Rasch measurement unidimensionality and local independence (Part 1)

[@richardarnett2138]

Very useful. Looking forward to complete version of tutorial 2.

[@VahidAryadoust]

It is available from here: kzbin.info/www/bejne/aIHKgKuDrppnn7M

[@soehartosoeharto8471]

Dear prof. Vahid To be honest. I hope there is a video about Stacking and Racking in this channel Thank you for your help. I hope you and your family are always healthy and get blessing by God. Thank to help us

[@VahidAryadoust]

Thanks for the suggestion. I has been on my to-do list.

[@soehartosoeharto8471]

@@VahidAryadoust thanks Prof

[@MichaelMenke]

any videos on monotonicity?

[@VahidAryadoust]

not yet; I will make a video on rating scale models and discuss monotonicity.

[@sandamaliwijekoon202]

Dear Prof.Vahid, I am self learning about RASCH analysis to apply it in a validation research under my postgraduate studies. I am a beginner in this field and your videos on Rasch analysis are great help for me. In some research papers I read, results on unidimensionality were presented under unidimensional t tests (CI). I further read that "strict unidimensionality " can be formally tested by conducting independent t tests to compare the person logit estimates derived from subsets consisting all positively or negatively loaded items on the first residual component(Smith,2002). Can you please kindly explain me about this "person logit estimates derived from subsets consisting of all positively or negatively loaded items of the first residual component". Does WINSTEP has an option to run t test or do I have to run it with SPSS? Your kind help is greatly appreciated.

[@VahidAryadoust]

Find out what items load positively and negatively on the first contrast. Next, run separate analysis, one including the negative-load items and the other on the positive-load items. You will get two estimates of person ability (one per analysis); next, compare them using a paired samples t test.

[@sandamaliwijekoon202]

@@VahidAryadoust Thank you very much Professor.

[@MoonLight-nc4fi]

Dear Dr. Vahid, thank you so much for your generous videos. I am new to facets. Do you have any video/tips for analysing PCA unidimensionality for analytical rubrics? Thanks a lot

[@VahidAryadoust]

Please also watch the second installment of this video. And the following paper will also be of help: journals.sagepub.com/eprint/YXCQB3NYCICYVFQ4TSCG/full

[@MoonLight-nc4fi]

@@VahidAryadoust hi Dr, Vahid. Thank you for your reply. I have read this paper and your other papers as well. I watched the 2nd video. Unfortunately, it does not show how the data (analytical rubric scores? from facets can be transferred to winsteps for unidimensionality and pca analysis. I appreciate if you could consider this in your videos. Thanks a lot

[@VahidAryadoust]

@@MoonLight-nc4fi I see what you mean now. Generally, it is not recommended to conduct unidimensionality for facets data because of the effect of different facets and missing data.

[@MoonLight-nc4fi]

@@VahidAryadoust thanks Dr. Vahid.

[@npd05]

Hi Vahid. Many thanks for your first tutorial - I thought your first tutorial was very informative and useful. I was very much looking forward to your second tutorial. And as I was watching it, I was looking forward to hearing your thoughts on the difference between PCA-R and traditional PCA, but the video stopped short. :( If you have time, could you upload a full version. Cheers!

[@VahidAryadoust]

Hi Nigel, thanks for your interest in the videos. I have noted the abrupt ending of the video. I am looking into it. You should see an update soon. :)

[@npd05]

Vahid Aryadoust great. I hope you will be able to.sort it out. :)

[@VahidAryadoust]

Nigel, here is the link to the second part of the video: kzbin.info/www/bejne/aIHKgKuDrppnn7M

[@npd05]

Just watched it. It was very informative and worth the wait. Many thanks. :)

[@VahidAryadoust]

Good to hear that!

[@VahidAryadoust]

Just an update: Our comprehensive review of Rasch measurement in language assessment has been published: journals.sagepub.com/eprint/YXCQB3NYCICYVFQ4TSCG/full