non-teachers about their knowledge of human intelligence. The results were not surprising . . . but still disappointing.
#psychology #edutwitter
doi.org/10.1177/016235โ€ฆ
First the good news. On several fundamental topics, the sample was generally correct. Large majorities of both groups believe:
โœ…Logical thought is part of intelligence
โœ…Men and women have equal avg IQs
โœ…Brain dmg can lower IQ
โœ…Abstract thinking is part of intelligence
For two groups that aren't psychologists, that's a pretty good grasp of the basics about intelligence. ๐Ÿ‘

The results for the non-teachers, only 35.9% of whom have had a college-level #psychology course, are especially encouraging.
Read 11 tweets
14 Oct 19
I just finished reading Shewach et al.'s (in press) meta-analysis of 181 stereotype threat effect sizes (total n = 10,436), published in @APA_Journals's JOURNAL OF APPLIED #PSYCHOLOGY. And it is sobering. 1/
psycnet.apa.org/doi/10.1037/apโ€ฆ
@APA_Journals First, the strongest effect sizes are found in the stereotype threat studies with the least realistic conditions. As the results get more realistic, the ST effect grows weaker. Studies in real-world settings show zero effect. 2/
@APA_Journals Most stereotype threat studies use low-stakes tests; examinees are likely unmotivated. As they get more motivated, the effect drops. Paying people for good scores eliminates ST.

Stereotype threat probably doesn't generalize to high-stakes tests w/motivated examinees. 3/
Read 8 tweets
23 Aug 19
This essay by Turkheimer about the supposed dangers of using polygenic scores in education amounts to: "People might use genetic information to make decisions I don't like. ๐Ÿ˜ญ Therefore, it's an awful technology."

1/ leapsmag.com/the-shiny-and-โ€ฆ
Turkheimer worries about using genetics instead of test scores to assign kids to educational tracks. Given the lockstep education that most kids experience, this is comically unlikely.
2/
Before middle school grades, schools do little differentiation of the educational experience, even when they have data (e.g., test scores) that would justify it. And suddenly, they'll create tracks and differentiated curricula when they suddenly get DNA test results? 3/
Read 9 tweets
10 Aug 19
My next #APA2019 presentation is from David MacKinnon: "Modern Causal Inference and Third-Variable Effects." #psychology
MacKinnon: As the number of variables increases, the number of possible relationships among them increases exponentially.

#APA2019 #psychology
MacKinnon explains the difference between mediators, confounders, and colliders. #APA2019 #psychology

Confounder: a causal variable that needs to be controlled for to understand true X&Y relationship.
Collider: a variable caused by X&Y. Don't adjust from colliders!
Read 5 tweets
21 Jul 19
Just finished reading the big UK mindset training study. It's extremely well designed:
โœ…Randomized control study
โœ…Sample size >4000
โœ…High treatment fidelity
โœ…High statistical power
โœ…Controlled for cluster sampling
educationendowmentfoundation.org.uk/public/files/Pโ€ฆ
And here are the results... Mindset training had ZERO impact on any of the dependent variables:
โŒReading achievement
โŒWriting mechanics
โŒMath achievement
โŒSelf-worth
โŒSelf-efficacy
โŒTest anxiety
โŒSelf-regulation
That 0 of 7 DV's that mindset changed.
You would think that if mindset training were useful, then it would improve SOMETHING. Nope. Mindset theory had 7 chances and struck out each time. #education #edutwitter #psychology
Read 8 tweets
13 Jul 19
Next #ISIR2019 speaker is @eawilloughby's talk entitled "The first longitudinal investigation of cognitive ability in adult adoptive families." #psychology
.@eawilloughby recaps the literature showing conflicting results of prior adoption studies. Some show zero shared environment. But some studies show lingering adoption effects at age 18.
#psychology #ISIR2019
Today @eawilloughby reports preliminary results from a sample of 1200 kids who have been tracked for 15-21 years, since an avg age of 14.9. These "kids" are all 26-40 years old. They're adults & don't have many active parental environmental influences.
#ISIR2019 #psychology
Read 9 tweets
12 Jul 19
Next #ISIR2019 talk is from Gregory Clark comparing family networks and genetics in determining social outcomesn England from 1750- 2019. #ISIR2019 #psychology
Clark: Economic outcomes among relatives are correlated. The big question is whether that is causal (that relatives help out one another) or whether shared genes create these correlations. #ISIR2019 #psychology
Clark: The correlation is stronger the more closely related people are. (Sir Francis Galton showed this in 1869.)

