A Psychometric Evaluation of the Measurement of Agency in Adults with Physical Disabilities in Long-Term Care
Abstract
Background: Evaluations of healthy aging tend to focus on a person’s lived environments, their
intrinsic capacities (physical and mental capacities to act), and the interplay between the two, as
these factors shape the aging process. Currently, there is a lack of research that has identified
positive indicators of intrinsic psychological capacities, particularly for adults living with
disabilities in long-term care (LTC) homes. Personal agency (beliefs about one’s own abilities to
act to achieve goals) and interpersonal agency (beliefs about one’s ability to engage with others
to achieve goals) might be useful strengths-based indicators of intrinsic psychological capacities
because they reflect individuals’ beliefs about the strategies that they use to achieve control over
life. The Personal Agency Scale (PAS) and Interpersonal Agency Scale (IPAS) are measures of
these constructs, but little is known about their psychometric properties.
Objectives: The primary aim of the study was to test the psychometric characteristics of the PAS
and IPAS within the context of adults who are aging with physical disabilities in a LTC home.
The specific objectives were to estimate item response variation and internal reliability of these
scales, test concurrent validity, and assess their sensitivity to change over time.
Methods: Repeated-measures data taken 6-months apart from a sample of 59 adults aging with
physical disabilities in a long-term care home were analysed. Between waves of data collection,
participants experienced changes to their physical and social environments that were likely to
have changed subjects’ levels of agency. Item-analyses were conducted to estimate internal
consistency of the scales, including Cronbach’s alpha and assessment of response distributions
for all items. Generalized Estimating Equations was used to test for known education-group
comparisons in agency scores and known correlations between agency scores and depression to
assess the construct validity of the PAS and IPAS. Paired Test-Statistics were used to test for
changes in individuals’ levels of agency from baseline to 6-month follow-up.
Results: This study found mixed evidence regarding the psychometric properties of the PAS and
IPAS. PAS measurements were found to have sufficient internal consistency at both timepoints
whereas the IPAS measurements were only internally consistent at baseline. Item response
variation was low, particularly for the IPAS at the 6-month follow-up. No education group
differences in PAS scores or IPAS scores were found. Furthermore, there was no evidence of an
association between depression and agency scores. Individuals’ PAS and IPAS measurements
were of similar magnitude at both timepoints.
Interpretation: This study did not provide sufficient evidence to endorse the use of the PAS and
IPAS as an indicator of intrinsic psychological capacity for older adults living with physical
disabilities in LTC, in their current form. Future psychometric evaluations of these measures
should consider the impact adopting a 5-point verbal response scale and adding a time-reference
to item stems on measurement performance.