From: Novel insights regarding the measurement properties of the SCOPA-AUT
Concept | Explanation |
---|---|
Coefficient alpha | An index of reliability (internal consistency) |
Differential Item Functioning (DIF) | Assessment of weather items work invariantly in different subgroups of respondents, e.g., age and gender groups |
Fit residual | Quantification of differences between observed and expected item responses. The expected value is 0 (perfect fit) and values between -2.5 and + 2.5 are considered acceptable |
Item Characteristic Curve | Displays the relationship between observed and expected item responses at various levels, from less to more dysautonomia. This is one aspect of Rasch model fit (see below) |
Local dependency | Occurs when there is item redundancy (different items represent similar content), or can be due multidimensionality (see below) |
Multidimensionality | Items do not represent one common construct |
Person separation index (PSI) | An index of reliability that is conceptually analogous to coefficient alpha (see above) |
Rasch model fit | Different related approaches to assess to what extend empirical data fit the Rasch model |
Response category functioning | Weather response categories (for example never [0], sometimes [1], regularly [2], and often [3]) function as intended. Response category thresholds (see below) that are disordered imply that response categories are not functioning as expected from less to more |
Response category thresholds | The locations where there is equal probability of responding in either of two adjacent categories |
Strata | The number of statistically distinct groups of persons that can be identified. For instance, with two strata, the scale can differentiate between persons with less versus more dysautonomia |
Targeting | To what extent items in a scale represents the levels of the construct (in this study dysautonomia) reported by the sample |
Unidimensionality | Items represents one common construct (in this study dysautonomia) |