Hello Everyone,
I am working on an IRT analysis using the graded response model to evaluate a scale. In the reference manual there is a postestimation command that provides the predicted theta (latent trait) value for each respondent in the dataset. Is there a similar command that will provide the "expected score" for the entire scale for each respondent using the IRT model, and save that score into a new variable. I keep thinking there has to be a way given the graph option of the test characteristic curve, which plots the expected scores on a scale or test given a range of thetas. I'd just like to have an actual variable in the file that has the total expected score as predicted by the fitted graded response model for each participant.
I am working on an IRT analysis using the graded response model to evaluate a scale. In the reference manual there is a postestimation command that provides the predicted theta (latent trait) value for each respondent in the dataset. Is there a similar command that will provide the "expected score" for the entire scale for each respondent using the IRT model, and save that score into a new variable. I keep thinking there has to be a way given the graph option of the test characteristic curve, which plots the expected scores on a scale or test given a range of thetas. I'd just like to have an actual variable in the file that has the total expected score as predicted by the fitted graded response model for each participant.
Comment