I am trying to better understand and explain maximum likelihood estimation. To explain the intuition of many ML aspects I find it easiest to explain them graphically, like for example the ML-based tests LR/Score/Wald:
My question: Is there a similar sketch to be drawn for the most common/simple (OPG, Hessian, Sandwich) estimators for standard errors/confidence bands?
Idea: Intuitively I expect the uncertainty of my point estimate to be captured by the degree of convexity around the max. This fits with the definitions of some of the estimators. But I can’t come up with a good way of drawing this