Skip to main content

Table 3 Difficulty levels, standard errors, fit statistics and Rank order for two Rasch ASES-p models (n = 161)

From: Patient self-report section of the ASES questionnaire: a Spanish validation study using classical test theory and the Rasch model

Item no.

Item description

All ASES-p itemsa

Function itemsb

δ (logit)

SE

Infit MNSQ

Outfit MNSQ

Rank order

δ (logit)

SE

Infit MNSQ

Outfit MNSQ

Rank order

Function

 Item 7

Lift 10 lbs above shoulder

1.78

0.13

1.09

0.86

1

1.82

0.14

1.09

0.89

1

 Item 8

Throw a ball overhand

1.07

0.13

0.82

0.72

2

1.09

0.13

0.81

0.73

2

 Item 3

Wash back/do up bra

0.89

0.12

0.90

0.83

3

0.90

0.13

0.90

0.85

3

 Item 2

Sleep on painful side

0.37

0.12

0.98

1.01

4

0.35

0.12

1.02

1.10

4

 Item 6

Reach high shelf

0.36

0.12

0.85

0.79

5

0.35

0.12

0.86

0.82

4

 Item 10

Do usual sport

−0.12

0.14

1.81

1.78

6

−0.16

0.14

1.86

1.90

5

 Item 9

Do usual work

−0.28

0.12

0.73

0.74

7

−0.32

0.13

0.80

0.88

6

Pain

 Item

How bad is your pain today

−0.33

0.12

1.26

1.74

8

-

-

-

-

-

Function

 Item 1

Put on a coat

−0.97

0.13

0.71

0.75

9

−1.05

0.13

0.78

0.89

7

 Item 5

Comb hair

−1.28

0.13

0.79

0.78

10

−1.38

0.13

0.81

0.83

8

 Item 4

Manage toileting

−1.50

0.13

1.19

1.11

11

−1.61

0.14

1.23

1.16

9

Reliability index: Person/item

0.90/0.98

0.88/0.98

Separation index: Person/item

2.93/7.53

2.77/8.02

Variance %: Observed/Expected

59.7/59.5

60.4/60.2

  1. δ (logit) level of item difficulty, based on the Rasch model, SE standard error, MNSQ mean square fit index, Rank order difficulty level, based on the δ (logit) measure. Variance refers to the Rasch model raw variance explained by the measures
  2. aIn this model, the original pain item was categorized as having 4 response options: 0–1 = No pain; 2–5 = some pain; 6–8 = a lot of pain; 9–10 = maximum pain. bItems 2 and 6 obtained the same δ (logit) estimation and where thus assigned the same rank order