Message/Author |
|
sharon su posted on Sunday, September 25, 2016 - 5:31 am
|
|
|
Hello ~ I have a set of categorical(binary) data, but I don't know whether ex5.2 or ex5.5 should I use? What's the difference between the two analysis method(ex5.2 & ex5.5)? Do they have different link function assumption? Thanks ~ |
|
|
5.2 uses weighted least squares estimation and probit regression. 5.5 uses maximum likelihood estimation and logistic regression. |
|
|
Hello, I am using CFA with WLSMV and if I'm correct, theoretically I am carrying out an IRT. Because I use mac, I do not get the plots of the estimated model. However, I also do not get any information on the item discrimination and item difficulties in the output. Is this due to having mac or something missing in my input: VARIABLE: CATEGORICAL = ALL; NAMES = dw1-dw36; USEVARIABLES = dw1 dw2 dw3 dw4 !dw5 dw6 !dw7 dw8 dw9 dw10 dw11 dw12 dw13 dw14 dw15 !dw16 dw17 dw18 dw19 dw20 dw21 dw22 dw23 dw24 dw25 dw26 dw27 dw28 !dw29 dw30 dw31 !dw32 dw33 dw34 dw35 dw36; ANALYSIS: ESTIMATOR IS WLSMV; MODEL: f1 BY dw1 dw2 dw3 dw4 dw6; f2 BY dw17 dw18 dw19 dw20 dw21 dw22 dw23; f3 BY dw9 dw10 dw11 dw12 dw13 dw14 dw15; f4 BY dw24 dw25 dw26 dw27 dw28 dw30 dw31 dw33 dw34 dw35 dw36; dw4 WITH dw6; dw20 WITH dw21; OUTPUT: sampstat standardized modindices (all) tech4 res; PLOT: TYPE = PLOT3; Thank you. |
|
|
The printed output is not different for Mac and Window. We have a FAQ on this which explains that with multiple factors, IRT essentially uses the Mplus parameterization: IRT parameterization using Mplus thresholds |
|
|
Greetings! We met (many) years ago at the Psychometric Society annual meeting at the University of New Hampshire. According to the "IRT in Mplus" technical notes, it should be possible to convert WLSMV-estimated loadings and thresholds to IRT parameters. I am able to do this for ML estimates, but as an exercise, I would like to be able to do the same or WLSMV estimates. Would you, please, provide or point me to some sample code that I could use as an example? Best regards, Ethan |
|
|
See Section 3 of http://www.statmodel.com/download/MplusIRT.pdf |
|
|
Thanks for your reply. In Equation (24), I recognize phi as the factor variance. But, where in the output would I would find the residual parameters, theta_ik? |
|
|
Under Residual variances. |
|
|
Thank you so much! |
|
Back to top |