Ben Anderstone

Your ghost is a light show at night on the Grand Coulee Dam

[Ben Anderstone @BAnderstone]

@jbarro Correlation between conscientiousness and IQ is pretty dang weak, though (and slightly negative when self-reported conscientiousness is used.) https://t.co/fjZhGSDgTd

[Ben Anderstone @BAnderstone]

@infraa_ @ProducerCities Closure of the California Correctional Center at Susanville in 2023

[Ben Anderstone @BAnderstone]

@mrandystevens @loganb Weirdly, that's because Clinton Bliss got 25% there

[Ben Anderstone @BAnderstone]

@davidshor @SidKhurana3607 Can't even imagine the swing in a modeled electorate with 100% recent immigrant history. It hurts too much

[Ben Anderstone @BAnderstone]

@gmiller @GarethCanada @NickTerhaar1989 @eyeslasho @CounterTyrannyX That's not the analogy. The analogy is that they demand something from the bank through some level of coercion or implied violence, not merely protest. You're telling us that charging them with attempted robbery would be "retarded"?

[Ben Anderstone @BAnderstone]

@GarethCanada @NickTerhaar1989 @eyeslasho @gmiller @CounterTyrannyX I'd prefer "riot" too, but these reasons it "can't" be an attempted coup don't make sense. Obviously, many participants had that intent, even if they were childish idiots. Imagine if someone used this logic ("they left their guns at home") about a bank robbery! Silly, right?

[Ben Anderstone @BAnderstone]

@gmiller @eyeslasho Are you saying that the word "attempt" means something else besides "try to achieve an outcome"? It doesn't. Or are you saying you categorically oppose prosecuting adults who commit childishly unrealistic criminal attempts? ("Your honor, I was only LARPing!")

[Ben Anderstone @BAnderstone]

@gmiller @eyeslasho That doesn't really scan logically. You can't "attempt" to do something if you lack the realistic capacity to successfully complete it? Gonna be news to prosecutors.

[Ben Anderstone @BAnderstone]

@MiguelMadeira11 @batterystaple42 @GarettJones Construct validity tells us that a test truly measures what it’s supposed to. Without it, we'd mistake any correlation as a meaningful factor, when it could just be a random coincidence or reflect unrelated traits. Thanks for being so intellectually honest & chill about this btw!

[Ben Anderstone @BAnderstone]

@MiguelMadeira11 @batterystaple42 @GarettJones Big question, but statistical tests inc: convergent (correlation w/ related measures), discriminant (low correlation w/ unrelated traits), predictive (predicting outcomes), and factor analysis (structure fit). Plus, test-retest reliability, i.e.: do same people get same results?.

[Ben Anderstone @BAnderstone]

@MiguelMadeira11 @batterystaple42 @GarettJones 0.72 is definitely high for psych research, but it means Intuition only accounts for like 51% of Openness variation. That's miles away from the claim that they're comparably valid, let alone virtually identical.

[Ben Anderstone @BAnderstone]

@MiguelMadeira11 @batterystaple42 @GarettJones Interesting thought! But, factor analysis just clusters items; it doesn’t prove constructs are valid. MBTI overlaps with Big Five traits, and Cronbach’s alpha shows consistency, not meaning. At the end of the day, MBTI just doesn’t pass scientific rigor like the Big 5 does.

[Ben Anderstone @BAnderstone]

@MiguelMadeira11 @batterystaple42 @GarettJones FYI: As far as I know, Intuition and Openness aren't "strongly" correlated. Slightly to moderately, at most. If you have a source for the claim that MBTI has similar construct validity to the Big 5, I'm open! But, I'm pretty confident that's NOT the scientific consensus on this.

[Ben Anderstone @BAnderstone]

@MiguelMadeira11 @batterystaple42 @GarettJones Bad phrasing, maybe, but my point is that the Sensing/Intuition scale can't be used *consistently* to show things because it has poor construct validity. Are you claiming research doesn't indicate poor construct validity, even when used dimensionally?

[Ben Anderstone @BAnderstone]

@MiguelMadeira11 @batterystaple42 @GarettJones It's a valid criticism - rounding 51% to 100% doesn't help. But the bigger problem with the MBTI is that research suggests a lot of the things it's "testing" (like "sensing versus intuition") fail to consistently test/detect anything meaningful.

[Ben Anderstone @BAnderstone]

@MiguelMadeira11 @batterystaple42 @GarettJones That's not really accurate. Several of the MBTI dimensions don't really hold up to to re-testing or inter-rater reliability. The Big 5 ones do. A dimensional MBTI would not fix that fundamental problem.

[Ben Anderstone @BAnderstone]

WA Presidential results by city/town are also telling. D list has fewer working-class suburbs & Hispanic towns. Bainbridge Island, Mercer Island, and resort towns like White Salmon & Winthrop now top list. (Despite R swing in Native vote, Nespelem was #1 @ 46-2 Harris.) [2/2]

[Ben Anderstone @BAnderstone]

With King County results finalized, a few pearls. First, in Seattle, Broadmoor edged out the C/ID for the Trumpiest neighborhood. Still, the rest of the list shows class realignment. Much of South Seattle is now more R than Laurelhurst or Madison Park. 1/2 #waelex

[Ben Anderstone @BAnderstone]

@lymanstoneky The scripts run on the website's back-end, not for download. Accessing table format might require a membership ($35?), but it's the only website I'm aware of with highly-updated, centralized, national county-level data in table format.