hyunsooseol / snowirt

This module includes Item Statistics, Model fit, Differential Item Functioning, Wright Map, Expected Scores Curve,and Item Characteristic Curve for DIF using MML estimation of the Rasch measurement model. Furthermore you can analyze DIF, Distractor analysis and Many facet Rasch model.

snowIRT

This package provides distractor analysis, item and person statistics, Model fit, Differential Item Functioning, Wright Map, Expected Score Curve, and Item Characteristic Curves for DIF using the Rasch measurement model.

• These anlalyses are described in the book; https://bookdown.org/dkatz/Rasch_Biome/

snowIRT module can be installed and used in R as a standard R packages.

Installation

At the moment, snowIRT is not in CRAN yet, so you need to install it via devtools

devtools::install_github("hyunsooseol/snowIRT")

snowirt's Issues

Missing value where TRUE/FALSE needed

Dear Author,

Thank you for the add on Rasch analysis with jamovi (1.6.3). I am trying to test the snowIRT. When I open the snowIRT (dichotomous) and enter the dichotomous variables I get an error message "Missing value where TRUE/FALSE needed", even though there are no missing values. How do I proceed with the analysis?

Thanks a lot,

Nikolaos Tsigilis

Error in Jamovi - 'from' must be a finite number

Hi everyone,

I am new to Jamovi and trying to learn how to run IRT analyses with polytomous items. After importing my dataset to Jamovi, I get a 'from' must be a finite number error. Would you be able to help me ? I have also attached a screenshot of the error message.

Thank you!

'Breaks are not unique' message. Could you please help

Dear Professor,
I work with distructor analysis. I have 10 multiple choice questions and the key for each question but I take a message reading " 'breaks' are not unique ". My variables are nominal. I have tried data as 1, 2, 3, 5 and also a, b, c, d, e. I wonder if you could help me.
Best Regards,
Athanasios

QuestionAboutScoreTheResponse

Hello Soel, it's Larry Nelson here in Australia, experimenting with having Distractor Analysis take in and score raw test item responses (each item having options of 1,2,3,4) and output corresponding item scores of (0,1) for each student on each item.

The (0,1) items scores will then go into the Dichotomous Option.

"Score the response" does seem to do something (there's a pause of a couple of seconds) but I cannot find its output.

Not urgent at all, regards, thanks, Larry

Result generation issues with NA

Hola, Seol. Thank you for your contributions.

As observed below, I have more than 600 items in the database, categorized as continuous variables (scale) 0 and 1, and missing data marked as NA for each item. These are student evaluations where not all students answer the complete question bank, but only 60 items. When performing the dichotomous Rasch analysis, results are shown for some items, while for others, the message "length of 'dimnames' [3] not equal to array extent" is displayed.

For example, in the image, when entering item 3112, the error message "length of 'dimnames'not equal to array extent" appears. Analyzing the behavior of this item, it has fewer missing data than item 3111, and the standard deviation is similar to the introduced items that yield results (items 3108, 3109, 3110, 3111). So, I cannot find an explanation for this error because missing data is marked with NA. I appreciate your support in finding a solution. In Jmetrik, the same database works without issues.

Jamovi - WrightMap

Could you please add the feature to show the item names (exact names) in the wrightmap?
Could you please add the feature to show the dimensions (for example 3 factors in a test) and the related items (for example 10 items for each factor), together in Wrightmap? (I mean not separate maps, they can be seen together)..

Item Plot in MFRM

Hello. I am doing a many facet rasch model analysis with Jamovi. I want to see the item response curve graph for each item but the graph does not appear. Can you fix it? Thank you.

Distractor Analysis comments

A big hello from Western Australia, and also a thank you for the work you've done with Distractor Analysis.

https://lertap5.com/Documentation/DistractorAnalysisWithJamovi.pdf

The pdf above has comments made as I experimented with Distractor Analysis.

Yours, Larry

(Larry Nelson)

Break are not Unique

Dear Hyun,

I use Distractor analysis of snowIRT in Jamovi. I have problem and some warning written "Break are not unique". is it common problem? how do I fix that?

thank you very much

Unable to access 'pinfit' and 'poutfit' results when using the 'polytomous()' function

Hello,

I've been trying to use the polytomous() function from the snowIRT package in R, but I'm facing an issue when trying to access the 'pinfit' results.

Here's the code I'm using:

results <- polytomous(
    data = selected_df,
    vars = c("INTEGRITAS.1", "INTEGRITAS.2", "INTEGRITAS.3", "INTEGRITAS.4", "INTEGRITAS.5", "INTEGRITAS.6", "INTEGRITAS.7", "INTEGRITAS.8"),
    infit = TRUE,
    outfit = TRUE
)

When I run results$pinfit and results$poutfit, it doesn't return any output. However, the 'pinfit' and 'poutfit' element does appear as an 'active binding' when I inspect the results object using str(results).

I noticed that the snowIRT module is used in the jamovi software and it can produce the infit and outfit. I'm trying to replicate this in R, but facing the above issue.

Could you help me understand why I'm unable to access the 'pinfit' results and how I can resolve this issue? Any guidance would be greatly appreciated.

Thank you.