#ISIR2019 #psychology
Read 11 tweets
10 Jul 19
Next #ISIR2019 talk is from James J. Lee. His topic: "Genome-wide association studies of intelligence."

For those of you who don't know, a genome-wide association study (GWAS) identifies correlations between variations in segments of DNA and a phenotype. #psychology
Lee: Once we know that a trait is heritable, it seems reasonable to identify which segments of DNA are associated with a phenotype. #psychology #ISIR2019
Lee: In 2016, Okbay et al. published a GWAS of 293,723 for years of education. Lee et al. (2018) followed up with a sample size of 1.1 million. They found 1,271 SNPs (i.e., DNA variants) that are correlated with educational attainment.

#psychology #ISIR2019
Read 45 tweets
10 Jul 19
Next talk at #ISIR2019 is from Matt McGue, and it's entitled "What behavioral genetics can tell us about the origins of individual differences in human abilities." #psychology
McGue: Correlation is not causation, but with "methodological triangulation," correlation can produce data that can be interpreted as causal. #psychology #ISIR2019
McGue's goals:

โœ…Biometric approach to IQ
โœ…Why there probably isn't "one" heritability for IQ
โœ…Move beyond abstract variance components to identify underlying causal mechanisms

#ISIR2019 #psychology
Read 34 tweets
10 Jul 19
Thomas Bouchard speaking at #ISIR2019 now. His talk is entitled "Finding Out How Things Work."

He mentions Popper's requirement that science must test ideas that are refutable. "But I discovered how resistant psychologists were to refutation and science." #psychology
In graduate school he contested the idea that genetics might influence autism. "It was a bloodbath. No one else in the room wanted to entertain the idea that genes could influence psychopathology." #psychology #ISIR2019
Bouchard discussed the evidence of the time, such as the Tryon (1924) study showing that rats could be bred to select for behavior. "What more evidence do you need that genetics can influence behavior?" #psychology #ISIR2019
Read 39 tweets
25 May 19
Next #APS19DC presentation is from @JoHenrich: "W.E.I.R.D.: How Westerners became psychologically peculiar and particularly prosperous." Henrich is chair of the Dept of Human Evolutionary Biology at Harvard. #psychology
.@JoHenrich says this presentation is the culmination of 13 years of work. He has a paper in review and an upcoming book about the topic. #psychology #APS19DC
.@JoHenrich: Most #psychology studies were based on Westerners, but they are psychologically different from other peoples in the world. He wrote this up in a lit review:
cambridge.org/core/journals/โ€ฆ
#APS19DC
Read 33 tweets
2 Feb 19
Typical problem in #psychology and #education research: Study shows that a program has little/no effect. How do the authors (who are invested in the program) react? They try to salvage it and argue that it really does work. ๐Ÿ™„
Later passages in the article:
"Future studies should examine...moderator variables to better understand with whom and under what conditions Second Step differentially impacts academic outcomes."
"Researchers should continue to explore the impact of Second Step..."
This is a study with the strongest effect size for the intervention was d = .126. Of 28 outcomes, only 4 (14.3%) were statistically significant (p < .05). But it never occurs to the authors that teaching "social emotional learning" doesn't improve academics noticeably.
Read 5 tweets
10 Sep 18
This came in the mail over the weekend. I don't get to read much fiction, so I think I'll enjoy it. #psychology
This whole paragraph (Gardner, 2011, p. xix) is a train wreck of scientific logic. Basically, Gardner says because he doesn't want to be a psychometrician, he doesn't have to gather data to multiple intelligences... and the theory isn't testable anyway.
#psychology
Gardner (2011, p. xxi) thinks the Mozart effect is real, even...
-17 yrs after @timothycbates's failed replication: sciencedirect.com/science/articlโ€ฆ
-11 yrs after Waterhouse's debunking: tandfonline.com/doi/abs/10.120โ€ฆ
-1 yr after a meta showed zero effect: sciencedirect.com/science/articlโ€ฆ
#psychology
Read 10 tweets