Hello, everyone. I am currently working with cross-sectional survey data from seven countries, comprising 35,893 observations. My dependent variable is "satisfaction with democracy" (categorized into four categories), and my independent variable is "populism attitudes." Additionally, I am considering the moderating variable, "corruption levels."
The model includes an interaction between an individual-level continuous predictor.
I would like to calculate the marginal effects and predicted probabilities of populism attitudes and satisfaction with democracy based on different levels of corruption.
However, I am facing challenges in finding the appropriate information on how to calculate a continuous-by-continuous interaction in a multilevel model.
Here is my code.
I hope these revisions help improve the clarity and readability of your text.
I'm not sure "margins , dydx(Populism) at(corruption=(0.08(0.05)0.79)) expression(predict (outcome(4) mu fixed) + predict(outcome(3) mu fixed)) vsquish atmean" is predicted probabilities or marginal effects.
I really tried to do this correctly, however I apologize if this is too wordy, too vague, or difficult to read. Thank you in advance.
The model includes an interaction between an individual-level continuous predictor.
I would like to calculate the marginal effects and predicted probabilities of populism attitudes and satisfaction with democracy based on different levels of corruption.
However, I am facing challenges in finding the appropriate information on how to calculate a continuous-by-continuous interaction in a multilevel model.
Here is my code.
I hope these revisions help improve the clarity and readability of your text.
Code:
meologit demo_satisfy c.Populism c.corruption///
c.Populism#c.corruption///
$control ///
|| n_country: , diff
margins , dydx(Populism) at(corruption=(0.08(0.05)0.79)) expression(predict (outcome(4) mu fixed) + predict(outcome(3) mu fixed)) vsquish atmean
# satisfaction with democracy is predicted by combining 3 (satisfied) and 4 (very satisfied)
------------------------------------------------------------------------------
| Delta-method
| dy/dx std. err. z P>|z| [95% conf. interval]
-------------+----------------------------------------------------------------
Populism |
_at |
1 | .2285128 .0448994 5.09 0.000 .1405116 .316514
2 | .226162 .0380684 5.94 0.000 .1515493 .3007747
3 | .2232461 .0318626 7.01 0.000 .1607965 .2856957
4 | .2197591 .0263693 8.33 0.000 .1680763 .2714419
5 | .2156988 .0216909 9.94 0.000 .1731855 .2582121
6 | .2110671 .0179506 11.76 0.000 .1758845 .2462496
7 | .2058698 .0152808 13.47 0.000 .17592 .2358197
8 | .2001175 .0137635 14.54 0.000 .1731415 .2270934
9 | .1938246 .0133329 14.54 0.000 .1676925 .2199567
10 | .1870103 .0137537 13.60 0.000 .1600535 .213967
11 | .1796978 .0147392 12.19 0.000 .1508096 .208586
12 | .1719148 .0160741 10.70 0.000 .1404101 .2034195
13 | .1636927 .0176435 9.28 0.000 .1291121 .1982734
14 | .1550669 .0194036 7.99 0.000 .1170365 .1930973
15 | .1460762 .0213452 6.84 0.000 .1042402 .1879121
------------------------------------------------------------------------------
I really tried to do this correctly, however I apologize if this is too wordy, too vague, or difficult to read. Thank you in advance.
