Rosenhan’s study can be criticised because a small sample of only 8 participants was used, therefore we cannot be certain that the DSM-11 process of diagnosis is always low in validity and high in reliability. The results form this study may have been due to other factors and not merely reliability and validity affecting the psychologist’s diagnosis, and because the sample is so small we cannot determine this. The study can also be criticised because it was low in ecological validity and not very true to real-life. The psychologists are not normally confronted by patients who wish to be admitted, and so therefore their diagnosis may have been affected by this unusual occurrence.
Temerlin conducted an experiment and found that clinically trained psychiatrists and clinical psychologists could be influenced in their diagnosis by an authority figure. The patient was an actor who had no symptoms, but just expressed an interest in psychotherapy; however, because the psychologists had heard someone suggesting that the patient seemed neurotic, but was actually psychotic, this affected their diagnosis. This study again highlights the problems in the DSM-11 classification system. The study supports the suggestion raised by the Rosenhan’s study, that the diagnosis system is low in validity, because may of the diagnosticians were influenced, which resulted in them giving the wrong diagnosis.
Beck et al found that agreement on diagnosis for 153 patients was only 54%, which is a lot less than the 100% diagnosis for the Rosenhan study. This study therefore shows that the system is low in reliability, as different psychologists could not agree on the same diagnosis. This may be due to vague criteria for diagnosis or inconsistencies in the techniques used to gather data. The sample used in this study is much greater than the one used in the Rosenhan study. The results from this study goes against the Rosenhan study, therefore suggesting that the small sample size may have affected the overall results, and actually the DSM-11 classification is unreliable which cause different psychiatrists to diagnose the same patient with a different disorder.
Cooper et al also conducted an experiment that showed that effects that reliability can have on a patients diagnosis. Cooper found that psychiatrists from one country were more likely to diagnose a patient with schizophrenia than psychiatrists from a different country when shown the same video of recorded clinical interviews. Each psychiatrist was shown the exact same patient, and using the same classification system diagnosed them with different disorder. This suggests that the system used by the psychiatrists is not reliable. The study was not very reliable, and so therefore once again a patient was misdiagnosed. However, this study does not look at the affects of the two cultures used. The psychiatrist from one culture may be more used to seeing schizophrenia than the other psychiatrist, so is more likely to diagnose the patients with it. Despite this possible difference in cultures, the study still shows that the classification system is unreliable.
Di Nardo et al found that reliability for some disorders can be quite high, whereas reliability for others is much lower. This suggests that the classification system is biased towards some disorders, because of problems with interpreting what the patient was feeling. If a patient was misdiagnosed, then they would be treated for the wrong disorder, and would not be helped. The DSM was able to correct this fault in their classification system to prevent problems in interpretation, and so to increase overall reliability for all disorders, and hopefully result in more correct diagnosis.
Much of the research into the reliability and validity of the DSM-11 classification system, was conducted a number of years ago. The diagnostic procedures have now improved, and we now have the DSM-IV-TR-VI, which is much more reliable and valid. These studies therefore cannot show how reliable and valid the new classification system is. Psychiatrists now generally agree on a diagnosis. These studies have therefore benefited many patients, as now it is less likely that they will be misdiagnosed because the classification system is unreliable and low in validity.