Simulated file specifications 
Winsteps uses two methods to simulate data:
1) probabilisticallygenerated data based on anchored parameter estimates
2) resamplingwithreplacement from the current dataset
Winsteps uses the estimated (or anchored) person, item and Andrich Thresholds or personresponsestring resamplingwithreplacement to simulate a data file equivalent to the raw data. This can be used to investigate the stability of measures, distribution of fit statistics and amount of statistical bias. Each time SIFILE= is run, or selected from the Output Files pulldown menu, a simulated data file produced. Do simulated analyses with several simulated datasets to verify their overall pattern.
The parts of the dialog box outside the red rectangle are described in Output file specifications. The file format matches the input data file if both are in fixedfield format. When SIFILE= is written with CSV=Y, commaseparated or CSV=T, tabseparated, the item responses precede the person label. 

Simulated data files: 
invoked with SIFILE= 
Number of files: 
SINUMBER=, specifies the number of simulated files to produce. If SINUMBER= is greater than 1, then the data file name is automatically incremented, and so is the SISEED= preset seed value 
Seed number (0 for random): 
SISEED=, controls whether the pseudorandom number generator is seeded with the system clock (0 or 1), or with a userchosen value, (2 and above) 
Simulate: use measures or use the data 
SIMEASURE=, chooses whether the simulated data is generated from the estimated measures (use measure), or by resampling from the observed data (use the data). If you wish to override the estimated measures, then use IAFILE=, PAFILE= and SAFILE= 
Resample persons: No or Yes: Persons 
SIRESAMPLE=, controls whether resampling occurs (sampling with or without replacement), and, if it does, how many person records to include in the simulated data file 
Complete data: Yes or No  allow missing data 
SICOMPLETE=, Yes for complete data. No for missing data patterns to be repeated in the simulated data file 
Extreme scores: Yes or No  avoid extreme scores 
SIEXTREME=, Yes to allow the simulated data to include extreme (zero, minimum possible or perfect, maximum possible) scores. No to avoid generating extreme scores (when possible) 
Winsteps simulates data two ways:
i) using the parameter values (persons, items, Andrich thresholds) from the current analysis. This way generates simulated data according to the probabilistic distributions defined by the Rasch model and the generating Rasch parameters. This is for investigations relating to exact Rasch conditions.
ii) by resampling (with replacement) the data (observations, responses) in the current analysis. This way generates data according to the empirical distribution of the generating data and is for investigations relating to datasets like this one.
Example 0 . KCT.txt simulated with CSV=N fixed field format (resampling response strings):
Dorothy F 111111111100000000 .2594 13
Elsie F 111101111100000000 1.3696 14
Thomas M 111111111010000000 .2594 31
Rick M 111111111010000000 .2594 27
KCT.txt simulated with commaseparated, CSV=Y, HLINES=Y, QUOTED=Y format (resampling person measures):
"14","23","124","134","214", ... ,"KID","Measure","Entry"
1,1,1,1,1,1,0,1,1,1,0,0,0,0,0,0,0,0,"Rick M",.2594,27
1,1,1,1,1,1,1,1,1,0,1,0,0,0,0,0,0,0,"Helen F",.2594,16
1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,0,0,"Rod M",1.9380,28
1,1,1,1,1,1,1,1,1,1,0,0,0,0,0,0,0,0,"William M",.9229,34
Example 1. It is desired to investigate the stability of Rasch measures.
(1) Estimate measures from your control and data files (e.g., SF.txt)
(2) Choose SIFILE= from the Output Files menu.
(3) Choose to output a permanent file:
(4) Simulated data filename: SFSIMUL.TXT
(5) Rerun Winsteps with your Winsteps control file and DATA=SFSIMUL.TXT on the "Extra Specifications" line.
(6) Compare person, item and Andrich Thresholds.
Example 2. To estimate the measure standard errors in a linked equating design.
1. Do a concurrent calibration with Winsteps
2. Simulate data files SIFILE= from the Output Files menu.
Specify "complete data" SICOMPLETE= as "No" to maintain the same data pattern.
Save 10 simulated sets, SINUMBER=, as S.txt S2.txt .....
3. Rerun your Winsteps analysis 10 times
Specify in Extra Specifications "DATA=S.txt PFILE=P1.txt CSV=TAB" etc.
This will produce 10 PFILE=s. Export them in Excel format.
