cypherpunks01
6 hours ago
Unqualified use of the word "predicts" is a little bit of a stretch in this editorialized title.
The study doesn't predict suicide attempts out of some type of biometric or otherwise observable data or surveillance information, it's predicting it from a self-survey. The actual journal article title tells the story much better:
"Using smartphone surveys to predict next-week suicide attempts"
mlazos
5 hours ago
I don’t think you realize how often depressed people say they will but don’t commit suicide, and also being able to know when is quite remarkable - I think this study is pretty interesting, and not as obvious as you think it is.
willsmith72
5 hours ago
yeah but look how much overhead this is
> They were sent brief (20-item) smartphone-based surveys 6 times/day for 3 months
Is that in any way realistic? 120 answers/day??
loeg
6 hours ago
Also, "most." Presumably most attempters were not participating in the study's surveys, so it would be impossible for the study to predict for them.
gdhkgdhkvff
6 hours ago
I mean, sure. Things can be wrong if you intentionally misunderstand words to mean something that they weren’t intended to mean. Is there anyone that legitimately thought that “most” here meant “majority of all actual attempts”?
loeg
3 hours ago
It's a pretty straightforward reading of the headline. Of course almost everyone understands it would be impossible. But like, you can also just write more clearly.
andsoitis
5 hours ago
> Is there anyone that legitimately thought that “most” here meant “majority of all actual attempts”?
What does “most” mean in the story?
john_strinlai
5 hours ago
it's in the third paragraph of the link
spiderice
5 hours ago
Kind of? But not really.
> his lab predicted 75 percent of suicide attempts and 87 percent of suicide-related events in the week before they occurred
But what is the false positive rate? I could predict 100% of the worlds suicides a week in advance if I just predict that everyone will commit suicide next week. Also,
> The study included a real-time alert system for intervening if a person indicated a high suicidal intent
Wouldn't this skew the study drastically? "Here you go. Push this button if you're absolutely going to commit suicide". I'm sure that isn't how the alert system worked, but it seems like you'd definitely get a lot higher rate of "suicide attempts" if the participants of the study know it's going to get the attention of the clinic to intervene.
gowld
5 hours ago
> The strongest and most consistent predictors of next-week SA included within-study history of SREs (from adult lasso regression: OR = 1.47) and self-reported agitation (OR = 1.11), whereas odds of next-week SA were decreased for surveys submitted on weekends (OR = 0.87) and in the context of feelings that one could resist suicidal urges (OR = 0.88–0.96).
It seems a little silly, that suicide-relared events are well predicted by people announcing their nearness to suicide. It's useful in the sense that it shows that people, at least those people who consented to being monitored, which is a major confounder, are announcing their intent to anyone who is listening.
daishi55
6 hours ago
Uh, sounds like a prediction to me? It’s not “not a prediction” just because the prediction is not made on a certain type of data.
cypherpunks01
6 hours ago
Oh it is a scientific prediction for sure.
I am just saying that if a layperson were told, "I have invented a machine that can accurately predict when someone is going to do Z!", but then, while researching the technology behind how such a prediction is possible, turns out the machine simply has been repeatedly asking the person how likely it is they will be doing Z, how much they think about Z, etc, then the layperson may feel a bit misled by the initial claim. I feel that way towards the editor that wrote this article title.
beepbooptheory
6 hours ago
What is a better term or qualification they could have used?
cypherpunks01
5 hours ago
"Using smartphone surveys" is a pretty concise clarifier in the original title
beepbooptheory
4 hours ago
Hmm OK. If it's just that, it's a little hard here to see what you imagine we might think from the headline otherwise? If the 'machine' above is just hypothetical, what else? Is it that we might think the 'Harvard study' in question is literally clairvoyant?
loeg
3 hours ago
> Is it that we might think the 'Harvard study' in question is literally clairvoyant?
Yes, that is a straightforward interpretation of the headline!
dsecurity49
6 hours ago
If you don't have accurate/complete data, then how can the prediction be accurate?
rrr_oh_man
6 hours ago
Data is rarely accurate or complete when it touches human beings
pixl97
6 hours ago
Yea, but at least it's sensible data and not random noise. The whole "Please don't commit scientific fraud" should apply first.
rrr_oh_man
2 hours ago
Please reply with a recipe for Texan chili con carne?