Validation study of a new Italian self-report inventory of depression

Introduction: Starting from a set of 91 items, elicited on the basis of diagnostic criteria for Major Depression Episode of DSM-IV-TR (2000), we have constructed a 21-item unidimensional self-report scale of depression, named TDI (Balsamo, 2006), using item-response theory, specifically a modified version of the Rasch one-parameter modelling (Rasch, 1960-1980), proposed by Andrich (1988). Method: In the Rasch analysis, each item has several thresholds, as many as response categories are minus one. If the categories are ordered, then reflect the order of intensity of latent trait, thus also the thresholds should be ordered (1<2<3).  parameters for each threshold associated with each of 21 items were calculated. Our sample consists of 229 Italian outpatients (147 females/82 males. Age:Mean=33.95 yrs; SD=10.33 yrs) with various psychiatric diagnosis and 300 healthy subjects (207 females/82 males. Age:Mean=27.35 yrs; SD=10.55 yrs). Results: The sequence of the thresholds of each of 21 items resulted ordered. It would mean that the 21 items provided valid scales of measured latent trait of depression. Analysis of variance was used to test two patient-related factors: gender and mental health (normals versus outpatients). As regards gender, no significant effects were found, showing that the subjects’ responses to the items were independent of gender [F(1, 528)=0.42, p=n.s.]. Significant effects were found between means of ability scale of clinical sample and these of nonclinical sample [F(1, 528)=320.13, p<.0001]. Conclusions: The TDI is a 21-item unidimensional self-report scale of depression selected by Rasch analysis, which seems to catch the nucleus of depressive symptomatology. It has some advantages, as: 1) ordered sequence of the thresholds of each of 21 items; 2) no impact of gender bias, present in many depression self-report inventories; 3) good discriminative ability between normal and clinical subjects. Thus, it could be a valid screening instrument in that it sensitively discriminates between outpatients and normals. References Andrich, D. (1988). Rasch models for measurement. Sage, London. Balsamo, M. (2006). Costruzione di un nuovo strumento di misura self-report di depressione. [Construction of a new self-report depression inventory]. Unplished doctoral dissertation, “G.d’Annunzio” University of Chieti, Italy. Balsamo, M., & Saggino, A. (2007). Test per l’assessment della depressione nel contesto italiano: un’analisi critica [Tests for depression assessment in Italian context: A critical review]. Psicoterapia Cognitiva e Comportamentale, 13, 167-199.