A landmark study by Rosenhan (1973) exposed poor inter-rater reliability: pseudo-patients admitted to US psychiatric hospitals were diagnosed with schizophrenia by different clinicians despite displaying identical, normal behaviour after admission.