The study of the validity and reliability of suicide probability scale (sps) in clinical sample

Suicide which has been a big public healt problem in the world, also has been risen in our country, recently. The purpose of the current study is to investigate the validity and realibility of Suicide Probability Scale (SPS) developed by Cull and Gill to aid in the assessment of suicide risk in adolescent and adults, in clinical population. SPS is 36-item selfreport measure and has 4 subscales which are hopelessness, suicide ideation, negative selfevaluation and hostility. Each item are rated on a 4-point Likert-type scale. There were 428 (217 psychiatric patients and 211 healty people) people participated in the study. Study group was consisted of volunteer patient who are treated in the psychiatry department of the Medical School of the University of Aydın Adnan Menderes and control group was consisted of people who are living in Aydın, could be approached with personal relations and were accepted to participate in the study. The realibility of the scale was computed with Cronbach Reliability Analyze methods. The result of the analyses suggested that the internal consistency and split-half reliabilities of the scale were high. Validity of the scale was assessed in three ways. First, convergent validity of the SPS was assessed by examining the relationship between the SPS total score and Beck Depression Inventory (BDI), Beck Hopelessness Scale (BHS) and Rosenberg Self-Esteem Scale (RSES). To be expected there were significant pozitive correlations between SPS and BDI and likewise SPS and BHS; and negative correlation between SPS and RSES. Second, the discriminant validity was examined with the methods of t-test and discriminant analyze. The results of analyses showed that scores on the SPS differentiated between suicide attempters and controls and also the groups which have suicide ideation and which have not suicide ideation. Third, construct validity of SPS was tested with Confirmatory Factor Analyze (CFA) using LISREL 8.30. As indexed by a variety of fit indices, the model suggested an acceptable fit to data, providing evidence for the construct validity of SPS. Finally, in this study cutting point of the scale was investigated to use on normal and psychiatric population. The results showed that for the both group of normal and psychiatric patients the most suitable cutting point is 110. As a result the findings obtained showed that SPS is a valid and reliable measure to use in clinical population.