This seems appropriate
8 Likes
Won’t somebody think of the children!
No it isn’t; it could mean that Facebook use has a positive correlation with future diminished well being. That doesn’t imply causation. It could imply that there is a root cause that affects both things.
For causation to be involved a mechanism has to be demonstrated.
3 Likes
Ugh. I went and checked, and I haven’t even earned that badge yet. Now I feel very depressed…
6 Likes
4 Likes
So what’s missing here is a mechanism by which the one causes the other? I guess I had assumed the time phased relationship was sufficient. It seems like semantics to suggest there is a root cause that affects one and then the other in sequence, rather than the one causing the other. I guess that’s why I’m not in research or statistics.
No, it says association, which is not causation. It says that it might be that A causes B, or it might be that C causes both A and B. Saying it is an association means that you don’t have enough evidence to state which of those two options represents the actual cause.
3 Likes
Well, a lot of self-medicating exacerbates the condition. It’s not like people who deal with depression by binge eating, drinking, or taking heroin are increasing their “well-being”.
But that’s what I was concerned about with the use of the word “cause”. I doubt facebook causes depression in the way that lighting matches causes fires. That might be senseless criticism because probably nothing causes depression in that strong sense of the word “cause”, but I guess I feel like it’s important to remind everyone of that every time something like this comes up.
That’s not semantics at all. A carelessly tossed cigarette might cause cause the sofa to start to smoke and then cause it to set on fire. The smoke didn’t cause the fire. Correlation isn’t causation even when the correlated things are separated by time. From what they describe, I get the sense they did some valid math to demonstrate that the two things were inter-related, and it was not merely the case that depressed people tended to use facebook more.
But it’s still a big gap between that and saying, “You shouldn’t use facebook because you’ll get depressed” the way we’d say, “You shouldn’t play with asbestos because you’ll get lung cancer.” It doesn’t rule out that various other factors have to be aligned to create the interaction between facebook and depression.
5 Likes
Finding that I was using Facebook (or that they were using me) would certainly cause me to be depressed.
You cannot do maths to demonstrate cause rather than correlation. All mathematics is tautological.
For instance, Kepler’s planetary laws demonstrated an association between various orbital parameters which could be expressed in a mathematical way. But the cause of that association is gravitation, which is “outside” the mathematics. As to what gravitation “really is” that’s another long involved story, but the postulate of a central force between objects can be shown to be entirely consistent with Kepler’s equations.
Now take Tycho Brahe’s geo-centric system. Mathematically the epicycles are an improvement on Copernicus. But the geocentric model, despite giving more accurate results, is wrong because it is contrary to gravitation.
So, in the present case, perhaps I am suffering from an as yet undiagnosed but common illness which results in increasing depression with time. The symptoms cause me to spend more time on Facebook. But, despite the fact that the Facebook use is an indicator of future depression, it is the illness that is the cause of the Facebook use.
This problem, by the way, bedevils a lot of sociological and psychological research.
5 Likes
That’s only because you don’t subscribe to my newsletter.
1 Like
I thought that all the researchers wanted to demonstrate is that you can’t write off the facebook-depression correlation by just saying, “Well, maybe depressed people spend more time on Facebook.” I’m trying to think of a model where depression caused facebook use and facebook use did not cause depression, and somehow facebook use comes out before depression in the data. Maybe it could be an issue where facebook use is just easier to measure and so we think it goes up first when really depression is already on the rise?
Anyway, you are right. They could never use math alone to rule out unknown causes being responsible for the whole thing, but that problem doesn’t really just plague sociology and psychology. It’s true of all science: you make a prediction and then you support it by repeatedly trying and failing to disprove it. So it’s not that we know that facebook causes depression because the facebook usage comes before the decline in well-being, it’s that if depression came before facebook usage in the data we’d know that facebook usage couldn’t account for the depression that came before it, and that didn’t happen.
I am not a Popperian. Popper was a philosopher, not a scientist, and he was trying to resolve what he saw as the problem of induction. Working scientists will tell you science doesn’t really work this way. When Newton said “non fingo hypotheses” I think he was being psychologically correct.
Very often nowadays predictions are made and then evidence is looked for to support them. Popper didn’t really envisage a world in which testability was about standard deviations rather than yes/no. To take a well known example, the search for the Higgs resulted in a “signature” produced by the analysis of a very large number of collision events. Once this reached a certain level of significance, there was a parallel search for alternative explanations that had not been discounted in the original analysis. Popperian falsification is a great oversimplification of what actually happens.
The predictions themselves arise from a theory, and the theory usually arises from observations. So these observations about Facebook are just that; they are prescientific with no theory to support them. If someone comes up with a theory it will need to make testable predictions. But the predictions will be positive (i.e. in these circumstances this will happen). Popper’s attempt to shoehorn in falsification is to confuse, in my view, the quality assessment for the product.
1 Like
And it’s people whom (on some level) you know
Knowing that there exist people who are richer/stronger/thinner/more successful (etc etc) than you is not the same as feeling that way about the people who actual comprise your social circle. And then there’s also the much greater potential for feeling excluded by FB – look at my two friends going to this awesome place and they did not invite me! Seeing people on tv or in films doesn’t really have that effect.
1 Like
I don’t know what distinction you are trying to draw or even what you are trying to distinguish between. I think there is plenty of theory around what increases and decreases human well-being that these observations would be supported by. I think they must have started with a prediction of these results because why else would they have even checked for this correlation? Why test for the ordering of events if they didn’t have a prediction that facebook use would precede depression? I don’t think this was random application of math to a dataset until someone said, “Maybe we could publish this relationship” I think they were looking for it when they set out.
I don’t know about “shoehorning in” falsifiability. Are you saying that there are important scientific claims that could never be disproven by checking them against the real world? I don’t know what that would look like.
So for the attempts to falsify being merely the QA, observing reality and coming up with ideas from that is what everyone does every day, scientist or no, and its what all of humanity did before science. I think that QA process is the only thing that distinguishes science from dumbass guesses (or, worse, economics).
Alcoholism?
6 Likes
It’s the complete opposite for me. To the point i have a friend that checks up on me when i vanish from FB for long periods…
2 Likes
This topic was automatically closed after 5 days. New replies are no longer allowed.
