Introduction
Method
Design
Item development
Cognitive interviews—developing evidence of content validity
Rasch analysis—developing evidence of structural validity and internal consistency
Measurement properties | Aim | Criteria |
---|---|---|
Rating scale functioning | Assess item responses and their fit to the Rasch model assumptions | Assessment of distribution At least 10 observations of each category Categories advance monotonically Outfit MNSQ < 2.0 Step calibrations (Andrich Thresholds) advance between 1 and 5 logits |
Item goodness of fit | Assess individual items’ fit to the Rasch model | Infit MNSQ between 0.7 and 1.3 [42] |
Structural validity | Principal components analysis (PCA) of the residuals to assess unidimensionality and local independence | Unidimensionality is supported when: raw variance explained by the measure is > 50% of the total variance, and that the variance explained by the first principal component is small, i.e., < 2 eigenvalues [35] |
Internal consistency | Targeting of persons with mean and S.E of θ (theta) | Report direction and distance from mean item measure at 0 [43] |
Internal consistency | Assess person separation index and reliability Assess item separation index and reliability | Person separation > 2 Person reliability ≥ 0.8 Item separation > 3 Item reliability > 0.9 [35] |
Internal consistency | Assess consistency in correlations between items | KR-20 ≥ 0.70 for each unidimensional scale or subscale [43] |
Person goodness of fit | Detect improbable item-score patterns | Infit MNSQ ≤ 1.4 and/or associated z-value < 2 [14] |
Uniform differential item function (DIF) | Assessed across gender and three age groups (24–44, 45–60, 61–83) selected based on previous known curvilinear relationship with PSF and age [38] |
Structural validity
Internal consistency
Score-to-measure conversion
Ethical considerations
Results
Instrument development by expert panel
Version 2.0 of the Norwegian Fatigue Characteristics and Interference Measure (FCIM)a | ||
In this questionnaire, we want to assess post-stroke fatigue. It is very normal to feel tired during periods when you have a lot to do, but being fatigued means that you are more tired than you would expect considering what you have done | ||
Please choose the answer that best describes how you have been feeling in the past 7 days | ||
To what degree did you feel | Not at all—a little bit—somewhat—quite a bit—almost always | |
1. Fatigued | ||
2. Exhausted | ||
3. Mentally fatigued | ||
4. Tired in your head | ||
5. Physically fatigued | ||
6. Tired in your body | ||
7. Fatigued in the morning | ||
8. Fatigued around noon | ||
9. Fatigued in the afternoon | ||
10. Fatigued in the evening | ||
Challenges due to fatigue | Never – Rarely – Sometimes – Often – All the time | |
11. How often were you so fatigued that you had problems concentrating? | ||
12. How often were you so fatigued that you had problems making decisions? | ||
13. How often were you so fatigued that you had problems following a conversation? | ||
14. How often were you so fatigued that you had problems gathering your thoughts? | ||
15. How often were you so fatigued that you had problems taking a bath or shower? | ||
16. How often were you so fatigued that you had problems getting dressed/undressed? | ||
17. How often did you have problems starting your tasks/activities because of fatigue? | ||
18. How often did you have problems completing your tasks/activities because of fatigue? | ||
19. How often did you have to give up on your tasks/activities because of fatigue? | ||
20. How often have tasks/activities taken more time because of fatigue? | ||
21. How often did you avoid activities outside your home because of fatigue? | ||
22. How often has it been difficult to plan activities ahead of time because of fatigue? | ||
23. How often have you felt fatigued even if you have not done anything? | ||
