Abstract:

Objective    To study the reliability and validity of Trait Anxiety Inventory (TAI) in rural suicides and their paired controls. Methods    200 suicide deaths that occurred from September 1st, 2008 to August 31st, 2009 were collected from three counties selected from disease surveillance points of Shandong Province. Informants of each suicide or the paired control with the same age (±3 years), gender and district as the suicide were interviewed by trained interviewers using questionnaires and TAI. Results    The Cronbach′s α of suicides and controls were 0.906 and 0.877, respectively. The corrected item-total correlations in suicides and controls were 0.1830.697 and 0.1220.659, respectively. The total scores of trait anxiety had significant differences between the suicides(45.35±10.29) and controls(32.85±7.67)(t=-15.137, P<0.001), and had a positive correlation with the scores of Beck Hopelessness Scale(r=0.52，0.64；both P<0.001). Exploratory factor analysis was carried out in the suicidal group and the control group and yielded 3 factors and 4 factors, respectively. The adscription of items 5 and 8 in the two groups was different, which indicated that all factors extracted from the two groups were different from the hypothesis of the scale′s original designer. Conclusion    The reliability, discrimininative validity and criterion validity of TAI in suicide research are acceptable, but further research is needed to modify and improve its structural validity.

Key words: Trait anxiety； Scale； Suicide； Rural； Reliability； Validity