|
|
 |
| Interpretation of covariate on LGM |
 
|
| Message/Author |
|
|
| Ppudah Ki posted on Friday, March 13, 2020 - 9:15 pm
|
|
|
Hello
I have trouble interpreting covariate (predictor variable) on decreasing LGM trajectory.
My LGM model is 5-time point depression, and the covariate is marital satisfaction (wave 1).
The model fit is good, and the means and variances of intercept and slope are significant. The slope is negative, which means depression has a decreasing trajectory every year. But then when I put marital relationship satisfaction as a covariate, the beta coefficient of marital satisfaction on the slope is positive (significant).
So I would like to get some help interpreting this.
1. Marital satisfaction positively affected the depression slope (which is decreasing). Does this mean that when people have higher marital satisfaction, the decreasing pattern of depression becomes more decelerated (slope flattened)? (by the way, I would expect the opposite based on the literature)
2. Also, can I say that people with lower marital satisfaction at the beginning have decreasing pattern of depression becomes more decelerated (slope flattened)? (which makes sense based on the literature).
I am not sure if I can interpret the results using people with higher vs. lower marital relationship satisfaction.
I would really appreciate how you would interpret my results.
Thank you. |
|
|
|
|
1. Yes. Increasing marital satisfaction produces a higher slope (while possibly still negative), so yes a flattening.
2. Lower marital satisfaction gives a steeper negative slope.
Note also that people with low depression to start with may have little space to lower it. Those with high marital satisfaction may have depression levels at time 1 that are so low that the marital satisfaction can't lower it further. |
|
|
| Back to top |
|
|
