Score and measure output files simplify the transfer of measures and calibrations to other computer software. you may specify a file name that will be used to hold the numeric results presented in Facets measure tables. This is called interactively from the Output Files menu as Score and measure file output.
Score file name
|
What happens
|
Scorefile= filename.xls
Scorefile= filename.xlsx
Excel file
|
For Excel-formatted output files, the scores and measures for the element of each facet are written to a separate workbook.
|
Scorefile= filename.sav
SPSS file
|
For SPSS-formatted output files, the scores and measures for the elements of all facets are written to one SPSS .sav file, which includes facet number as a variable.
|
Scorefile= filename.rda
Scorefile= filename.rdata
R Statistics file
|
For R Statistics-formatted output files, the scores and measures for the element of each facet are written to a separate R data file.
|
Scorefile= filename with any other suffix or no suffix
CSV=
CSV=Tab
CSV=CSV
DOS Text file
|
The measures for each facet are written to a separate file, whose name is that specified by Scorefile=, with the facet number appended. E.g., if Scorefile=Results, then the measures for facet 1 are written to Results.1, for facet 2 to Results.2 etc.
The file format is specified by CSV
|
All fields, except the Label, are reported with two decimal places, and leading blanks. To select fields, use the Select Fields screen.
Heading lines are included or omitted, depending on Heading lines=.
QM quotation marks, controls whether labels are withing quotation marks.
Example: Write out score files for each facet in fixed field format:
Score file= kct
This puts the measures for facet 1 in file "kct.1.txt", for facet 2 in "kct.2.txt", etc. The results depend on the score file field selection, but can be:
1
|
Children
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
T.Score
|
T.Count
|
Obs.Avge
|
FairMAvge
|
Measure
|
S.E.
|
InfitMS
|
InfitZ
|
OutfitMS
|
OutfitZ
|
PtMea
|
PtMeExp
|
Discrim
|
Displace
|
Status
|
Group
|
1
|
Children
|
F-Number
|
F-label
|
7
|
18
|
0.29
|
0.05
|
-0.96
|
0.83
|
0.61
|
-1.19
|
0.29
|
1.99
|
0.8
|
0.75
|
1.71
|
-0.03
|
-1
|
-1
|
1
|
Boy
|
1
|
Children
|
10
|
18
|
0.5
|
0.43
|
1.74
|
1.11
|
4.1
|
2.47
|
6.15
|
1.94
|
0.65
|
0.87
|
-0.96
|
0
|
-1
|
-1
|
2
|
Boy
|
1
|
Children
|
The fixed field format is:
Field
|
Columns
|
Notes
|
0. Element number
|
241-250
|
Element number is the first field when "Element number after statistics" is not checked in the Score File Field Selection dialog box. Otherwise it is before the Element label.
|
1. Raw score
|
1-10
|
T.Score includes extreme scores, when Totalscore=Yes
Score excludes extreme scores, when Totalscore=No
|
2. Response count
|
11-20
|
T.Count includes extreme scores, when Totalscore=Yes
Count excludes extreme scores, when Totalscore=No
Elements with no observations are omitted unless Omit-unobserved=No
|
3. Observed average score
|
21-30
|
(Raw Score)/(Raw Count)
|
4. Fair average score
|
31-40
|
FairMAvge: Fair=Mean computes from the mean of the other facets
FairZAvge: Fair=Zero computes from the local zero of the other facets
|
5. Measure
|
41-50
|
Decimal places set with Umean=x,x,x
|
6. Standard error
|
51-60
|
Decimal places set with Umean=x,x,x
|
7. Infit mean-square
|
61-70
|
|
8. Infit standardized
|
71-80
|
|
9. Outfit mean-square
|
81-90
|
|
10. Outfit standardized
|
91-100
|
|
11. Point-measure correlation
|
101-110
|
or Point-biserial correlation. See Pt-biserial=
|
12. Point-measure expectation
|
111-120
|
.00 if not computed.
|
13. Discrimination
|
121-130
|
|
14. Displacement
|
131-140
|
Decimal places set with Umean=x,x,x. This is limited to the range -10 to +10 logits.
|
15. Status code
|
141-150
|
See Status Code list below
|
16. Group number
|
151-160
|
|
17. Weight
|
161-170
|
|
18. Sign
|
171-180
|
|
19. Infit Chi-squared
|
181-190
|
= infit information-weighted mean-square * d.f.
|
20. Infit Chi-squared d.f.
|
191-200
|
|
21. Infit Chi-squared 2-sided probability
|
201-210
|
2-sided probability of the chi-squared value. <.05 is significant underfit or overfit.
|
22. Outfit Chi-squared
|
211-220
|
= outfit conventional mean-square * d.f.
|
23. Outfit Chi-squared d.f.
|
221-230
|
|
24. Outfit Chi-squared 2-sided probability
|
231-240
|
2-sided probability of the chi-squared value. <.05 is significant underfit or overfit.
|
25. Element number
|
241-250
|
Element number is here when "Element number after statistics" is checked in the Score File Field Selection dialog box. Otherwise it is the first field.
|
26. Element label
|
251- ?
|
field length depends on the length of the element labels
|
27. Facet number
|
10 columns
|
field positions depends on the length of the element labels
|
28. Facet label
|
?
|
field length depends on the length of the facet label
|
Status code
|
Meaning for element measure
|
1
|
Anchored measure
|
0
|
No data available to estimate measure
|
-1
|
Measure estimated from data
|
-2
|
Measure estimated for extreme low score: XTREME=
|
-3
|
Measure estimated for extreme high score: XTREME=
|
-4
|
Unmeasurable, because all relevant data are generated by more than one extreme element
|
-5
|
Only one active data point for this element
|
Example: Compute the Spearman Reliability and Separation Index for score file.
1. Compute the sample S.D. of the Measures, then square the S.D. = Measure Variance (MV)
2. Square all the Standard errors and average the squares = Error Variance (EV)
3. Spearman (Kuder-Richardson, Cronbach) Reliability = (MV-EV) / MV
4. Separation Index = square-root ( (MV-EV) / EV )
