I did this test on my 3 children around the time they turned four. For one of them, I thought, as I watched the video, the reason my child couldn't wait the whole 15 minutes was because of feelings of anxiety – "Where has daddy gone, is he still hear?". In that sense, the test, in my interpretation, was more a measure of anxiety or ability to control negative thoughts or emotions. Another child sat easily the 15 minutes, and here I could suspect the test revealed a measure of self control or maybe just a very strong sweet-maximizing-instinct. On the other hand this child has also always been at ease just in its own company and never been bothered or worried if, there wasn't any adults around. Only the presence of spiders is frightening.
Still I do suspect – and fear –. that in all its simplicity the mashmellow test might still have some predictive power for future outcomes, although it might not exactly be self controll which is "revealed" – it might be anxiousness, self reliance, gluttony, intelligence etc.
I am feeling a bit judged by Stuart here. I literally JUST flew from Oakland to New York (so cross-country) on Sunday and listened to the Studies Show the entire time (I fell WAY behind okay). I am not a freak 😭
Okay last part is a lie I am a lil bit of a freak but that's besides the point! There are dozens of us that would listen to this podcast for five hours on a plane! Or at least one, which is one of the two digits in a dozen.
A cynical part of me wonders if this would be a much shorter podcast if it instead covered the big findings in psychology that *do* replicate with similar effect sizes / p-values…
The experimental setup gives rise to other reasons why the bell could be rung than failure to defer gratification . What if the child decides that the adult may never come back unless they ring the bell? There is nothing much to lose by ringing it once they have lost confidence. And how long they wait before that happens will also depend on their perception of time . So we do t really know what, in terms of underlying psychological variable, the time delay measures.
As part of my undergraduate psychology degree I was involved in running an experiment that tried to establish how long people waited before hanging up on a phone call that didn’t connect. At the time some appreciable % of long distance calls would fail to connect and there was a variable delay before connection. There was a trade off to be made between waiting too long (way past the point that a call might still connect) and giving up too early as there was a cost to redialling (in the days of rotary phones). Issues of time perception and how you assessed the likelihood of the call connecting after a delay of x seconds made this rather complicated. I think the marshmallow test suffers from similar confounds.
It would actually be a perfectly valid experiment if perhaps a bit cruel on a 4 year old for the experimenter to never come back until the bell is rung. Depending how sophisticated the 4 yr olds are in their understanding of psychologists tricks they may conclude that quite early on!
Interesting Washington Post article about cultural confounders in the marshmallow test and other executive function instruments: https://wapo.st/3WZ9HRQ
You sort of buried the lede on this. Finding r = .17 for educational attainment, with a seemingly unrelated child behavior test from decades earlier? That seems like a big deal. Sure, it’s interesting that controlling for SES eliminates it. But it’s not like all the causal complexities with SES are nailed down, such that controlling for it means we understand what’s really going on here.
You also talk like there’s no evidence for a trainable skill, but I don’t see how we can know that from these studies. That’s a bigger topic than just the old “willpower as muscle” finding that never replicates. There are other strategies for overcoming temptation than effortful resistance. Clinical psychologists teach people skills for healthier ways of responding to their emotions all the time. I don’t see any reason to assume non-clinical populations can’t learn different responses to cravings such that they achieve more life goals. (Think about therapy for procrastination.)
I did this test on my 3 children around the time they turned four. For one of them, I thought, as I watched the video, the reason my child couldn't wait the whole 15 minutes was because of feelings of anxiety – "Where has daddy gone, is he still hear?". In that sense, the test, in my interpretation, was more a measure of anxiety or ability to control negative thoughts or emotions. Another child sat easily the 15 minutes, and here I could suspect the test revealed a measure of self control or maybe just a very strong sweet-maximizing-instinct. On the other hand this child has also always been at ease just in its own company and never been bothered or worried if, there wasn't any adults around. Only the presence of spiders is frightening.
Still I do suspect – and fear –. that in all its simplicity the mashmellow test might still have some predictive power for future outcomes, although it might not exactly be self controll which is "revealed" – it might be anxiousness, self reliance, gluttony, intelligence etc.
I am feeling a bit judged by Stuart here. I literally JUST flew from Oakland to New York (so cross-country) on Sunday and listened to the Studies Show the entire time (I fell WAY behind okay). I am not a freak 😭
Okay last part is a lie I am a lil bit of a freak but that's besides the point! There are dozens of us that would listen to this podcast for five hours on a plane! Or at least one, which is one of the two digits in a dozen.
A cynical part of me wonders if this would be a much shorter podcast if it instead covered the big findings in psychology that *do* replicate with similar effect sizes / p-values…
The experimental setup gives rise to other reasons why the bell could be rung than failure to defer gratification . What if the child decides that the adult may never come back unless they ring the bell? There is nothing much to lose by ringing it once they have lost confidence. And how long they wait before that happens will also depend on their perception of time . So we do t really know what, in terms of underlying psychological variable, the time delay measures.
As part of my undergraduate psychology degree I was involved in running an experiment that tried to establish how long people waited before hanging up on a phone call that didn’t connect. At the time some appreciable % of long distance calls would fail to connect and there was a variable delay before connection. There was a trade off to be made between waiting too long (way past the point that a call might still connect) and giving up too early as there was a cost to redialling (in the days of rotary phones). Issues of time perception and how you assessed the likelihood of the call connecting after a delay of x seconds made this rather complicated. I think the marshmallow test suffers from similar confounds.
It would actually be a perfectly valid experiment if perhaps a bit cruel on a 4 year old for the experimenter to never come back until the bell is rung. Depending how sophisticated the 4 yr olds are in their understanding of psychologists tricks they may conclude that quite early on!
Interesting Washington Post article about cultural confounders in the marshmallow test and other executive function instruments: https://wapo.st/3WZ9HRQ
You sort of buried the lede on this. Finding r = .17 for educational attainment, with a seemingly unrelated child behavior test from decades earlier? That seems like a big deal. Sure, it’s interesting that controlling for SES eliminates it. But it’s not like all the causal complexities with SES are nailed down, such that controlling for it means we understand what’s really going on here.
You also talk like there’s no evidence for a trainable skill, but I don’t see how we can know that from these studies. That’s a bigger topic than just the old “willpower as muscle” finding that never replicates. There are other strategies for overcoming temptation than effortful resistance. Clinical psychologists teach people skills for healthier ways of responding to their emotions all the time. I don’t see any reason to assume non-clinical populations can’t learn different responses to cravings such that they achieve more life goals. (Think about therapy for procrastination.)