3 ms·

>In an effort to provide a firmer grounding for their arguments, Kaufman and her McLean colleagues used artificial intelligence to develop a model of the connec

>In an effort to provide a firmer grounding for their arguments, Kaufman and her McLean colleagues used artificial intelligence to develop a model of the connections between diverse brain networks that could account for dissociative symptoms. They fed the computer MRI data on 65 women with histories of childhood abuse who had been diagnosed with PTSD, along with their scores on a commonly used inventory of dissociative symptoms. “The computer did the rest,” Kaufman says. Knowing a little of the issues with MRI studies (see dead fish fMRI study, https://www.discovermagazine.com/mind/fmri-gets-slap-in-the-face-with-a-dead-fish https://www.discovermagazine.com/mind/fmri-gets-slap-in-the-... ) and with ML models, and with the replication crisis in psychology, it does not inspire confidence at all that a team who strongly backs dissociative amnesia, feeds a bunch of MRI data into an ML model, and lo and behold after "the computer does the rest" they get the answer they had been hoping for.

Hnrobert42 6y ago
Do we know that they got the “answer they had been hoping for”? A good scientist hopes for nothing more than data that conclusive validates or fails to invalidate the null hypothesis. I appreciate that you admit you have strong biases. That does not mean these scientists also do or that they let their biases get on the way of doing a good job.
- RcouF1uZ4gsC 6y ago
  In science, rarely is the individual scientist neutral. People have built careers and staked reputations on one thing or another being true. This is why you have publication bias, p-hacking, because scientists are ultimately human and they would like to publish something and something that agrees with their position. This is why we do double-blind studies for those studies were the answer is critical (like does a medicine work), because humans can't be trusted to completely disregard their biases. We also pre-register trials, so we don't only get results that the person is happy publishing. The wonder and genius of science, is that there is a process that works in spite of human bias. If instead, we had to depend on the ideal, bias-free scientist, science would become a lot less useful pretty fast. If you want some evidence that scientists are not just neutral reporters of fact, see the story with Susan Fiske's accusation of "methodological terrorism" against people who pointed out math errors. https://www.businessinsider.com/susan-fiske-methodological-terrorism-2016-9 https://www.businessinsider.com/susan-fiske-methodological-t...
  - Hnrobert42 6y ago
    You offer no evidence that these scientists were biases for affected by their bias. You offer only speculation.

Hnrobert42 6y ago

Do we know that they got the “answer they had been hoping for”? A good scientist hopes for nothing more than data that conclusive validates or fails to invalidate the null hypothesis. I appreciate that you admit you have strong biases. That does not mean these scientists also do or that they let their biases get on the way of doing a good job.

RcouF1uZ4gsC 6y ago
In science, rarely is the individual scientist neutral. People have built careers and staked reputations on one thing or another being true. This is why you have publication bias, p-hacking, because scientists are ultimately human and they would like to publish something and something that agrees with their position. This is why we do double-blind studies for those studies were the answer is critical (like does a medicine work), because humans can't be trusted to completely disregard their biases. We also pre-register trials, so we don't only get results that the person is happy publishing. The wonder and genius of science, is that there is a process that works in spite of human bias. If instead, we had to depend on the ideal, bias-free scientist, science would become a lot less useful pretty fast. If you want some evidence that scientists are not just neutral reporters of fact, see the story with Susan Fiske's accusation of "methodological terrorism" against people who pointed out math errors. https://www.businessinsider.com/susan-fiske-methodological-terrorism-2016-9 https://www.businessinsider.com/susan-fiske-methodological-t...
- Hnrobert42 6y ago
  You offer no evidence that these scientists were biases for affected by their bias. You offer only speculation.