One of the main problems facing the psychiatrist in forensic psychiatry is the distinction between malingering and mental illness especially in Zulu speaking patients. This study identified twenty items from the literature and clinical practice that separate malingering from mental illness. The validity of these items was assessed through an experimental, cross -sectional study design which compared two groups. These were a sample of fifty malingering African patients, male and female and a control group of fifty mentally-ill African forensic patients who were classified as State Patients. Since the data was categorical, that is, the outcome was either positive (that is malingering) or negative (that is mentally ill) the groups were compared by employing such methods as the chi-square test and Fisher's exact test. Seventeen items we re found to be statistically significant and were regarded as valid items that separate malingering from mental illness. Then the effectiveness of these seventeen items in separating malingering from mental illness was determined by calculating their sensitivity, specificity, their false positive rate and their false negative rate. The items fell into Group four categories or groups. Group I are those three items with a high sensitivity, a high specificity, a few false positives, a few false negatives, high positive predictive values and high negative predictive values. They were able to diagnose both malingering and sickness with a high degree of accuracy. Group 11 consisted of eight items with a high specificity, a few false negatives and high positive predictive values. i1 These items are good at diagnosing malingering patients directly. Group III consisted of six items with a high sensitivity, a few false positives and high negative predictive values. These items are good at diagnosing sick patients and therefore diagnose malingering indirectly by excluding mental illness. Group IV consisted of three items which did not show statistical significance between malingering and mentally ill patients. This study proved that seventeen items were able to separate malingering from mental illness to a statistically significant extent and are effective for the use in the diagnostic assessment of Zulu speaking forensic patients.

Thesis (Ph.D)-University of Natal, Durban, 1997.