4. Use Excel to compute the standard deviation of the measures for each person based on the 10 person measures
5. These are the model standard errors for the equating design for the measures.
6. Inflate these values by 20%, say, to allow for systematic equating errors, misfit, etc.
Example 3. If you do need estimationbiascorrection (STBIAS=) that is as accurate as possible with your data set, you will need to discover the amount of bias in the estimates and correct for it:
1. In your control file, STBIAS=No and USCALE=1
2. Obtain the Winsteps estimates for your data
3. Simulate many datasets using those estimates. (SIFILE= on the Winsteps Output Files menu).
4. Obtain the Winsteps estimates from the simulated data sets
5. Regress the simulated estimates on your initial estimates. These will give a slope near 1.0.
6. Obtain the Winsteps estimates for your data with USCALE = 1/slope. The set of estimates in 6 is effectively unbiased.
Example 4. You need to simulate data from generating values. You can use Winsteps to simulate a dataset.
1. Obtain the generating item difficulties, person abilities and threshold values. If you need a normal distribution of person abilities, you can generate this with Excel.
a. From your standard analysis, output IFILE=if.txt, SFILE=sf.txt
b. Use Excel or similar to simulate a normal distribution of person abilities with the mean and S.D. that you want.
In Excel:
Cell A1 = Mean
Cell B1 = S.D.
Cell A2 = =ROW()1
Cell B2 = =NORMSINV(RAND())*$B$1 +$A$1
then copy A2, B2 for as many rows as you want the sample size.
c. Copy Columns A and B into a text file, pf.txt. Delete row 1.
For a random uniform distribution of item difficulties, use the Excel formula:
=(item difficulty range)*(RAND()  0.5) + (mean item difficulty)
d. In your Winsteps control file:
IAFILE=if.txt
SAFILE=sf.txt
PAFILE = pf.txt
SIFILE= simulated.txt
2. Construct a Winsteps control file including the generating values in IAFILE=, PAFILE=, SAFILE=
3. Make a rectangular dataset with a row of valid responses (can be the same one) as wide as the number of items
and with a column of valid responses (can be the same one) as long as the number of persons.
For instance, number of persons = 2000, number of items =100, then an artificial dichotomous Winsteps control file and dataset can be:
ITEM1=1
NAME1=101
NI=100
CODES=01
EDFILE=*
12000 1100 1
*
IAFILE=*
1 2.3
2 1.5
.....
*
PAFILE=*
1 3.1
2 2.8
.....
*
SAFILE=*
0 0
1 0
*
&END
END LABELS
and nothing else.
4. Run Winsteps. Ignore the results of this analysis. Choose SIFILE= option from the output files menu. Click on "complete data" to simulate the entire data matrix.
Example 5. Bootstrap estimation of the confidence interval for a reliability.
Bootstrapping is based on generating new datasets using "sampling with replacement" from the current dataset. The new datasets can be generated in Winsteps using:
Simulate: use the data
Resample persons: yes, with same number of rows as the original data.
Compute the reliability of each new dataset. See "Performing multiple simulations in Batch mode".
After transformation with Fisher's z, the distribution of the linearized reliabilities (mean and S.D.) are the linearized expected value and linearized S.E. of the observed linearized reliability for the original data set. Transform the linearized (meanS.D.), mean, and (mean+S.D.) back to the original reliability metric by reversing the Fisher 's z transformation.
Example 6. Multiple simulations in Batch mode. See BATCH=
These can construct bootstrap confidence intervals for DIF estimates, etc.
Set up 100 simulations in a batch file, and extract the relevant numbers from the 100 output DIF tables.
PowerGREP is great software for extracting values from files. For instance:
To pick out lines 1035 in the output files (after line 9, for 26 lines):
Action type: Search
File sectioning: Search and collect sections
Section search: \A([^\r\n]*+\r\n){9}(([^\r\n]*+\r\n){0,26})
Section collect: \2
Search: the search string: .* for everything
1. Use NotePad to create a text file called "Simulate.bat"
2. In this file:
REM  produce the generating values: this example uses example0.txt:
START /WAIT c:\Winsteps\Winsteps BATCH=YES example0.txt example0.out.txt PFILE=pf.txt IFILE=if.txt SFILE=sf.txt
REM  initialize the loop counter
set /a test=1
:loop
REM  simulate a dataset  use anchor values to speed up processing (or use SINUMBER= to avoid this step)
START /WAIT c:\Winsteps\Winsteps BATCH=YES example0.txt example0%test%.out.txt PAFILE=pf.txt IAFILE=if.txt SAFILE=sf.txt SIFILE=SIFILE%test%.txt SISEED=0
REM  estimate from the simulated dataset
START /WAIT c:\Winsteps\Winsteps BATCH=YES example0.txt data=SIFILE%test%.txt SIFILE%test%.out.txt pfile=pf%test%.txt ifile=if%test%.txt sfile=sf%test%.txt
