4 ms·

> The therapeutic action in mice, which is replicated in the study I linked to and many others. Would that be the same therapeutic effect that was falsified in

by lutusp 12y ago

> The therapeutic action in mice, which is replicated in the study I linked to and many others. Would that be the same therapeutic effect that was falsified in the most recent study, the one that showed no correlation between serotonin and depression? Neither study represents anything like a conclusive outcome, but both must be read with an open mind -- who funded the studies, how strong is the evidence, what are the p-values, and so forth. But let's step back. Do you know what's missing from this exchange, and do you know why such exchanges tend to be so wordy and inconclusive? Both of us are discussing symptoms, descriptions, not causes. This issue will never move forward until people move beyond psychiatric descriptions and locate a biological cause for depression. Science requires a focus on causes, explanations -- descriptions aren't enough. And this is not science, because psychiatry is not science. For those who doubt the central role of explanations in science, I have an anecdote -- Doctor Dubious invents a new treatment for the common cold. His treatment is to shake a dried gourd over the cold sufferer until the patient gets better. Sometimes the treatment takes a week, but it always works — the cold sufferer always recovers. So, why doesn't Doctor Dubious get a Nobel Prize for his breakthrough? The answer is that the procedure is only a description — shake the gourd, patient recovers — without an explanation, without a basis for actually learning anything or being truthful about the connection between cause and effect. It's the same with psychiatry.

dllthomas 12y ago
"For those who doubt the central role of explanations in science, I have an anecdote -- Doctor Dubious invents a new treatment for the common cold. His treatment is to shake a dried gourd over the cold sufferer until the patient gets better. Sometimes the treatment takes a week, but it always works — the cold sufferer always recovers. So, why doesn't Doctor Dubious get a Nobel Prize for his breakthrough?" You always go to this example, but it does not convince anyone of anything because you are not only foregoing explanation but also a control group. If you do have a randomly assigned group of people who get gourd-shook and a randomly assigned group of people who do not, and those who do recover significantly faster than the those who do not, and you have no idea how gourd-shaking might cause that... maybe it's still incorrect to say "the study showed gourd-shaking effective in treating the cold" but that's not intuitively obvious - explain why.
- lutusp 12y ago
  > You always go to this example, but it does not convince anyone of anything because you are not only foregoing explanation but also a control group. Yes, because my example is meant to caricature psychological "science" -- that's its purpose. Therefore it must have the same logical pitfalls as the class of study it caricatures. Is there a suggestion that psychiatrists or psychologists have control groups that work? Does the expression "no-treatment control" sound familiar? It's an often-seen expression in psychiatry and psychology studies. It results from an awareness that real scientists have control groups, but because of the nature of psychological studies, that's often not practical (imagine a study comparing therapeutic methods -- how would one design a realistic faux therapy for control purposes?). But these people want the appearance of science, so they call those denied treatment a "no-treatment control". "Send him home, we're not treating him. Oh, and add him to the 'no-treatment control' group." http://www.academia.edu/186808/The_Selection_and_Design_of_Control_Conditions_for_Randomized_Controlled_Trials_of_Psychological_Interventions http://www.academia.edu/186808/The_Selection_and_Design_of_C... Quote: "Although no-treatment controls have an appealing simplicity, they also have a number of potential disadvantages." I always laugh when I read that. > ... maybe it's still incorrect to say "the study showed gourd-shaking effective in treating the cold" ... Remember that I was mimicking a psychiatrist or psychologist, the sort of person willing to publish a paper containing this kind of reasoning in a theoretical vacuum. Anyway, the absence of a central corpus of falsifiable theory in psychiatry speaks for itself. All real sciences have such a corpus of theory, it's falsifiable and it informs all work in the field, psychiatry doesn't have one. Psychiatrists are free to go in any direction they please, even contradict each other on matters of substance, and without a basis in theory. It's not science.
  - dllthomas 12y ago
    'Yes, because my example is meant to caricature psychological "science" -- that's its purpose. Therefore it must have the same logical pitfalls as the class of study it caricatures.' You explicitly stated that the example was "for those who doubt the central role of explanations in science". That you can throw together an example with a bunch of flaws that produces a flawed result is both unsurprising and not especially damning of any specific one of the flaws - one flaw is enough to be wrong. Therefore, your example is not very persuasive on the issue you claimed it directed at. If you're just trying to get your jollies lampooning everything that sometimes goes wrong with psych research, that's fine - but present it that way. If you're trying to help people reason better - and possibly find flaws in your own reasoning - isolating particular problems is a much better approach.

by lutusp 12y ago

dllthomas 12y ago

"For those who doubt the central role of explanations in science, I have an anecdote -- Doctor Dubious invents a new treatment for the common cold. His treatment is to shake a dried gourd over the cold sufferer until the patient gets better. Sometimes the treatment takes a week, but it always works — the cold sufferer always recovers. So, why doesn't Doctor Dubious get a Nobel Prize for his breakthrough?" You always go to this example, but it does not convince anyone of anything because you are not only foregoing explanation but also a control group. If you do have a randomly assigned group of people who get gourd-shook and a randomly assigned group of people who do not, and those who do recover significantly faster than the those who do not, and you have no idea how gourd-shaking might cause that... maybe it's still incorrect to say "the study showed gourd-shaking effective in treating the cold" but that's not intuitively obvious - explain why.

lutusp 12y ago
> You always go to this example, but it does not convince anyone of anything because you are not only foregoing explanation but also a control group. Yes, because my example is meant to caricature psychological "science" -- that's its purpose. Therefore it must have the same logical pitfalls as the class of study it caricatures. Is there a suggestion that psychiatrists or psychologists have control groups that work? Does the expression "no-treatment control" sound familiar? It's an often-seen expression in psychiatry and psychology studies. It results from an awareness that real scientists have control groups, but because of the nature of psychological studies, that's often not practical (imagine a study comparing therapeutic methods -- how would one design a realistic faux therapy for control purposes?). But these people want the appearance of science, so they call those denied treatment a "no-treatment control". "Send him home, we're not treating him. Oh, and add him to the 'no-treatment control' group." http://www.academia.edu/186808/The_Selection_and_Design_of_Control_Conditions_for_Randomized_Controlled_Trials_of_Psychological_Interventions http://www.academia.edu/186808/The_Selection_and_Design_of_C... Quote: "Although no-treatment controls have an appealing simplicity, they also have a number of potential disadvantages." I always laugh when I read that. > ... maybe it's still incorrect to say "the study showed gourd-shaking effective in treating the cold" ... Remember that I was mimicking a psychiatrist or psychologist, the sort of person willing to publish a paper containing this kind of reasoning in a theoretical vacuum. Anyway, the absence of a central corpus of falsifiable theory in psychiatry speaks for itself. All real sciences have such a corpus of theory, it's falsifiable and it informs all work in the field, psychiatry doesn't have one. Psychiatrists are free to go in any direction they please, even contradict each other on matters of substance, and without a basis in theory. It's not science.
- dllthomas 12y ago
  'Yes, because my example is meant to caricature psychological "science" -- that's its purpose. Therefore it must have the same logical pitfalls as the class of study it caricatures.' You explicitly stated that the example was "for those who doubt the central role of explanations in science". That you can throw together an example with a bunch of flaws that produces a flawed result is both unsurprising and not especially damning of any specific one of the flaws - one flaw is enough to be wrong. Therefore, your example is not very persuasive on the issue you claimed it directed at. If you're just trying to get your jollies lampooning everything that sometimes goes wrong with psych research, that's fine - but present it that way. If you're trying to help people reason better - and possibly find flaws in your own reasoning - isolating particular problems is a much better approach.