Hi,
I'm new to ordinal logistic (xtologit) models but I try to interpret the effect of eye fixation time on recall from a video. I have a panel data set since I have different information from different scenes in the video. The model also contains dummies that represents the precens of different individuals and texts in the scenes. To my understanding this gives me the odds-ratios. How do I interpret the coefficients? Do I take the exponential function of the coefficients or is this only true for the dummy variables?
I have tried to read what the sigma2_u means but I'm not sure what to aim at, the model shows me Coef.: 0.5578779 and Robust Std.Err.: 0.1566713.
I have also calculated the rho since the output didn't reveal it (like I have seen in others' outputs). I used the following commands:
local logistic_variance = (c(pi)^2)/3
local rho = _b[sigma2_u:_cons]/(_b[sigma2_u:_cons] + `logistic_variance')
display `"`rho'"'
The results show rho=0.1449882252888197. How do I interpret this? Is it the variance?
Have I missed something important?
Hope that someone can help me!
Best,
Martina
(Don't know if it helps but I copied the output)
. xtoprobit RC EFT MalePA FemalePA MaleOA FemaleOA Text, vce(robust)
Fitting comparison model:
Iteration 0: log likelihood = -270.92752
Iteration 1: log likelihood = -252.07064
Iteration 2: log likelihood = -251.9922
Iteration 3: log likelihood = -251.99218
Refining starting values:
Grid node 0: log likelihood = -238.87537
Fitting full model:
Iteration 0: log pseudolikelihood = -238.87537 (not concave)
Iteration 1: log pseudolikelihood = -237.1539
Iteration 2: log pseudolikelihood = -234.14512
Iteration 3: log pseudolikelihood = -234.11837
Iteration 4: log pseudolikelihood = -234.11834
Iteration 5: log pseudolikelihood = -234.11834
Random-effects ordered probit regression Number of obs = 200
Group variable: Individ Number of groups = 20
Random effects u_i ~ Gaussian Obs per group:
min = 10
avg = 10.0
max = 10
Integration method: mvaghermite Integration pts. = 12
Wald chi2(6) = 37.86
Log pseudolikelihood = -234.11834 Prob > chi2 = 0.0000
(Std. Err. adjusted for 20 clusters in Individ)
------------------------------------------------------------------------------
| Robust
RC | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
EFT | .0002417 .0001221 1.98 0.048 2.33e-06 .000481
MalePA | .3223983 .260019 1.24 0.215 -.1872295 .8320261
FemalePA | .4146489 .245673 1.69 0.091 -.0668614 .8961592
MaleOA | -.4807377 .3294408 -1.46 0.144 -1.12643 .1649545
FemaleOA | .7115445 .2447673 2.91 0.004 .2318094 1.19128
Text | -.7980798 .2842195 -2.81 0.005 -1.35514 -.2410198
-------------+----------------------------------------------------------------
/cut1 | .0086747 .5029984 -.9771841 .9945335
/cut2 | .109286 .512719 -.8956248 1.114197
/cut3 | .4577737 .5205282 -.5624427 1.47799
/cut4 | 1.187028 .550084 .1088829 2.265173
-------------+----------------------------------------------------------------
/sigma2_u | .5578779 .1566713 .3217307 .9673547
I'm new to ordinal logistic (xtologit) models but I try to interpret the effect of eye fixation time on recall from a video. I have a panel data set since I have different information from different scenes in the video. The model also contains dummies that represents the precens of different individuals and texts in the scenes. To my understanding this gives me the odds-ratios. How do I interpret the coefficients? Do I take the exponential function of the coefficients or is this only true for the dummy variables?
I have tried to read what the sigma2_u means but I'm not sure what to aim at, the model shows me Coef.: 0.5578779 and Robust Std.Err.: 0.1566713.
I have also calculated the rho since the output didn't reveal it (like I have seen in others' outputs). I used the following commands:
local logistic_variance = (c(pi)^2)/3
local rho = _b[sigma2_u:_cons]/(_b[sigma2_u:_cons] + `logistic_variance')
display `"`rho'"'
The results show rho=0.1449882252888197. How do I interpret this? Is it the variance?
Have I missed something important?
Hope that someone can help me!
Best,
Martina
(Don't know if it helps but I copied the output)
. xtoprobit RC EFT MalePA FemalePA MaleOA FemaleOA Text, vce(robust)
Fitting comparison model:
Iteration 0: log likelihood = -270.92752
Iteration 1: log likelihood = -252.07064
Iteration 2: log likelihood = -251.9922
Iteration 3: log likelihood = -251.99218
Refining starting values:
Grid node 0: log likelihood = -238.87537
Fitting full model:
Iteration 0: log pseudolikelihood = -238.87537 (not concave)
Iteration 1: log pseudolikelihood = -237.1539
Iteration 2: log pseudolikelihood = -234.14512
Iteration 3: log pseudolikelihood = -234.11837
Iteration 4: log pseudolikelihood = -234.11834
Iteration 5: log pseudolikelihood = -234.11834
Random-effects ordered probit regression Number of obs = 200
Group variable: Individ Number of groups = 20
Random effects u_i ~ Gaussian Obs per group:
min = 10
avg = 10.0
max = 10
Integration method: mvaghermite Integration pts. = 12
Wald chi2(6) = 37.86
Log pseudolikelihood = -234.11834 Prob > chi2 = 0.0000
(Std. Err. adjusted for 20 clusters in Individ)
------------------------------------------------------------------------------
| Robust
RC | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
EFT | .0002417 .0001221 1.98 0.048 2.33e-06 .000481
MalePA | .3223983 .260019 1.24 0.215 -.1872295 .8320261
FemalePA | .4146489 .245673 1.69 0.091 -.0668614 .8961592
MaleOA | -.4807377 .3294408 -1.46 0.144 -1.12643 .1649545
FemaleOA | .7115445 .2447673 2.91 0.004 .2318094 1.19128
Text | -.7980798 .2842195 -2.81 0.005 -1.35514 -.2410198
-------------+----------------------------------------------------------------
/cut1 | .0086747 .5029984 -.9771841 .9945335
/cut2 | .109286 .512719 -.8956248 1.114197
/cut3 | .4577737 .5205282 -.5624427 1.47799
/cut4 | 1.187028 .550084 .1088829 2.265173
-------------+----------------------------------------------------------------
/sigma2_u | .5578779 .1566713 .3217307 .9673547
