From interviews with adolescent and adult patients with EoE, pain was not highlighted as an important symptomatic factor and question 4 was thus considered an exploratory item and not included as part of the psychometric evaluation.
Further exploratory item level analysis was undertaken to identify potential sources of bias.
Exploratory item factor analysis (EFA) was used instead to give LD, or additional dimensions, an opportunity to become visible.
Known groups validity was also demonstrated on the baseline sample, and construct validity was examined using principal components analysis and exploratory item response theory analysis.
A sequence of analyses using LD X statistics computed in the course of concurrent unidimensional calibration of the five new items with the existing 29-item Upper Extremity item pool, exploratory item factor analysis, and fitting increasingly refined bifactor models led to the conclusion that two of the new items were involved in an LD triplet with one of the existing items.
The LD X statistics computed in the course of concurrent unidimensional calibration of the two new items with the existing 23-item Fatigue item pool did not show any clear pattern, so exploratory item factor analysis was used to investigate potential multidimensionality.
After removal of two items, exploratory factor analysis and parallel analysis revealed two distinct factors broadly consistent with existing two-factor formulations of pathological perfectionism, but with modest internal consistency and test retest reliability.
Based on exploratory results, indicator items are specified as anchor variables for factors on which they are hypothesized to load highest.
As an exploratory approach, single items from ODI (except sex and pain intensity) were entered in a stepwise procedure together with workload, work satisfaction and seated work, using a likelihood ratio based criterion with p < .05 for entry and p < .10 for removal.
The exploratory factor analysis of items 1 11 suggested that the items loaded onto 2 factors – items 1 4 and 9 11 loading onto one factor (PPSM-Global) and all the pain items loading onto another (PPSM-Pain).
