Table 5.
Descriptive and Psychometric Data of the APPLIQue Among Spousal Caregivers of People With AD.
| Time 1 | Time 2 | |
|---|---|---|
| Data completeness | ||
| Missing item responses (min-max %)a | 1–6 | 1–7 |
| Scaling assumptions | ||
| Corrected polyserial item-total correlation (min-max)b | 0.36–0.78 | 0.32–0.80 |
| Targeting | ||
| Possible score range (midpoint) | 0–25 (12.5) | 0–25 (12.5) |
| Mean (SD) scorec | 13.4 (6.2) | 13.5 (5.9) |
| Median (q1-q3) scorec | 15 (8–18) | 14 (8.5–18) |
| Min-max scored | 0–25 | 0–23 |
| Floor/ceiling effects (%)e | 1 / 1 | 2.4 / 0 |
| Skewnessf | –0.25 | –0.27 |
| Reliability | ||
| Ordinal αg | 0.94 | 0.93 |
| Ordinal α when item deleted (min-max)h | 0.94–0.94 | 0.93–0.93 |
| SEM, ordinal α based (% of total score)i | 1.5 (6.1) | 1.6 (6.3) |
a Should be <10%.
b Should be >0.3 to support summation of raw item scores, and >0.3-0.4 to support a single underlying variable.
c Should be close to scale midpoint.
d Should span most of the scale’s score range.
e Should be <15-20%.
f Should be between -1 and +1.
g Should be ≥0.80.
h Should not increase compared to α for the total score.
i Should be less than half of the total score SD
APPLIQue, the Alzheimer’s Patient Partners Life Impact Questionnaire; SD, standard deviation; q1-q3, 1st-3 rd quartile (25th-75th percentile); SEM, standard error of measurement.