[Rasch] I have two very different instruments that I think measure the same thing.

Parisa Daftari Fard pdaftaryfard at yahoo.com
Fri Apr 22 09:17:31 EST 2011


Dear Michael,
 
I had the same problem in estimating two sets of items with different scale. I hope Dr. Linacre corrects me if I am wrong but I guess you should follow something like this
 
&INST            ; initial line (can be omitted)
  TITLE = "Bond & Fox BLOT data: Appendix 2" 
PERSON = Person ; persons are ...
   ITEM = Item   ; items are ...
  ITEM1 = 5      ; column of response to first item in data record
     NI = 52     ; number of items
  NAME1 = 1      ; column of first character of person label
NAMELEN = 3      ; length of person identifyinglabel
  XWIDE = 1      ; number of columns per item response
  CODES = 12345    ; valid codes in data file
    DPF = $S1W3  ; task codes are in the first 3 columns of the item label
ISUBTOT = $S1W3 
ISGROUPS = AAAAAAAAAAAAAAAAAAAABBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB  ; items 1-20 and items 21-52 have different rating scales
 UIMEAN = 0      ; item mean for local origin
 USCALE = 1      ; user scaling for logits
 UDECIM = 2      ; reported decimal places for user scaling
  TOTAL = Yes    ; show total raw scores
  CHART = Yes    ; produce across-pathway picture
   MNSQ = No     ; use Standardized fit statistics
CONVERGE= L      ; Convergence decided by logit change
LCONVERGE=.00001 ; Set logit convergence tight because of anchoring
IAFILE = *       ; Item anchor file to preset the difficulty of an item
4 0              ; Item 4 exactly at 0 logit point.
*                ; End of anchor list
&END
12345 01 ; g1
12345 02 ; g2
12345 03 ; g3
12345 04 ; g4
12345 05 ; g5
12345 06 ; g6
12345 07 ; g7
12345 08 ; g8
12345 09 ; g9
12345 10 ; g10
12345 11 ; g11
12345 12 ; g12
12345 13 ; g13
12345 14 ; g14
12345 15 ; g15
12345 16 ; g16
12345 17 ; g17
12345 18 ; g18
12345 19 ; g19
12345 20 ; g20
123   21 ; g21
123   22 ; g22
123   23 ; g23
123   24 ; g24
123   25 ; g25
123   26 ; g26
123   27 ; g27
123   28 ; g28
123   29 ; g29
123   30 ; g30
123   31 ; g31
123   32 ; g32
123   33 ; g33
123   34 ; g34
123   35 ; g35
123   36 ; g36
123   37 ; g37
123   38 ; g38
123   39 ; g39
123   40 ; g40
123   41 ; g41
123   42 ; g42
123   43 ; g43
123   44 ; g44
123   45 ; g45
123   46 ; g46
123   47 ; g47
123   48 ; g48
123   49 ; g49
123   50 ; g50
123   51 ; g51
123   52 ; g52
END LABELS
 
I cut and paste the syntax I wrote for my own data. Read Fox and Bond and compare this with that of yours. I have seen T test is used to see if the two tests are measuring the same or different things. I guess you use both Rasch and ttest.
 
Best,
Parisa Daftarifard
 

--- On Thu, 4/21/11, Michael Lamport Commons <commons at tiac.net> wrote:


From: Michael Lamport Commons <commons at tiac.net>
Subject: Re: [Rasch] I have two very different instruments that I think measure the same thing.
To: "Rense Lange" <rense.lange at gmail.com>
Cc: "Andrew M. Richardson" <andrewmrichardson at gmail.com>, Rasch at acer.edu.au
Date: Thursday, April 21, 2011, 9:46 PM


They are of different forms.  One is on a 1-6 scale, the other right and wrong
My Best,

Michael Lamport Commons, Ph.D.
Assistant Clinical Professor

Department of Psychiatry
Beth Israel Deaconess Medical Center
Harvard Medical School
234 Huron Avenue
Cambridge, MA 02138-1328

Telephone   (617) 497-5270
Facsimile   (617) 491-5270
Cellular    (617) 320–0896
Commons at tiac.net
http://dareassociation.org/



On 4/21/2011 1:09 PM, Rense Lange wrote: 
Why not scale them (both tests = all items) together? If they really measure the same thing, everything should fit. And, in this case you just have defined a more reliable test of the construct.  


No problem if some items don't fit, then you learned something useful also ... 


Rense


2011/4/21 Michael Lamport Commons <commons at tiac.net>


I have two very different instruments that I think measure the same thing.  What tests do you think would be useful in comparing person scores.  Note that the means will be close but the standard deviations might be more different.  Should a simple t-test do?
-- 
My Best,

Michael Lamport Commons, Ph.D.
Assistant Clinical Professor

Department of Psychiatry
Beth Israel Deaconess Medical Center
Harvard Medical School
234 Huron Avenue
Cambridge, MA 02138-1328

Telephone   (617) 497-5270
Facsimile   (617) 491-5270
Cellular    (617) 320–0896
Commons at tiac.net
http://dareassociation.org/



_______________________________________________
Rasch mailing list
Rasch at acer.edu.au
Unsubscribe: https://mailinglist.acer.edu.au/mailman/options/rasch/rense.lange%40gmail.com



-- 
Rense Lange, Ph.D.
via gmail

-----Inline Attachment Follows-----


_______________________________________________
Rasch mailing list
Rasch at acer.edu.au
Unsubscribe: https://mailinglist.acer.edu.au/mailman/options/rasch/pdaftaryfard%40yahoo.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20110421/8cb2b743/attachment.html 


More information about the Rasch mailing list