Instagram profiles reveal user depression in new machine learning study

Ad hominem name calling, that is. My name-calling "crybabies" refers to a general set of people that fail to deal with their blues by reading a book or listening to music or doing photography or whatever -- engaging and productive activities that bring pleasure and fulfillment. The fact that so many people of privilege are depressed is a social malaise and I think the treatment can be quite natural and effective. Instead, they think they need a drug and/or a psychoanalyst. THAT I would argue is idiotic, unless the case is seriously advanced or a result of trauma.

Hit and run insult? Come back and argue your point, or else apologize.

You are deflecting, shifting attention to me and refusing to debate the accusation you made. Come on, tell me why you think I am a callous idiot. Be specific and be clear.

Indifference to the suffering of others is the very definition of "callous".

The idea that comparative material advantage is a valid predictor for depression; casting the seeking of "attention and palliatives" as a strategy worthy of derision; proposing innocuous activities as an adequate strategy to deal with a serious illness; the post hoc construction of a self-serving strawman in the form of "...a general set of people that fail to deal with their blues..."; the artificial separation of illness and "social malaise" - hyper-competitive societies where your net social worth is tied to your material status, where support is discouraged and where those that struggle are referred to as "crybabies", will inevitably produce more sick people; announcing "I happen to think" when discussing an illness you evidently lack any personal or clinical authority to comment on - the preceding are all things that mark you as an "idiot" .

Now that's the kind of reasoned debate that is worth posting. I find every one of the points worthy of a response. However, your expressing an opinion doesn't give you the right to hurl insults, so knock it off.

Suffering comes in many forms, but equating the suffering of privileged folks with that of masses of deprived people is specious because the former have all manner of comforts, support structures, and venues to alleviate this. Yes, I do think people in Western societies especially are spoiled, and their suffering is greatly overstated.

Photography is an innocuous activity? Who is being derisive here? Certainly not I, who propose engagement in learning and productive and artistic endeavors as an alternative to drugs. I did not offer this as a strategy to deal with serious illness, nor did the article suggest that people posting dark photos on Instagram are afflicted with serious illness.

(continued)

The extraordinary increase in cases of depression in recent years suggests that most are not serious. That is not a strawman nor is it artificial, but rather it is an assessment (in my opinion) of this sudden rash of "illness" that people claim to have. How do you account for it? All you do is criticize -- you don't have answers.

This is not a venue for professional psychologists. So, you evidently think you possess greater "clinical authority" to assess otherwise? Does that kind of authority bring wisdom and solutions? Hardly, when you look at all the misdiagnoses, the ineffectual pharmaceutical solutions, and the continued malaise.

(continued)

If I were a practitioner, who's to say whether I might have better success than the pill-pushers in urging people to get out of their self obsession? Who's to say that the existing treatment regime by the mainstream medical profession is not malpractice? That's what this study was about -- pharma looking for a new market to enhance their obscene profits. You and others bought into it as a way to help sufferers.

The sample quality has to be significant enough to represent the population otherwise it is not accurate.

This is actually what I meant with sample quality - it should be representative of the population. Then a sample of 166 could be ok. I don't know how thorough they were in assembling their sample.

All proof of principle studies have to start small. On the other hand that creates a lot of interesting initial results that never turn into anything useful.

"sampled 166 individual Instagram users, out of +400 Million"

For the study to make sense or have any credibility they must have an accurate assessment of the participants otherwise how would they know if anyone is actually clinically depressed? Obviously there are no credible assessments of 400mil cases.

It was likely based on a different study that had amassed a larger pool of psychologically assessed individuals. They probably examined web pages for a willing subset. As with all such studies there will be limits to the statistics, which should be well documented in the study papers that no one here will ever bother to find, much less read or study ("gut" opinions are so much easier).

The selection criteria are in the paper.

https://arxiv.org/ftp/arxiv/papers/1608/1608.03282.pdf

"Qualified participants were asked to share their Instagram usernames and history." one had to qualify to be randomly tested

Thank you very much for your kind answers. I did not expect that.

What about the Apple "Lightning to SD Card reader", which works with iPhones?

I do not understand why you assume I am "going insane" or being a "layman." Have you read the study itself? I am reading it now, and find the study to be somewhat troubling. I went on to see the credential of the authors. They're mathematicians and statisticians.

They draw the conclusion based on popular believes. I am listing them some I find logically "oversimplified." (possibly because they do no have medical background)

1. Colorful = cheerful = happy
2. Cold spectrum of color = gloomy
3. Darker color = sad
4. Unhappy individual = depressed individuals

Some of these have positive correlation, but using these popular beliefs as premises of study, if not understood in its contextual sense, can result strange results, as we see in the study's methodology.

DSM (IV-TR or V, take your pick) specifically requires a specific numbers of criteria must be met, and authors picked only "sadness" to identify depression. And link it "questionably" to color perception.

If you learned some medical sciences about how a person perceive color, you'd know it's actually "negatively correlated" with your mood. When your mood is tense, in sympathetic mode (aka, fight or flight), your rods and cones in your retina actually relax, allow keener light and colors. That is the reason why depressed individuals prefer darker colors, because normal color can be too eye-blinding. Knowing this as background, drawing a quick conclusion based on Instagram Hue, Saturation, and Value data is a big jump, don't you think?

Please do read the study and we can discuss it if you would like. This is not the right place for such in depth discussion of a subject most people might not understand.