[Rasch] Measures for groups

Juho Looveer juho.looveer at gmail.com
Thu Jan 20 11:44:51 EST 2011


Hmm a bit of a puzzler where I need some suggestions.



Literacy and Numeracy tests were administered to a country last year.

For each item, student responses were coded as Level 0 (bottom; indicating
no skill or knowledge) up to Level 5 (top level, indicating mastery or high
level achievement).


To avoid comparison of students, the data was aggregated by gender within
schools, so that we have two lines of data for each school, one being for
males and the second for females.

e.g.

In the table below, at school B, 5 males were assessed as being at Level 1,
but zero females at Level 1.



*OPERATIONS*

* *

* *

*Level 0*

*Level 1*

*Level 2*

*Level 3*

*Level 4*

*Level 5*

*Total*

School A

males













0

females













0

School B

males

2

5

8

6

4

1

26

females

1

0

7

9

3

2

22

School C

males













0

females













0





There would have been similar tables for 4 other areas in numeracy (e.g.
measurement, number, space, proportion).



I have been asked if I can provide some Rasch measures on the fit of the
students to the items, and the fit of the items to the construct.



I do NOT have access to the actual student level data.



My first thought was that this can’t be done.



Any ideas on how to progress with this issue?





Thanks



Juho





*Dr  Juho Looveer*

Psychometrician / Analyst

“Improving the Assessment of Literacy and Numeracy Across the Pacific”

Secretariat of the Pacific Board for Educational Assessment

26 McGregor Road

PO Box 2083, Government Buildings,  Suva, Fiji

Phone: +679-368 0051

Mobile: (00679) 945 6055

Fax: +679-3302898

Email: jlooveer at spbea.org.fj <mshafraz at spbea.org.fj>

Web Site: www.spbea.org.fj
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20110120/d0891c93/attachment.html 


More information about the Rasch mailing list