The thing that caught my attention (35 minutes into the show) was the description of an experiment reported by Evelyn Hooker in 1956, testing if homosexual men were maladjusted people. Prior to Hooker's study, medical opinions about homosexuality were based on anecdotal evidence of psychologists who were attempting to "cure" homosexuality:
This procedure compromises the validity of the psychoanalytic conclusions in at least two important ways. First, the analyst's theoretical orientations, expectations, and personal attitudes are likely to bias her or his observations. ...For bringing rigor to psychology and discrediting a very influential form of pseudoscience, Hooker is now one of my scientific heroes.
A second problem with psychoanalytic studies is that they have only examined homosexuals who were already under psychiatric care – in other words, homosexuals who were seeking treatment or therapy. Patients, however, cannot be assumed to be representative of the general population.