Jason Collins

I have read the paper[1], and its discussion section supports the conclusions drawn by other readers, who are smarter than I am[2][3][4].

Here's one excerpt from the discussion:

> "When considering how our results might inform intervention development, recall that models with con-trols for concurrent measures of cognitive skills and behavior reduced the association between delay of gratification and age-15 achievement to nearly zero."

To put it another way, when controls for education and socioeconomic status are added, there is no significant effect.

1. https://sci-hub.se/10.1177/0956797618761661

2. https://jasoncollins.blog/2018/05/31/the-marshmallow-test-held-up-ok/

3. https://www.theatlantic.com/family/archive/2018/06/marshmallow-test/561779/

4. https://www.vox.com/science-and-health/2018/6/6/17413000/marshmallow-test-replication-mischel-psychology

The experiment is part of a larger research program to study delayed gratification. It was never intended to help parents raise their children. Asking for immediate applications is an odd standard for basic science research. (Unless you reject the value of basic science research entirely.)

Further, the result in a recent replication [1] was a correlation of .28 between the time to ring the bell (to get the marshmallow) and academic achievement. That's not exactly the trivial "4% less likely" effect in your caricature.

(You may also wish to double check your spelling of "marshmallow.")

[1]: https://www.jasoncollins.blog/posts/the-marshmallow-test-held-up-ok

Here are a few examples: https://jasoncollins.blog/2016/06/29/re-reading-kahnemans-thinking-fast-and-slow/

I would just avoid factoids from psychology altogether for the next few years while the researchers in that field sort everything out. https://www.theatlantic.com/science/archive/2018/11/psychologys-replication-crisis-real/576223/

Worse, some of the famous studies aren't merely underpowered or based on nonrepresentative samples (like college kids) or sloppily executed — some famous studies are outright fraudulent. https://www.vox.com/2018/6/13/17449118/stanford-prison-experiment-fraud-psychology-replication

This is somewhat eye opening to me, in the same way that Miller and Sunurjo's work was.

> In other words, selecting which part of the data to analyze based on information regarding where streaks are located within the data, restricts your choice, and changes the odds.[0]

I'm not particularly great at statistics, so I fall trap to a lot of these kinds of misconceptions. I appreciate articles like this getting shared, it's incredibly interesting.

[0]: https://jasoncollins.blog/2018/06/28/explaining-the-hoACt-hand-fallacy-fallacy/

Thinking, Fast and Slow is still worth it, but anyone thinking of reading it should probably be aware that it's taken a pretty hard hit in the psychology replication crisis. A lot (the majority, at least at one point) of its cited studies have failed to replicate. They may still be true (I think Kahneman has said that he believes most of them will bear out eventually) but worst case there may be entire chapters that are basically no good.

https://www.jasoncollins.blog/posts/re-reading-kahnemans-thinking-fast-and-slow

And you're well served with this response, as the actual work he did is basically not reproducible. No matter how many best selling books made into movies; it's still bullshit.

https://replicationindex.com/category/thinking-fast-and-slow/

https://slate.com/technology/2016/12/kahneman-and-tversky-researched-the-science-of-error-and-still-made-errors.html

https://jasoncollins.blog/2016/06/29/re-reading-kahnemans-thinking-fast-and-slow/

I read it on advice of a good pal, and it sure seemed like bullshit to me.