Determining the psychometric properties of Siddiqui anxiety scale - revised

Abstract

Siddiqui Anxiety Scale – Revised is a self-report screening instrument in Urdu language, for the assessment of anxiety, which was originally developed from the verbatim of the local patient population by Hasnain and Siddiqui (1993). It is based on a 4 point Likert Type Scale based on the frequency of symptoms of anxiety. The present study consisted of two phases and was aimed at the determining the psychometric properties of the revised version of the scale and explore the underlying factor structures in a Pakistani community sample. The instrument was first reviewed in a pilot study conducted with the local individuals (N = 30) to assess its comprehensibility and linguistic fluency. The items were later discussed with professionals through committee approach to assess their relevance. The final version of the tool now consists of 27 items out of which 23 are positively keyed and 4 are negatively keyed. In the second phase, a convenient sample of 500 Pakistani adults (215 males and 279 females) with a mean age of 20.8 (± 3.03) was selected. They filled four questionnaires namely Life Orientation Test – Revised, Siddiqui Anxiety Scale – Revised, Beck Depression Inventory II and Beck Anxiety Inventory, all in Urdu, after giving their verbal consent and assent. The Cronbach alpha for the scale was 0.895 indicating its sound internal consistency and high reliability. The factor analysis revealed two factors: Cognitive-Affective (Factor 1) and Somatic (Factor 2) with 13 and 14 items respectively. Both the factors were internally sound. The results also revealed a strong positive correlation between SAS-R, BAI (r = 0.763, p <0.001) and a negative correlation with Optimism and Pessimism scales of LOT-R respectively (r = -0.336 and r = -0.156, p < 0.001), which establish the convergent and divergent validity of SAS-R. Also, with a modest correlation with BDI II (r = 0.639) the items of these two scales loaded on Determining Psychometric Properties of SAS-R xii two separate factors in the exploratory factor analysis. This suggests the ability of this tool to differentiate between the constructs of anxiety and depression, unlike other indigenously developed tools. Furthermore, The ROC Curve analysis yielded a cutoff score of 30 in the community sample at which the scale has 81.4 % sensitivity, 77.66% specificity, 53.3% positive predictive percentage and 93% negative predictive percentage. Overall, these results provide preliminary evidence of reliability and validity of SAS-R for use in Pakistani community. The scale efficiently taps on the cognitive-affective and somatic components of anxiety. It can be used in various settings by mental health professionals, counselors, teachers, and parents; it can also be self-administered. It requires no specific training for administration and usually takes 10-15 minutes to complete.