REM  do 100 times
set /a test=%test%+1
if not "%test%"=="101" goto loop
PAUSE
3. Save "Simulate.bat", then doubleclick on it to launch it.
4. The simulate files and their estimates are numbered 1 to 100.
5. The files of estimates can be combined and sorted using MSDOS commands, e.g.,
Copy if*.txt combinedif.txt
Sort /+(sort column) <combinedif.txt >sortedif.txt
6. Individual lines from the output files can be written to one file using MSDOS batch commands. For instance, using an MSDOS batch routine (.bat or .cmd), the same text line can be extracted from many text files and output into a new text file. The new text file can be be pasted into Excel. Save these MSDOS commands as extract.bat in the folder that has the files of statistics. Double click on extract.bat to execute it.
rem replace 2 with the number of lines to skip before the line you want
@echo off
setlocal EnableDelayedExpansion
if exist result.csv del result.csv
for %%f in (*.txt) do (
echo %%f
set i=a
for /F "skip=2 delims=" %%l in (%%f) do (
if "!i!" == "a" echo %%f, %%l >> result.csv
set i=b
)
)
notepad result.csv
Help for Winsteps Rasch Measurement Software: www.winsteps.com. Author: John Michael Linacre
For more information, contact info@winsteps.com or use the Contact Form
Facets Rasch measurement software.
Buy for $149. & site licenses.
Freeware student/evaluation download Winsteps Rasch measurement software. Buy for $149. & site licenses. Freeware student/evaluation download 

Stateoftheart : singleuser and site licenses : free student/evaluation versions : download immediately : instructional PDFs : user forum : assistance by email : bugs fixed fast : free update eligibility : backwards compatible : money back if not satisfied Rasch, Winsteps, Facets online Tutorials 

Forum  Rasch Measurement Forum to discuss any Raschrelated topic 
Click here to add your email address to the Winsteps and Facets email list for notifications.
Click here to ask a question or make a suggestion about Winsteps and Facets software.
Coming Raschrelated Events  

Jan. 5  Feb. 2, 2018, Fri.Fri.  Online workshop: Practical Rasch Measurement  Core Topics (E. Smith, Winsteps), www.statistics.com 
Jan. 1016, 2018, Wed.Tues.  Inperson workshop: Advanced Course in Rasch Measurement Theory and the application of RUMM2030, Perth, Australia (D. Andrich), Announcement 
Jan. 1719, 2018, Wed.Fri.  Rasch Conference: Seventh International Conference on Probabilistic Models for Measurement, Matilda Bay Club, Perth, Australia, Website 
Jan. 2224, 2018, MonWed.  Inperson workshop: Rasch Measurement for Everybody en español (A. Tristan, Winsteps), San Luis Potosi, Mexico. www.ieia.com.mx 
April 1012, 2018, Tues.Thurs.  Rasch Conference: IOMW, New York, NY, www.iomw.org 
April 1317, 2018, Fri.Tues.  AERA, New York, NY, www.aera.net 
May 22  24, 2018, Tues.Thur.  EALTA 2018 preconference workshop (Introduction to Rasch measurement using WINSTEPS and FACETS, Thomas Eckes & Frank WeissMotz), https://ealta2018.testdaf.de 
May 25  June 22, 2018, Fri.Fri.  Online workshop: Practical Rasch Measurement  Core Topics (E. Smith, Winsteps), www.statistics.com 
June 27  29, 2018, Wed.Fri.  Measurement at the Crossroads: History, philosophy and sociology of measurement, Paris, France., https://measurement2018.sciencesconf.org 
June 29  July 27, 2018, Fri.Fri.  Online workshop: Practical Rasch Measurement  Further Topics (E. Smith, Winsteps), www.statistics.com 
July 25  July 27, 2018, Wed.Fri.  PacificRim Objective Measurement Symposium (PROMS), (Preconference workshops July 2324, 2018) Fudan University, Shanghai, China "Applying Rasch Measurement in Language Assessment and across the Human Sciences" www.promsociety.org 
Aug. 10  Sept. 7, 2018, Fri.Fri.  Online workshop: ManyFacet Rasch Measurement (E. Smith, Facets), www.statistics.com 
Sept. 3  6, 2018, Mon.Thurs.  IMEKO World Congress, Belfast, Northern Ireland www.imeko2018.org 
Oct. 12  Nov. 9, 2018, Fri.Fri.  Online workshop: Practical Rasch Measurement  Core Topics (E. Smith, Winsteps), www.statistics.com 
Our current URL is www.winsteps.com
Winsteps^{®} is a registered trademark
Concerned about aches, pains, youthfulness? Mike and Jenny suggest Liquid Biocell 
