Comparisons of measures at two time-points or in two situations can become complicated. Example 12 illustrates a straightforward situation. It uses the 13 motor items of the FIM®. This example uses the 18 items of the FIM, 13 motor and 5 cognitive, at two time points. There is a 7-level rating scale that is intended to operate in the same way across all items.

 

In exam17s.txt, the data have been stacked. The items are modeled to maintain their difficulties across the two time points, and the 52 patients are entered twice, once at admission to rehabilitation and once at discharge from it, so there 104 data records. Changes in patient independence can be identified by cross-plotting the admission and discharge measures for each patient, as in Example 12. This can be done by using the Plots menu to plot the measures against themselves, and then, in Excel, pasting the discharge measures over the top of the admission measures for the y-axis. Here is the procedure:

 

---

Easier:

Arrange time 1 and time 2 person records in the same order. All Time 1 then all Time 2.

1. you have your time 1 + time 2 data file

2. also create a time 2 + time 1 data file (simply move time 1 data lines after time 2 data lines)

3. analyze  time 2 + time 1 data file: output PF2.txt - delete time time 1 lines from PF2.txt ((second half of the file)

4. analyze time 1 + time 2 data file

5. scatterplot this analysis (which is still running) and PF2.txt - PF2 (time 2) will match time 1 in this analysis.

 

---

 

1.Do the analysis with all time 1 data above all time 2 data, and the persons in the same order at both time-points.

2.Output Files menu: PFILE=pf.txt

3.Edit pf.txt

4.Place time 1 persons in pf1.txt

5.Place time 2 persons in pf2.txt

6.Make sure that the person labels match (put in blank rows to make them align)

7.Winsteps Plots menu: Scatterplot

8.x-axis file is pf1.txt

9.y-axis file is pf2.txt

10.Plot the measures

11.Plot is produced in Excel

 

In exam17r.txt, the data have been racked. The persons are modeled to maintain their abilities across the two time points, and the 18 FIM items are entered twice, once at admission to rehabilitation and once at discharge from it, so there 36 items. (To rack datasets from several time points follow the Transpose procedure.) Average changes in patient performance on individual items can be identified by cross-plotting the admission and discharge measures for each item. This can be done by using the Plots menu to plot the measures against themselves, and then, in Excel, pasting the discharge measures over the top of the admission measures for the y-axis.

 

A further feature is the contrasts in the Principal Components Analysis of Residuals, Tables 23 and 24. The 13 motor items and 5 cognitive items are probing different aspects of patient independence. Do they function as one variable in this sample? See Principal components/contrast. Patients also have different impairments. Are their measures comparable? See Differential Item Function.

 

; This control file is EXAM17S.TXT

TITLE='Example 17S: 18-item FIM control file: stacked admission then discharge'

ITEM1=1  ; Responses start in column 7

NI=18  ; 13 motor items + 5 cognitive items

NAME1=20 ; start of person label

; variables in person label

@SETTING=$S1W1 ; setting: A=Admission, D=Discharge

@SEX=$S3W1 ; Gender: 1=Male, 2=Female

@IGC=$S5W2 ; Impairment Group Code: 8=Orthopedic, 9=Cardiac, 10=Pulmonary, 11=Burns, 12=Congenital, 13=Other

@ID=$S8W2 ; Patient sequence number

 

; variables in item label

@SUBTEST=$S1W1  ; subtest: M=Motor, C=Cognitive

@ILETTER=$S3W1 ; code letter of item

 

CODES=1234567 ; 7 level rating scale

CLFILE=* ; Defines the rating scale

1 0% Independent

2 25% Independent

3 50% Independent

4 75% Independent

5 Supervision

6 Device

7 Independent

*

&END

M A. EATING

....

C R. MEMORY

END NAMES

334412312331123112 A 1  8  1

432322442223134444 A 2  8  2

....

 

; This control file is EXAM17R.TXT

TITLE='Example 17: 18-item FIM control file: racked admission and discharge'

ITEM1=1  ; Responses start in column 7

NI=37  ; 13 motor items + 5 cognitive items: admission and discharge

NAME1=39 ; start of person label

 

; variables in person label

@SEX=$S1W1 ; Gender: 1=Male, 2=Female

@IGC=$S3W2 ; Impairment Group Code: 8=Orthopedic, 9=Cardiac, 10=Pulmonary, 11=Burns, 12=Congenital, 13=Other

@ID=$S5W2 ; Patient sequ. number

 

; variable in item label

@SETTING=$S1W1 ; setting: A=Admission, D=Discharge

@SUBTEST=$S3W1  ; subtest: M=Motor, C=Cognitive

@ILETTER=$S5W1 ; code letter of item

 

CODES=1234567 ; 7 level rating scale

CLFILE=* ; Defines the rating scale

1 0% Independent

2 25% Independent

3 50% Independent

4 75% Independent

5 Supervision

6 Device

7 Independent

*

&END

A M A. EATING

....

A C R. MEMORY

-      Blank for ease of seeing

D M A. EATING

....

D C R. MEMORY

END NAMES

334412312331123112 554535546665345545 1  8  1

432322442223134444 777677777666567777 2  8  2

....