Development of improved balance measures for community-dwelling older adults
Pardasaney, Poonam K
MetadataShow full item record
Balance measures used in community-dwelling elderly suffer from important psychometric limitations including limited comprehensiveness in content, ceiling effects, and limited sensitivity to change and responsiveness. This dissertation examines conceptual and methodological reasons underlying psychometric limitations of balance measures, and proposes strategies to overcome these limitations. Conceptually, psychometric limitations of balance measures may be related to gaps in representation of essential task and environmental components within their items. The first study conducted a systematic item-level content analysis of balance measures for community-dwelling elderly to examine the extent to which essential task and environmental components were represented within each measure. One hundred sixty-seven items from 26 measures were coded on seven criteria related to task and environment. Important gaps were noted across measures, with limited comprehensiveness in content areas represented and minimal incorporation of environmental variations. Most measures focused on single-task assessment in quiet, static environments, underrepresenting postural control demands in daily life situations, which involve dynamic changing environments, person-environment interactions, and multi-tasking. Methodologically, most balance measures are traditional fixed-form tests, which require administration of a fixed set of items to every individual. Developing comprehensive and precise fixed-form balance measures is challenging due to the large number of items needed to cover the spectrum of balance ability and components. Item response theory (IRT) and computer-adaptive testing (CAT) offer contemporary methodological approaches to develop comprehensive, tailored, and efficient balance tests with improved psychometric properties across the balance continuum. The second study built a computer-adaptive balance test from three traditional fixed-form balance measures using existing data on 187 community-dwelling older adults. Rasch IRT analysis was applied to compile a 23-item pool from the traditional measures. A 10-item CAT developed from the item pool showed excellent accuracy, acceptable reliability, and superior validity compared with the traditional measures. The CAT was the only measure to discriminate between fallers and non-fallers, and was a stronger predictor of self-reported function compared with the traditional measures. In summary, this dissertation reveals important content gaps in existing balance measures for community-dwelling elderly, highlights content areas that should be incorporated in new measures, and demonstrates advantages of computer-adaptive balance testing over traditional fixed-form measures. Development of new measures that better reflect postural control demands in daily life situations is recommended. Prospective investigation of computer-adaptive balance testing in community-dwelling elderly is also recommended.
Thesis (Sc.D.)--Boston University PLEASE NOTE: Boston University Libraries did not receive an Authorization To Manage form for this thesis or dissertation. It is therefore not openly accessible, though it may be available by request. If you are the author or principal advisor of this work and would like to request open access for it, please contact us at firstname.lastname@example.org. Thank you.