If you are on HN you would believe that all studies are wrong due to trivial sample issues, but the researchers are usually aware of the issues and try to control for it. For your analogous study nobody would mistake the cause for the effect like that.
"We, therefore, investigated data from nearly 500,000 people from the UK Biobank, with a mean age of 57. People were classified as socially isolated if they were living alone, had social contact less than monthly and participated in social activities less than weekly."
"Our study also included neuroimaging (MRI) data from approximately 32,000 people. This showed that socially isolated people had poorer cognition."
They focused on older people, severely isolated, for whom brain MRI data was available.
If you're not already having some sort of brain problem, you won't be listed in a database where your brain MRI data is available. "People with brain MRI data" is biased toward cognitive problems.
On the other hand even as a passionate programmer I see some form of mental decay in people describing debugging as fun. It can be fun but it probably shouldn't.
Sorry I don’t understand this sentence. Are you saying he is not coping with the findings of the experiment? Did you typo “crap”?
If it’s slang for “You are misinterpreting the facts to support your own biased opinion” it’s not slang I’ve seen before
For instance, I suspect that the raw data which was used by the study could be examined in ways such as: find the top 5% cognitive performers and determine how many of those socially isolate compared to the average.
Also, the study's definition of isolation is rather severe: "People were classified as socially isolated if they were living alone, had social contact less than monthly and participated in social activities less than weekly." The mean age in the data is 57, too; the investigation is focused on those who are getting on in years.
The source of the data would almost certainly have to be self-reporting, rooted in the answers to some questionnaires. A following problem could occur there: some of those who have a full-time job could answer yes to having daily social contact, whereas others could interpret social as anything but work. That would then cause a bias: over-representation of the unemployed in the isolated category. Speaking of work, based on the mean age, there is going to be a bunch of retirees in there too. Retired old people who don't see anyone not doing well mentally is hardly a secret.
I don't suspect any of this applies to people who isolate specifically so they can do something with their brain: study something, or make something. If you're at a party, and people basically like you, but you're thinking that you'd rather be finishing some debugging session that you left half done, this is not about you.
Rather, those who are shunned by others for being dullards.
Imagine they made an analogous study about running, following thousands of participants in distance running events. What would the finding be? "Runners who are separated from others are uncoordinated, inefficient and unfit, and have a higher BMI than those who stick with the pack". But, of course, front-runners are separated from others; there are necessarily far fewer of them in the study, leading to a misleading generalization.