[Rasch] Funny codes problem in Bigsteps

Stuart Luppescu s-luppescu at uchicago.edu
Tue Aug 29 07:40:04 EST 2006


Hello, It seems that either Bigsteps is ignoring the codes directive, or
I'm doing something very dumb.

We're trying to put together a measure of teacher mentor value from
survey items that asked new teachers how helpful their mentors were in
various areas. However, some teachers do not have mentors, so we coded
all the items for those teachers `1' and increased the codes for the
other teachers by 1. So, ``No mentor'' is coded 1; ``did not receive
help'', 2; ``received help but it was not valuable'', 3; ``received help
and it was somewhat valuable'', 4; etc. Of course, all the teachers who
do not have mentors will get the minimum extreme measure, but that will
be better than arbitrarily giving them some low number.

The control file has codes=12345, but people who don't have mentors
don't get any measure at all. Here's one such person:

23    0.00 -2     0      0   0.00   1.00   0.00   1.00   0.00   0.00   0.00 7940 E794049508 11111111

The -2 in the status column indicates that this teacher has no valid
responses. Also, the step table does not include category `1':

SUMMARY OF MEASURED STEPS
+----------------------------------------------+
|CATEGORY OBSERVED|AVGE  INFIT OUTFIT|  STEP   |
| LABEL    COUNT  |MEASURE MNSQ  MNSQ| MEASURE |
|-----------------+------------------+---------|
|   2         804 | -1.38  1.19  1.46|  NONE   |
|   3         607 | -0.71  0.67  0.58|   -0.70 |
|   4        2234 |  0.65  0.89  0.88|   -1.27 |
|   5        1291 |  2.35  1.04  1.00|    1.97 |
+----------------------------------------------+

Is this correct? Am I misunderstanding something?

Thanks.

-- 
Stuart Luppescu -=-=- s-luppescu <AT> uchicago <DOT> edu
CCSR at U of C ,.;-*^*-;.,  www.consortium-chicago.org
     (^_^)/    才文と智奈美の父
Thank God I'm an atheist!
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 3437 bytes
Desc: not available
Url : https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20060828/fef80008/attachment.bin 


More information about the Rasch mailing list