Table 1.
Clinical and demographic features of people with PD (PwPD) and controls at T1 (5 years following the diagnosis of PD).
Fig 1.
Distribution of locations of people and SWLS response category thresholds.
Distribution of locations of people (pooled sample of people with PD and control subjects; upper panel) and SWLS response category thresholds (lower panel) on the common logit metric (x-axis; positive values = higher life satisfaction). All locations are relative to the mean item threshold location, which is set at 0 logits. Thresholds are the scale’s points of measurement and represent locations where there is a 50/50 probability of responding in either of two adjacent item response categories. There is thus one threshold less than the number of response categories for each item, rendering 5 items x (7 response categories-1) = 30 thresholds (points of measurement) for the SWLS. Superimposed on the person distribution graph is the information function curve (the inverse of measurement error; higher values = less error and more information in scores, i.e., better measurement precision). Maximum information (vertical line under the information function curve) corresponds to a location of -0.8 logits (representing a raw SWLS score of 16–17 on the original 5–35 score range). PD, Parkinson’s disease; SWLS, Satisfaction with Life Scale.
Table 2.
Descriptive and psychometric statistics according to classical test theory (CTT) for raw SWLS scores (possible score range, 5–35) from PwPD and control subjects.
Table 3.
Overall Rasch model fit statistics, reliability and targeting of the SWLS among PwPD and controlsa.
Fig 2.
Item characteristic curves (ICCs) of the five SWLS items.
ICCs representing SWLS items 1 (panel A), 2 (panel B), 3 (panel C), 4 (panel D) and 5 (panel E). Grey curves (ICCs) represent expected item responses (y-axis) for each person location (x-axis) on the life satisfaction continuum (positive values = higher life satisfaction). Black dots represent the observed responses from groups of people at similar locations on the measured continuum (x-axis). ICC, Item characteristic curve; SWLS, Satisfaction with Life Scale.
Table 4.
Rasch item and fit statisticsa.
Fig 3.
Example category probability curves.
Locations on the life satisfaction continuum are indicated on the x-axis (with threshold locations centred at zero; positive values = higher life satisfaction) and the y-axis represents the probability of affirming response categories 1 through 7 (rescored as 0 through 6 in the analysis) relative to the location on the measured construct (x-axis). Panel A shows item 3, representing the pattern with disordering between the two first response category thresholds (also seen with item 1). Panel B displays item 2, representing items without disordered response category thresholds (also seen with items 4 and 5). Thresholds are the locations where there is a 50/50 probability of responding in either of two adjacent item response categories. SWLS, Satisfaction with Life Scale.
Fig 4.
Differential Item Functioning (DIF) between people with PD and control subjects.
Panel A displays uniform Differential Item Functioning (DIF) between people with PD and control subjects for item 5 of the SWLS. The item characteristic curve (ICC; grey curve) represents the expected response category endorsement (y-axis) at various levels of life satisfaction (x-axis). Superimposed plots represent the observed responses by people with PD (x) and control subjects (⭕), as ivided into eight class intervals according to their levels of life satisfaction. People with PD score systematically higher than control subjects in all class intervals but the two lowest. Observed differences indicate that the item does not work the same way in the two groups. For comparison, panel B represents an item without DIF (item 4). DIF, Differential Item Functioning; PD, Parkinson’s disease; SWLS, Satisfaction with Life Scale
Fig 5.
Scatterplot of the independently estimated linear logit locations.
Matrix scatterplot of the independently estimated linear logit locations associated with each possible raw total SWLS score from the pooled sample (people with PD & controls), people with PD and control subjects (see Table 5). Inserted in each panel are the respective Pearson product-moment correlations (r). Nonparametric Spearman correlations were 1.0 in all three instances. Intraclass correlation across the three sets of estimates is 0.990. PD, Parkinson’s disease; SWLS, Satisfaction with Life Scale.
Table 5.
Interval level logit locations (measures) and standard errors associated with each possible ordinal level raw total SWLS score for the pooled sample (PwPD and controls) and for PwPD and controls separatelya.