[Rasch] 回复：Re: Rasch analysis of mixed format data using jMetrik
chkollias at outlook.com
Wed Dec 15 09:16:01 AEDT 2021
I missed this the first time reading your email… Do you have only one polytomous item? If so, then that’s why it’s dropped. I cannot recall whether jmetrik allows you to weight items. If not, you may need to try another software programme that allows weighting of items (e.g. Winsteps). With weighting you could split the one polytomous item into two and weight both items by 0.5.
But the data also needs to be entered in a special way for each of these two items.
See if jmetrik allows you to weight items or see if you have access to other software that does weighting. If so, contact me off list to discuss this further.
From: Yingchen <2596011329 at qq.com>
Sent: Tuesday, December 14, 2021 9:26:34 PM
To: Charalambos Kollias <chkollias at outlook.com>; rasch <rasch at acer.edu.au>
Subject: 回复：Re: [Rasch] Rasch analysis of mixed format data using jMetrik
Yes, it is correct. I checked. Exactly as you specified.
For the binary items, the original data is 0/1. So, it is easy to input it.
But the first time, I input the theoretical range (0-20), and clicked on the refresh button. After finding the error, I went back to change it to the empirical range (4-19), and refreshed it. Checked it in variable view, and it was correct.
Either way, jMetrik output window showed the range of 0-20. “ Dropped”.
收件人："Yingchen "<2596011329 at qq.com>,"rasch"
主题：Re: [Rasch] Rasch analysis of mixed format data using jMetrik
As it is hard to ascertain where the issue is in jmetrik - let's take it step by step.
Let's begin by checking whether key has been inserted correctly.
"Transform" - "Advanced Item Scoring" -
Click on item 1
Click arrow ">" (item 1 should be in box on right side)
Set scoring as follows: (left had box)
In "scoring syntax" text box you should have:
Options = (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19), scores = (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19), variables = (item1)
Go to "variables" tab
"Refresh Data View"
Does item 1 in type column say "polytomous item"?
Has key been entered correctly?
From: Rasch <rasch-bounces at acer.edu.au> on behalf of Yingchen <2596011329 at qq.com>
Sent: Tuesday, December 14, 2021 2:52 PM
To: Yingchen <2596011329 at qq.com>; rasch <rasch at acer.edu.au>
Subject: [Rasch] Rasch analysis of mixed format data using jMetrik
Hi, dear Rasch users:
I used jMetrik to analyze the mixed data, a combination of binary items and polytomous item. The polytomous item is scored by human raters, ranging from 4 to 19. I tried to input the practical range (4-19) as well as the theoretical range (0-20). None of the efforts succeeded. Each time, the polytomous item was dropped from the analysis.
Item formats are correct. Between 4 and 19, "5" is missing.
I checked at jMetrik online instruction, but didn't find any solution. Google is not available in China.
Would appreciate if anyone can help.
发件人:"Yingchen "<2596011329 at qq.com>
收件人:"rasch"<rasch at acer.edu.au>;
主题:Questions about partial credit model
Some simple questions.
(1). What is the maximum number of categories that partial credit model can handle? I recall that some professor said 12. Not 100% positive.
(2) What is the maximum number of categories that our current Rasch programs (Winsteps, jMetrik, conquest, etc) can handle?
(3) If my memory is correct (12 as maximum number of categories) and that is the limit of our current programs, are there any ways to bypass this limit? Certainly, the purpose is to run a program on polytomous data with more categories.
Thank you for your help.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Rasch