Thursday, January 12, 2012: 8:00 AM-12:00 PM
Independence B (Grand Hyatt Washington)
Speaker/Presenter:
Social work researchers in all fields of social work rely on scales, or sets of related questions, to measure attitudes, behaviors, relationships, emotions, and functioning. Scales are superior to individual items for measuring complex phenomena. Some of the important advantages of scales, however, are lost when scores from the items they comprise are simply summed or averaged to create a composite score. Measurement error is included in the composites, for example, and all items are weighted equally regardless of their actual relative importance to the overall construct. Structural equation modeling (SEM) exploits the advantages of scales in social work studies, whether they are descriptive, explanatory, or causal studies, leading to more accurate findings. Furthermore, confirmatory factor analysis (CFA), a subset of SEM analyses, is a mandatory step in the development and validation of new scales.
Natasha K. Bowen, PhD, University of North Carolina at Chapel Hill
Social work researchers in all fields of social work rely on scales, or sets of related questions, to measure attitudes, behaviors, relationships, emotions, and functioning. Scales are superior to individual items for measuring complex phenomena. Some of the important advantages of scales, however, are lost when scores from the items they comprise are simply summed or averaged to create a composite score. Measurement error is included in the composites, for example, and all items are weighted equally regardless of their actual relative importance to the overall construct. Structural equation modeling (SEM) exploits the advantages of scales in social work studies, whether they are descriptive, explanatory, or causal studies, leading to more accurate findings. Furthermore, confirmatory factor analysis (CFA), a subset of SEM analyses, is a mandatory step in the development and validation of new scales.
See more of: Other Events
