Control file and template.txt

Top Up Down  A A

The control file tells what analysis you want to do. The template file, TEMPLATE.TXT. gives you an outline to start from. The easiest way to start is to look at one of the examples in the next section of this manual, or on the program disk. The control file contains control variables. These are listed in the index of this manual. Only two control variables must have values assigned for every analysis: NI= and ITEM1=. Almost all others can be left at their automatic standard values, which means that you can defer learning how to use most of the control variables until you know you need to use them.

 

When in doubt, don't specify control variables, then they keep their standard values.

 


 

Here is a version of TEMPLATE.TXT. Copy and paste this, if your TEMPLATE.TXT is corrupted.

 

; this is a WINSTEPS specification control file template.

; Save it with your own name, e.g., control.txt

 

; a semi-colon means a comment: remove semi-colons as needed.

 

&INST ; optional

 

 TITLE = "Put your page heading here"

 

;Input Data Format

 NAME1 = 1       ; column of start of person information

NAMLEN = 30       ; maximum length of person information

 ITEM1 = ?        ; column of first item-level response

    NI = ??       ; number of items = test length

 XWIDE = 1        ; number of columns per response

PERSON = Person   ; Persons are called ...

  ITEM = Item     ; Items are called ...

 

; DATA =          ; data after control specifications

 

; For rescoring

;       0        1         2         3         4         5         6         7

;       1234567890123456789012345678901234567890123456789012345678901234567890

;ISGROUPS=0        ; specify that each item has its own response structure (partial credit)

;IREFER=AABBCC....

 

; Data Scoring

  CODES = "01"    ; valid response codes

;IVALUEA= "01"    ; for rescoring for item type A

;IVALUEB= "10"    ; for rescoring for item type B

;IVALUEC= "  "    ; for rescoring for item type C

;                 ; Codes in IREFER with no IVALUE are not changed

 

CLFILE = *          ; label the categories in Table 3.2

0 Strongly Disagree ; 0 in the data means "Strongly Disagree"

1 Strongly Agree    ; 1 in the data means "Strongly Agree"

*

 

;NEWSCORE = "10"    ; use to rescore all items

;    KEY1 =         ; key for MCQ items

 

; XWIDE = 2   ; for all codes 00 to 99

; CODES = "000102030405060708091011121314151617181920212223242526272829+

;         +303132333435363738394041424344454647484950515253545556575859+

;         +606162636465666768697071727374757677787980818283848586878889+

;         +90919293949596979899"

 

; codes reversed, in case needed

; NEWSCR= "999897969594939291908988878685848382818079787776757473727170+

;         +696867666564636261605958575655545352515049484746454443424140+

;         +393837363534333231302928272625242322212019181716151413121110+

;         +09080706050403020100"

 

; MISSCORE = -1     ; -1 = missing data treated as not administered

 

;User Scaling

UMEAN  = 50             ; user-set item mean - standard is 0.00

USCALE = 10             ; user-scaled measure units - standard is 1.00

UDECIM = 1              ; reported decimal places - standard is 2

MRANGE = 50             ; half-range on maps - standard is 0 (auto-scaled)

 

&END

 

;Put item labels here for NI= lines

 

END LABELS

 

;Put data here - and delete this comment to prevent it being processed as a data line.