Date of Award
Doctor of Philosophy
James P. Whelan
Andrew W Meyers
James G Murphy
George E Relyea
Retaining an empirically supported model while reducing assessment parameters becomes challenging in short form measurement development. In 2012 Larwin and Harvey proposed a systematic item reduction approach using structural equation modeling (SIR-SEM). The SIR-SEM permits retention of a strong connection to an empirically supported model while reducing some of the challenges of working with a large measurement battery. The application of the SIR-SEM strategy to reduce the number of items needed to assess an empirical supported multidimensional model of impulsivity (Ginley, Whelan, Meyers, Relyea, & Pearlson, 2014) is presented using a two-study procedure. To complete the item reduction, a SAS/Stat version of the SIR-SEM was developed and model fit statistics with extensive empirical support were adopted. In Study 1, the SIR-SEM approach successfully eliminated 84% of items while retaining 33 items to assess three impulsivity dimensions: behavioral activation, preference for stimulation, and inhibition control. Study 2 tested the resulting 33-item impulsivity measure, the Memphis Impulsivity Measure (MIM), in an independent sample of participants. This second study confirmed model fit. Each of the three MIM dimensions had similar moderate levels of internal consistency. The Pearson correlations for each dimension score indicated good two-week test-retest reliability. The MIM was found to be largely demographically invariant and to have a significant relation with target risk behaviors including: gambling frequency, symptomology, and classification, alcohol use problems, and alcohol use classification, and drug use involvement and complexity of involvement.
dissertation or thesis originally submitted to the local University of Memphis Electronic Theses & dissertation (ETD) Repository.
Ginley, Meredith Kathleen, "A Two Study Structural Modeling Based Approach for Ensuring Retention of Empirical Structures and Optimizing Short Form Development." (2016). Electronic Theses and Dissertations. 1422.