IPMATRIX= is only available from the Output Files pull-down menu. It constructs a rectangular matrix in which there is one row for each person and one column for each item, or vice-versa. The entries in the matrix are selected from the following screen:
The first rows and columns can be the entry numbers, measures, labels and/or fields from the labels.
The matrix must contain one of
3. Original response value (after keying/scoring) (I4) (OBS)
4. Observed response value (after recounting) (I4) (ORD)
5. Expected response value (F7.3) (EXPECT)
6. modeled variance of observed values around the expected value (F7.3) (VAR)
This is also the statistical information in the observation.
Square root(modeled variance) is the observation's raw score standard deviation.
7. Standardized residual: (Observed - Expected)/Square root Variance (F7.3) (ZSCORE)
8. Score residual: (Observed - Expected) (F7.3) (RESID)
11. Measure difference (Person measure - Item measure) (F7.3) (MEASURE)
12. Log-Probability of observed response (F7.3) (LOGe(PROB))
13. Predicted person measure from this response (F7.3) (PMEASURE)
14. Predicted item measure from this response (F7.3) (IMEASURE)
15. Response code in data file (A) (CODE)
Field numbers shown here are those for XFILE=.
Depending on CSV=, data values are separated by "tab" or comma characters. In fixed field format, all fields are 7 characters long, separated by a blank. Missing data codes are "." as standard, but can be any character, or nothing.
1234567-1234567-1234567-1234567-1234567-1234567-1234567 ; these indicate fields.
. 1 2 3 4 5 6
1 8.85 -.223 . . . .
2 3.917 . 3.917 . . .
3 . 6.585 -.298 . . .
Example: I want a table of probabilities with items as the columns and possible scores as the rows, like the one on page 166 of Doug Cizek's book, Setting Performance Standards, based on work by Mark Reckase.
From your main analysis, write out an IFILE=itemanc.txt
Create a data set with one record for each possible score (it doesn't matter what the actual pattern of 1's and 0's is).
Enter the intended raw score as the person label.
In the control file for the new data set, put IAFILE=itemanc.txt
Analyze this second data set. Ignore any "subset" warning messages.
Check that the items are anchored in Table 14.
Check that the reported raw score for each person match that in the person label in Table 18.
Use the Output Files pull-down menu to write out an IPMATRIX=
Select "Expected response value", "Persons are rows, items are columns", "Person label", "Item entry number"
Then "Tab-separated", "EXCEL"
Then, for each observation in your data set, you have the probability for each score for each item.
Here it is from Exam1.txt (using the non-extreme items):
TITLE='KNOX CUBE TEST' ; Report title
NAME1=1 ; First column of person label in data file
ITEM1=11 ; First column of responses in data file
NI=18 ; Number of items
CODES=01 ; Valid response codes in the data file
iafile = itemanc.txt ; item calibrations from the original analysis
&END
END NAMES
1 10000000000000 ; score and a response string for it
2 11000000000000
3 11100000000000
4 11110000000000
5 11111000000000
6 11111100000000
7 11111110000000
8 11111111000000
9 11111111100000
10 11111111110000
11 11111111111000
12 11111111111100
13 11111111111110
Here's the array in EXCEL, after some editing. Add across the columns to confirm that the probabilities add up to the scores.
Score |
Item 4 |
Item 5 |
Item 6 |
Item 7 |
Item 8 |
Item 9 |
Item 10 |
Item 11 |
Item 12 |
Item 13 |
Item 14 |
Item 15 |
Item 16 |
Item 17 |
1 |
0.29 |
0.19 |
0.13 |
0.19 |
0.05 |
0.13 |
0.02 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
2 |
0.51 |
0.37 |
0.28 |
0.37 |
0.12 |
0.28 |
0.06 |
0.01 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
3 |
0.68 |
0.55 |
0.44 |
0.55 |
0.22 |
0.44 |
0.11 |
0.01 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
4 |
0.81 |
0.71 |
0.60 |
0.71 |
0.35 |
0.60 |
0.20 |
0.02 |
0.01 |
0.01 |
0.00 |
0.00 |
0.00 |
0.00 |
5 |
0.90 |
0.83 |
0.76 |
0.83 |
0.53 |
0.76 |
0.34 |
0.05 |
0.01 |
0.02 |
0.00 |
0.00 |
0.00 |
0.00 |
6 |
0.95 |
0.92 |
0.88 |
0.92 |
0.73 |
0.88 |
0.55 |
0.10 |
0.03 |
0.04 |
0.01 |
0.00 |
0.00 |
0.00 |
7 |
0.98 |
0.97 |
0.96 |
0.97 |
0.89 |
0.96 |
0.79 |
0.26 |
0.08 |
0.10 |
0.03 |
0.01 |
0.01 |
0.01 |
8 |
1.00 |
0.99 |
0.99 |
0.99 |
0.96 |
0.99 |
0.93 |
0.54 |
0.21 |
0.27 |
0.08 |
0.02 |
0.02 |
0.02 |
9 |
1.00 |
1.00 |
1.00 |
1.00 |
0.99 |
1.00 |
0.97 |
0.76 |
0.43 |
0.50 |
0.20 |
0.06 |
0.06 |
0.06 |
10 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
0.99 |
0.89 |
0.65 |
0.72 |
0.38 |
0.13 |
0.13 |
0.13 |
11 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
0.95 |
0.82 |
0.86 |
0.60 |
0.26 |
0.26 |
0.26 |
12 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
0.98 |
0.92 |
0.94 |
0.78 |
0.46 |
0.46 |
0.46 |
13 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
1.00 |
0.99 |
0.97 |
0.98 |
0.91 |
0.72 |
0.72 |
0.72 |