Live data from Hacker News

Toward Psychoinformatics: Computer Science Meets Psychology (2016)

ncbi.nlm.nih.gov

21–26 of 26 posts

Re: Toward Psychoinformatics: Computer Science Meets Psychology (2016)

#21

Psychoinformatics... I think it just goes to show that computer science is less of a science and more of a tool / method for other disciplines. Yes, you can be scientific about the tool that you're building (and measure it, test it). You can even have scientific theories about the tool that you're using (theory of computation) but to call this psychoinformatics would imply that the informatics part is scientific [1].…

> In my opinion it isn't. Collecting data through the use of big data methods (or whatever you want to call it) has nothing to do with science and everything to do with tool use

There's way more to computation and psychology than just machine learning and big data. Those just happen to be trending presently.

Re: Toward Psychoinformatics: Computer Science Meets Psychology (2016)

#22

If you want to find out more about the project that sparked this paper you can find it here https://menthal.org/ I'm one of the researchers that works on the project.

Any relation to computational cognitive science?

Re: Toward Psychoinformatics: Computer Science Meets Psychology (2016)

#23

Earlier quoted context omitted.

The difference is it's much easier to choose not to walk into a bad psychologists office than it is to completely forsake using the internet, where nearly everywhere is overrun with advertisers tracking your every move and adding it to a psychological profile. Whether you're "giving" them your data or not, they're taking it. Suggesting the average person simply "be careful" about the data they divulge is absurd in li…

> The difference is it's much easier to choose not to walk into a bad psychologists office than it is to completely forsake using the internet I can't agree with this. Abstractly it's the same problem. Requires knowing information you don't have, before you have it.

Can't you read reviews about a psychologist on the internet before you decide whether or not to go into their office?

Re: Toward Psychoinformatics: Computer Science Meets Psychology (2016)

#24

If you want to find out more about the project that sparked this paper you can find it here https://menthal.org/ I'm one of the researchers that works on the project.

Any relation to computational cognitive science?

In my opinion, psychoinformatics is just that. Now we have better tools for data collection and analysis plus outside knowledge from computer science. When we talk to psychologists, they have ideas that have been there for a long time but they did not know how to actually run these experiments. We bring a large scale view into their field. They are usually overwhelmed and we need to guide them extensively. (I've had to specifically craft dataframes into SPSS data file formats in a specific order for them to actually be able to work with.)

You'll see that definitions have evolved over time. Each new wave has a new vision based on the latest research trends and tools.

Re: Toward Psychoinformatics: Computer Science Meets Psychology (2016)

#25

Earlier quoted context omitted.

> The difference is it's much easier to choose not to walk into a bad psychologists office than it is to completely forsake using the internet I can't agree with this. Abstractly it's the same problem. Requires knowing information you don't have, before you have it.

Can't you read reviews about a psychologist on the internet before you decide whether or not to go into their office?

I don't know if I can explain why a review from a patient isn't information that reduces the problem space. To me, it offloads the complexity of the initial problem to another constructed variant.

As analogy, suppose I tell you to only focus on information coming from other corporations about 'big data' and what these organizations are going to do with your data. From that, derive whether their security practices are held at a high standard.

You can't know these things until things proving to the contrary happen. Likewise, I can't know whether my personal experience in a psychologist office is going to be similar or different from another patient - yielding the same results (what is the goal?). Every single statement spoken in the office of a mental health care professional - all of these things matter to the psychologist, and I am aware of that (psychology can't cure neurosis because psychology is neurotic). Every single behavioral mannerism, every quirk, every habit. All 'data' collected, to yield diagnosis. The psychologist themselves might reason themselves out of that belief by telling themselves 'I ignore it, I validate my theories against the science' but the science is mutable because the mind has no set definition that is 'valid' and 'correct' with respect to a standard. They are not God - they do not have omniscience on the definition of mental wellness, mental clarity. I've had counselors tell me it's weird that my 'grounding' techniques are thinking about reality and math. They are not programmers, computer scientists, mathematicians!

It's the same as you worrying about every single statement or bit of information you divulge on the internet. Could be worried about all of it, could shrug it all off. Could go into a psychologist's office and not worry about any of that. Won't, because it fundamentally conflicts with my core values, self identity, principles, beliefs - things I need for my mental health, sanity, security, things that keep my future stable and under my control (being able to reason correctly with respect to code). Conflating all of that with a neurotic obsession over the functioning of my mind? Just as bad, if not worse, as having shit security with data. Data can't predict the future. It's a choice to trick oneself into that belief. That's the problem with these organizations that yields mental peace. Collect all the data they want! It controls their minds more than it controls mine.

Re: Toward Psychoinformatics: Computer Science Meets Psychology (2016)

#26

Earlier quoted context omitted.

Names are just a tool too. The debate is settled for me when I read about Norbert Wiener, Von Neumann, etc.

Hmm, names are tools, I agree. Could you give some sources? I'm curious.

The Dream Machine: J.C.R. Licklider and the Revolution That Made Computing Personal is a good source on Von Neumann, Wiener, Licklider, and more. The book makes computing look like a science.

The names comment is from me. I wish they'd replace "Artificial Intelligence" with "Algorithms++", for example. Either way though, it's not so much the name that matters, unless it's a terrible name... loop.

Post reply on HN