Marginal Effects

I know how to run marginal effects on STATA, but was never taught the theory behind them. Working through a textbook that I'm thinking of adopting and found the following, not even sure where to begin interpreting or solving.

Population regression formula: Y = BX + BW + BZ + BXZ + BWZ

X and W are ratio-level; Z is binary

1) Marginal effect of W?
2) Marginal effect of Z?
3) What can we conclude if BXZ were significant?
4) What can we conclude if BWZ is NOT significant?

Any help would be BEYOND appreciated at this point. I even went through and created similar equations in STATA to get effects and try to interpret, but it's to no avail.