[psych/pshrinkery, DSM5] "DSM 5 Has Flunked Its Reliability Tests"

[siderea]

Frances Allen, MD, (famously the chair of the DSM-IV Task Force and thorn in DSM-5's side) reported Sunday that at the APA meeting (convention), the results of the field trial of the new DSM definitions were presented... and they were dire: For context, in previous DSM's, a diagnosis had to have a kappa reliability of about 0.6 or above to be

Comments

[en_ki]

Those confidence intervals are enormous. Did they get, like, three patients?

[siderea]

I vote some of the sites discovered polyhedral dice.

[ron_newman]

Are there any similar studies of reliability for non-psychiatric medical diagnoses?

[siderea]

There must, somewhere.

[ron_newman]

Somewhat related, in today's Boston Globe: Proposed new diagnoses for bipolar disorder in children divides psychiatrists