Hi,
I am developing a psychometric scale. I have three factors and I have been told by my supervisors that I won't be able to run an IRT because there is more than one latent variable. Furthermore, I can't run an IRT on each factor to inspect the items because I have too few per factor (4, 4, 3).
Can anyone help me explain further why this is the case, and are there any references that people can recommend?
Many thanks,
Leng
Dear Leng, The multidimensional random coefficients multinomial logit model (MRCMLM) was proposed by Adams, Wilson, & Wang (1997). The approach integrates the Partial Credit Model (Masters, 1982) and is grounded in the 1-parameter logistic (1PL) IRT model, commonly referred to as the Rasch Model (Rasch, 1960). The model is applied when multiple dimensions are present within a single overarching construct.
The number of items within each of your domains seems to be the larger issue and will affect the size of the standard errors of your estimates. Consider adding more items to each domain. From there, you can decide whether to run a unidimensional or multidimensional model.
Keep us posted on your progress!
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running