24. How often did you avoid physical activity because of fatigue? | ||
25. How often did you limit your social activities because of fatigue? | ||
26. How often were you too fatigued to be together with your family? | ||
27. How often did you avoid engaging in hobbies or leisure activities because of fatigue? | ||
28. How often did you avoid pleasant activities because of fatigue? | ||
Pre-stroke fatigue | ||
29. Before you had a stroke, to what degree did you feel fatigued then? Not at all – A little bit – Somewhat – Quite a bit – Almost always | ||
30. Before you had a stroke, for how long had you felt fatigued? (If item 29 is greater than ‘Not at all’) 1 month or less – 2–6 months – 7–12 months – more than a year |
Content validity testing by cognitive interviews
Patient characteristics | Cognitive interviews (n = 15) | Rasch analysis (n = 169) |
---|---|---|
Age, mean (range) | 55.5 (40–75)a | 52.4 (24–83) |
24–44 | 2 | 35 (20.7) |
45–60 | 7 | 98 (58.0) |
61–83 | 4 | 36 (21.3) |
Male, n (%) | 7 (46.7) | 66 (39.1) |
Education, n (%) | ||
Primary school | 2 (13.3) | 11 (6.5) |
Secondary school | 5 (33.3) | 62 (36.7) |
Higher education < 4 years | 4 (26.6) | 61 (36.1) |
Higher education ≥ 4 years | 4 (26.6) | 35 (20.7) |
Marital status, n (%) | ||
Married | 11 (73.3) | 95 (56.2) |
Unmarried | 3 (20.0) | 38 (22.5) |
Widowed | 0 | 4 (2.4) |
Divorced/separated | 1 (6.6) | 32 (18.9) |
Years since stroke, n (%) | ||
1–24 months | 13 (86.6) | 69 (41) |
> 2 years | 2 (13.3) | 100 (59) |
Type of stroke, n (%) | ||
Cerebral infarction | 13 (86.6) | 126 (75) |
Hemorrhage | 2 (13.3) | 34 (20) |
Unknown/other | 0 | 9 (5) |
Work status, n (%) | ||
Working (full or part time) | 5 (33.3) | 60 (35.5) |
Stroke-related speech disorder, n (%) | Not collected | 55 (32.5) |
Weekly rehabilitation with speech therapist, n (%) | 1 (6.7) | 12 (7.1) |
Fatigue Severity Scale (FSS7) total scoreb, n (%) | ||
No/mild fatigue (1–3.9) | 3 (20.0) | 21 (12.4) |
Moderate fatigue (4–4.9) | 5 (33.3) | 23 (13.6) |
Severe fatigue (5–7) | 7 (46.7) | 125 (74.0) |
Evaluating structural validity and internal consistency with Rasch analysis
Characteristics subscale
Measurement property | Characteristics subscale | Interference subscale | |||
---|---|---|---|---|---|
Model | Step #1 | Step #2 | Step #1 | Step #2 | Step #3 |
Item goodness-of-fit statistics (Infit MNSQ) | |||||
1st iteration | Item 10 (1.38) | Item 16 (1.41) | |||
2nd iteration | Item 2 (0.56) | Item 15 (1.48) | |||
Removed item (reason) | Item 10 (misfit) Item 2 (misfit and flagged) | Item 4 (flagged and locally dependent) Item 6 (flagged and locally dependent) | Item 16 (misfit) Item 15 (misfit) | Item 14 (locally dependent) Item 18 (locally dependent) | Item 27 (flagged) Item 28 (flagged) |
Items left for further analysis | 8 | 6 | 16 | 14 | 12 |
Principal components analysis (PCA) of the residuals: | |||||
Latent trait % | 55.4% | 57.6% | 62.0% | 62.9% | 62.8% |
1st component (eigenvalue) | 2.3 | 1.5 | 2.5 | 2.3 | 2.0 |
Local independence (Yen’s Q3 residual correlations between items) | Items 5 and 6 (0.55) Items 3 and 4 (0.37) | No positive correlations | Items 12 and 14 (0.37) Items 13 and 14 (0.37) Items 18 and 19 (0.33) Items 17 and 18 (0.31) Items 27 and 28 (0.25) Items 19 and 20 (0.24) Items 25 and 28 (0.24) | Items 11 and 12 (0.25) Items 19 and 20 (0.24) Items 17 and 23 (0.23) | Items 19 and 20 (0.20) Items 11 and 12 (0.20) Items 17 and 23 (0.20) Items 21 and 25 (0.20) |
Targeting of personsa Mean person measure (mean model standard error) | 1.08 (0.58) | 0.95 (0.67) | 0.52 (0.42) | 0.57 (0.45) | 0.55 (0.48) |
KR-20 (internal consistency) | 0.89 | 0.86 | 0.95 | 0.95 | 0.94 |
Person separation index (person reliability)a | 2.45 (0.86) | 2.21 (0.83) | 4.11 (0.94) | 3.87 (0.94) | 3.52 (0.93) |
Item separation index (item reliability) | 4.11 (0.94) | 3.63 (0.93) | 5.51 (0.97) | 5.87 (0.97) | 6.28 (0.98) |
Person misfit, n (%) | 11 (6.5) | 10 (5.9) | 17 (10) | 17 (10) | 17 (10) |