I have a dataset where there are 5 concepts and each concept has multiple items and there are student observations for the items ie if they attempted it correctly or not. Now while running pyirt I have these questions.
Is there any way to set the default values of alpha, beta, theta initially? From what I could learn that you have just mentioned bounds.
How is the algorithm running? I mean does it run on the entire database which has multiple concepts or is it running on individual concepts that have multiple items.
I ran the algorithm at concept level for beta range of 0-1 and found that beta values come out mostly 0. Do you know any reason why this could be happening?
How did you come up with the default values and what is the explanation for it?
adding our data for actual vs predicted output by the pyirt library -
I have a dataset where there are 5 concepts and each concept has multiple items and there are student observations for the items ie if they attempted it correctly or not. Now while running pyirt I have these questions.
Is there any way to set the default values of alpha, beta, theta initially? From what I could learn that you have just mentioned bounds.
How is the algorithm running? I mean does it run on the entire database which has multiple concepts or is it running on individual concepts that have multiple items.
I ran the algorithm at concept level for beta range of 0-1 and found that beta values come out mostly 0. Do you know any reason why this could be happening?
How did you come up with the default values and what is the explanation for it?
adding our data for actual vs predicted output by the pyirt library -
@mvj3 @bryketos @bybunni @junchenfeng @wxiaoguang