Hello, Dr Faulkenberry. Your video is so helpful. Thanks for uploading this tutorial. I have a quick question, The MI showed that I should add residual covaries inQ13 and Q14. So my question is how can I put this in real questionnaires
Hi April! I'm glad you found the video helpful. Residual covariance between items (these are technically called "indicators" of your factors) is often added to improve model fit -- that is because it helps to recover some of the actual covariance between items that is not captured by the model itself. Adding things like this will almost always improve model fit, at the expense of making the model more complicated (i.e., more parameters). As far as "putting in real questionnaires"...that's not really the best question, as the only things you can "put in the questionnaire" are the items themselves. Instead, you can ask yourself "does this residual covariance help me understand the hidden structure of what I"m trying to measure?". I hope this helps!
Hi, Dr Faulkenberry. Your video is so helpful. Thank you so much.
Hello, Dr Faulkenberry. Your video is so helpful. Thanks for uploading this tutorial. I have a quick question, The MI showed that I should add residual covaries inQ13 and Q14. So my question is how can I put this in real questionnaires
Hi April! I'm glad you found the video helpful. Residual covariance between items (these are technically called "indicators" of your factors) is often added to improve model fit -- that is because it helps to recover some of the actual covariance between items that is not captured by the model itself. Adding things like this will almost always improve model fit, at the expense of making the model more complicated (i.e., more parameters). As far as "putting in real questionnaires"...that's not really the best question, as the only things you can "put in the questionnaire" are the items themselves. Instead, you can ask yourself "does this residual covariance help me understand the hidden structure of what I"m trying to measure?". I hope this helps!
@@TomFaulkenberry Thank you so much,Dr Faulkenberry