A hundred cases of parasuicide: V. Validation of the Amharic Version of Hopelessness Scale at St. Paul’s General Specialized Referral Hospital, AA, Ethiopia

Abstract

Abstract Background: Hopelessness is an important variable in parasuicide and suicide. The need for a valid and reliable Amharic version (AV) of Hopelessness Scale (HS) for clinical screening is obvious. Objective: To test the validity and the reliability of the AV of HS. Methods: 100 consecutive parasuicide cases arriving at the OPD were included in this ‘retrospective descriptive clinical case study’. Self-rated AV of HS and interviewer-rated Expanded BPRS was administered. The AV of HS was validated against the corresponding item/items on the Expanded BPRS. Reliability test was also performed. Results: Concurrent validity: Yule’s Q, between AV of HS and Expanded BPRS depression and suicidality items were 0.66265, p=0.00052 and 0.55585, p=0.04144 respectively. Construct validity: It was shown that cases with the intention ‘to die’ had significantly higher (p=0.0028) HS scores than those without the intention. The association between the dichotomous measures of AV of HS and intention ‘to die’ was very highly positive and significant (Yule’s Q=0.89563, P=0.00011). No such relationships between the AV of HS scores and other 9 endorsed intentions/‘reasons’ for parasuicidal act with different themes. These findings indicate that the AV of HS has an acceptable construct validity to measure Hopelessness. Reliability: in item-total correlation test, only item no. 5 was eliminated due to low correlation coefficient (0.0370, p=0.715). Three more items were eliminated in factor analysis which yielded 3 factors similar to Beck’s finding. Conclusion: The AV of HS has acceptable validity, reliability and factor loading. Items recommended to be discarded have to be re-translated/re-written to keep the meaning as close to the original (English version) as possible. Further validation studies are advisable to be done. In the meantime, the AV of HS has to be applied clinically in vulnerable groups. [Ethiop.J.Health Dev. 2008;22 (3):275-281]
PDF