BACKGROUND: It has long been thought that individual activities of daily living (ADLs) can be combined to form a hierarchy or Guttman scale. The purpose of this study is to determine if ADLs fit into a single hierarchical structure, and to examine how such a hierarchy should be evaluated. METHODS: We use data from the baseline year of the Longitudinal Study of Aging, a nationally representative survey of noninstitutionalized persons 70 years of age and older. For each of the 360 permutations of the ADLs within the Katz hierarchy, we calculate the standard measures of fit of ordered data to a Guttman scale: the coefficient of reproducibility, the minimum marginal reproducibility, the percentage improvement, and the coefficient scalability. RESULTS: We find that although the Katz hierarchy does satisfy the traditional requirements for scalability, many other ADL hierarchies also satisfy these criteria. Specifically, our analysis shows that there are 4 hierarchies at least as good as the Katz hierarchy, and 103 hierarchies which satisfy the minimum standard for scalability. CONCLUSIONS: We conclude that the typical scalogram methodology may not be sufficient to summarize data, and that a multiplicity of disability profiles may exist.
BACKGROUND: It has long been thought that individual activities of daily living (ADLs) can be combined to form a hierarchy or Guttman scale. The purpose of this study is to determine if ADLs fit into a single hierarchical structure, and to examine how such a hierarchy should be evaluated. METHODS: We use data from the baseline year of the Longitudinal Study of Aging, a nationally representative survey of noninstitutionalized persons 70 years of age and older. For each of the 360 permutations of the ADLs within the Katz hierarchy, we calculate the standard measures of fit of ordered data to a Guttman scale: the coefficient of reproducibility, the minimum marginal reproducibility, the percentage improvement, and the coefficient scalability. RESULTS: We find that although the Katz hierarchy does satisfy the traditional requirements for scalability, many other ADL hierarchies also satisfy these criteria. Specifically, our analysis shows that there are 4 hierarchies at least as good as the Katz hierarchy, and 103 hierarchies which satisfy the minimum standard for scalability. CONCLUSIONS: We conclude that the typical scalogram methodology may not be sufficient to summarize data, and that a multiplicity of disability profiles may exist.
Authors: Sherry L Willis; Sharon L Tennstedt; Michael Marsiske; Karlene Ball; Jeffrey Elias; Kathy Mann Koepke; John N Morris; George W Rebok; Frederick W Unverzagt; Anne M Stoddard; Elizabeth Wright Journal: JAMA Date: 2006-12-20 Impact factor: 56.272
Authors: Justine M Nagurney; William Fleischman; Ling Han; Linda Leo-Summers; Heather G Allore; Thomas M Gill Journal: Ann Emerg Med Date: 2017-01-06 Impact factor: 5.721
Authors: Margaret G Stineman; John T Henry-Sánchez; Jibby E Kurichi; Qiang Pan; Dawei Xie; Debra Saliba; Zi Zhang; Joel E Streim Journal: Am J Phys Med Rehabil Date: 2012-02 Impact factor: 2.159
Authors: George W Rebok; Karlene Ball; Lin T Guey; Richard N Jones; Hae-Young Kim; Jonathan W King; Michael Marsiske; John N Morris; Sharon L Tennstedt; Frederick W Unverzagt; Sherry L Willis Journal: J Am Geriatr Soc Date: 2014-01-13 Impact factor: 5.562