'Effect on Marginal Predicted Mean' in marginsplot

[Otto Meyer]

After running a multi-level logit regression that includes a continuous-by-continuous interaction, I want to visualize the results. I wonder how the effect can best be visualized and interpreted. Here is some example data:

* Example generated by -dataex-. For more info, type help dataex
clear
input double(inctxff eduyrs unemp_gen_12mo) long gndr double(agea chldhhe green_skills gdp_pc carbonprice_share) long cntry
1 14 67 1 47 2 0 70867.3609970749 44.12539 18
1 19 67 2 56 1 0 70867.3609970749 44.12539 18
0 16 67 2 25 2 0 70867.3609970749 44.12539 18
1 14 67 1 65 2 0 70867.3609970749 44.12539 18
1 15 67 2 59 1 0 70867.3609970749 44.12539 18
end
label values gndr gndr
label def gndr 1 "1", modify
label def gndr 2 "2", modify
label values cntry cntry
label def cntry 18 "NOR", modify

Here is the regression code:

melogit inctxff c.eduyrs##c.unemp_gen_12mo i.gndr c.agea c.chldhhe c.green_skills c.gdp_pc c.carbonprice_share || cntry: eduyrs, or

To improve interpretability of the interaction, I want to estimate and visualize marginal effects. I fix predictor 1 (eduyrs) at its mean value, let values of x2 (unemp_gen_12mo) vary and seek to show how this affects the prediction of x1(eduyrs) on Y(inctxff). Covariates are fixed at their means.

margins, dydx(eduyrs) at(unemp_gen_12mo=(15(15)90)) atmeans
marginsplot

The resulting plot shows the "Effect on Marginal Predicted Mean"? What does this mean? I have attached the resulting plot below.

[Otto Meyer]

It seems the graph hasn't uploaded. Here it is: