Download files
Abstract
The large scale collection of psychiatric data in both community and clinical settings is crucial for the continual evolution of the psychiatric nomenclature and the overall understanding of how and why psychiatric disorders exist in nature. To facilitate the collection of large scale data, diagnostic instruments must be relatively free from respondent burden and cost effective to administer whilst remaining faithful to the psychiatric nomenclature. Although a large selection of diagnostic tools and questionnaires exist, the extant literature reveals that very few of these instruments possess an optimal balance between validity, reliability, and utility. Instruments that are valid and reliable are often limited in terms of utility, whilst instruments that are free from burden and cost effective are often limited in terms of validity and reliability.
The current thesis sought to design an instrument that possesses an optimal balance between validity, reliability, and utility by reducing the length and costs associated with an existing valid and reliable instrument, the Composite International Diagnostic Interview (CIDI) version 3.0. Signal detection theory was utilised to retain the smallest number of items in ten disorder modules of the CIDI 3.0 to predict a lifetime DSM-IV diagnosis. Preliminary concordance analysis between the CIDI 3.0 and the short form modules, known as the Mini-CIDI 3.0, demonstrated an excellent level of agreement in samples of the Australian and the American general population.
Two additional studies were conducted to investigate the content validity and the construct validity of the Mini-CIDI 3.0. Cognitive interviewing demonstrated that the respondents were adequately able to interpret, recall, judge, and respond to the item content of the Mini-CIDI 3.0 in a similar manner as the CIDI 3.0. Subsequently, confirmatory factor analysis demonstrated that a theoretical model of internalising disorders provided the best model fit for data generated from the Mini-CIDI 3.0 and the CIDI 3.0.
The good validity of the Mini-CIDI 3.0 presented in this thesis enables clinicians and researchers to substitute the CIDI 3.0 with the Mini-CIDI 3.0 in clinical and community studies when they wish to reduce respondent burden and associated administration costs. Future applications of the Mini-CIDI 3.0 are discussed.