Naked Science Forum

Life Sciences => Physiology & Medicine => Topic started by: jon_bondy on 13/05/2020 15:34:15

Title: How do we distinguish causation from correlation?
Post by: jon_bondy on 13/05/2020 15:34:15
In this podcast

[I tried to post a direct link to a Naked Scientist podcast, but was told that I cannot post "external links".  So, you will have to do the Google search yourself]

Brain development - decisions, decisions...

the assertion is made that "During the pandemic lots of us will be playing more video games than usual. So it's good to know that we can do so guilt-free because we are really just developing our cognitive skills"

I find this statement to be very difficult to defend. 

It presumes that highly skilled gamers became that way because of practice, rather than any other possible explanation.  In fact, there are other explanations that seem equally likely.  If I am not good at making rapid decisions, it is likely that I will find rapid-fire games to be frustrating, and I would quite likely avoid them.  This would result in a selection bias for those with quick responses being drawn preferentially to such games.  The statistics quoted in the podcast could well be due to this effect rather than a training effect.

My point is that excellent gamers need not have learned those skills; they could have already had them from the start.

A careful analysis of cause vs effect is essential in avoiding claims that sound plausible but need not be true.
Title: Re: How do we distinguish causation from correlation?
Post by: alancalverd on 13/05/2020 16:56:33
Let's segment the question.

Playing against a computer is mostly a futile waste of time. So we can discount the intellectual capacity of those who do.

"Playing" casino "games" is a waste of time and money. The operators are not charities. One can ascribe a negative IQ to the "players". 

Playing against another human with an electronic intermediary can be fun, and whether the game is of pure skill (chess) skill and chance (backgammon) or anticipation and speed of response (combat games)  there  is no doubt that anyone who sticks at it will improve, just as with actual face to face games. At the extremes we find people who are appalled by their first experience of cricket, and some who go on to make a very satisfying career from playing and teaching it. There's no doubt that initial aptitude has a lot of influence, but strength and interest won't get you far without a lot of practice to develop  understanding, anticipation and coordination, and I can't see why highly electronic gaming should be any different from  physical wrestling or kicking a ball in those respects.

An early air combat game was banned from my flying club bar in the 1970s after a 12-year-old thrashed our highly decorated Chief Flying Instructor, but everyone is encouraged to hone their skills with a good simulator, and the essence of training whether simulated or real is to move processes and decisions from the conscious to the unconscious brain so that you can cope with a gradually increasing workload  as you progress to more complex machines and unfamiliar environments. Some folk just can't do it, others are "natural pilots" who just absorb the drills without seemingly going through any conscious phase at all. I put myself in the "highly motivated slow learner" category, and I can see how someone without a lot of motivation might never develop the unconscious responses at all. 
Title: Re: How do we distinguish causation from correlation?
Post by: Bored chemist on 13/05/2020 17:29:33
Playing against a computer is mostly a futile waste of time. So we can discount the intellectual capacity of those who do.
The same can be said of those who argue on the internet...
There's no doubt that initial aptitude has a lot of influence, but strength and interest won't get you far without a lot of practice
But that practice won't happen without the initial aptitude.
So it's chicken and egg.
I'm fairly sure that's the OP's point, and they are right.
Title: Re: How do we distinguish causation from correlation?
Post by: evan_au on 13/05/2020 23:32:14
Quote from: OP
the assertion is made that "During the pandemic lots of us will be playing more video games than usual. So it's good to know that we can do so guilt-free because we are really just developing our cognitive skills" (emphasis added)
I wouldn't call it an assertion - more like a joke, as a leadin to a more serious story...
Title: Re: How do we distinguish causation from correlation?
Post by: jon_bondy on 17/05/2020 13:43:35
"I wouldn't call it an assertion - more like a joke, as a leadin to a more serious story."

Perhaps evan_au should have listened to the podcast before commenting.

It was not a leadin, it was a summation.  In a serious science podcast, any statement that conflates causation with correlation has the potential to have the listener take such statements seriously. Perhaps some of us are educated enough to understand that the statement was not true, but many, MANY people are not that careful in their analysis.

Personally, I believe that a retraction is in order, so that it is clear that this statement is unsupported by the facts stated in the podcast.
Title: Re: How do we distinguish causation from correlation?
Post by: jon_bondy on 26/05/2020 15:52:20
I was really excited to hear, on the most recent Naked Neuroscience, that my concerns were addressed directly and completely.  Thanks so much for this!