Summary of psychometric measurement progress
The distinction between formative and reflective indicators has important implications for validation approaches. For details, please refer to the WHO Ageism Scale Manual and User Guide.
For formative/causal indicators, psychometric measurement models such as factor analysis and item response theory, which assume an underlying latent trait are not appropriate. As a result, tests of internal consistency reliability and factorial analysis are not appropriate for assessing the psychometric quality of formative indicators. However, formative indicators could still be expected to show high test-rest reliability and convergent and criterion validity with related measures. Based on these methodological considerations, the current analyses of interpersonal and institutional experiences items focus on convergent and criterion validity, as well as test-retest reliability, while analyses of self-directed ageism additionally include the internal structure and internal consistency reliability of the items.
Summary of psychometric measurement progress
Psychometric property test | Tested? | Sample |
Exploratory factor analysis (EFA)
(Self-directed items only) |
✓ | UK mixed-age sample
Moldova older persons |
Confirmatory factor analysis (CFA)
(Self-directed items only) |
✓ | Colombia older persons |
Internal consistency
(Self-directed item only) |
✓ | UK mixed-age sample
Moldova older persons Colombia older persons |
Test-retest reliability | ✓ | UK mixed-age sample |
Convergent validity | ✓ | UK mixed-age sample
Moldova older persons Colombia older persons |
Discriminant (Divergent) Validity | ✗ | |
Criterion-Related Validity | ✓ | Moldova older persons (concurrent)
Colombia older persons (concurrent) |
Measurement invariance
(Self-directed items only) |
✗ |
Note. ✓ indicates that the test was performed, while ✗ indicates that the test was not performed.