Contemporary Psychometric Evaluation of the ObsessiveCompulsive Inventory– Revised
Wootton, BM and Diefenbach, G and Bragdon, LB and Steketee, GS and Frost, R and Tolin, D, Contemporary Psychometric Evaluation of the Obsessive Compulsive Inventory- Revised, ABCT 48th Annual Convention, 20-23 November, 2014, Philadelphia, Pennsylvania (2014) [Conference Extract]
PDF (ABCT14 - OCI-R Paper - Wootton) Not available 138Kb
Traditionally, hoarding symptoms were coded under ObsessiveCompulsive Disorder (OCD),
however, in DSM5 hoarding symptoms are classified into a new independent diagnosis,
Hoarding Disorder (HD). This change will likely have a considerable impact on the selfreport
scales that are commonly used to assess symptoms of OCD, as these scales often include
items measuring symptoms of hoarding. This study evaluated the psychometric properties ofone of the most commonly used selfreport measures of OCD symptoms, the ObsessiveCompulsive InventoryRevised (OCIR), in a sample of 474 individuals with either OCD (n = 118), HD (n = 201) or no current or past psychiatric disorders (healthy control; HC) (n = 155). For the purposes of this study the OCIR was divided into two scales; the OCIOCD (measuring the 5 dimensions of OCD) and the OCIHD (measuring the hoarding dimension). Criterion validity was demonstrated for both measures; on the OCIHD, the HD sample scored higher than the HC and OCD groups (p's = <.001) and on the OCIOCD, the OCD group scored significantly higher than did the HC and HD groups (p's = <.001). Both scales also demonstrated good convergent and divergent validity. The OCIHD correlated strongly with the SIR (r =.94) and HRS (r =.89) and only moderately with a measure of anxiety (BAI r =.36). Similarly, the OCIOCD correlated more strongly with a measure of anxiety (BAI r =.61) than with measures of hoarding (SIR r =.06; HRS r =.01). Cronbach's alphas for both scales were high (r =0.94 for the OCIHD and r =0.92 for the OCIOCD). Receiver operating characteristic analyses were conducted in order to ascertain the diagnostic sensitivity and specificity of the OCIHD and OCIOCD scales. On the OCIHD, the area under the curve (AUC) was .97 (95% CI: .95.98) and a cutoff score of 6 provided the best balance between sensitivity and specificity. On the OCIOCD the AUC was .91 (95% CI: .89.94) and a cut score of 12 provided the best balance between sensitivity and specificity. The results of this study highlight the potential of the OCIR as a screening measure for individuals with OCD and HD and demonstrate the utility of separate clinical cut offs for assessing likely diagnosis of both HD and OCD.