If you have a thing with lots of indices, such as a fourth-order sampling probability \(\pi_{ijk\ell}\) (the probability that individuals \(i\), \(j\), \(k\) and \(\ell\) are all sampled), there will likely be scenarios where it has lots and lots of symmetries.
A useful trick is to write a wrapper that checks them:
FourPi<-function(i,j,k,l){
answer <- FourPiInternal(i,j,k,l)
sym <- FourPiInternal(j,i,k,l)
if (abs((answer-sym)/(answer+sym))>EPSILON) stop(paste(i,j,k,l))
answer
}Other useful tricks:
- The score (deriviative of loglikelihood) has mean zero at the true parameters under sampling from the model, even in finite samples
- Quite a few design-based variance estimators are unbiased for the sampling variance even in small samples.
- Many (but not all) variance estimators should be positive semidefinite even in small samples.
- If you have two estimators of the same thing, do a scatterplot of them or of their estimating functions.
More generally, properties of estimating functions are often easier to check in small samples than properties of the estimators. This is especially useful when you have an estimator that takes \(\Omega\left(M^2N^2\right)\) time for large \(N\) and moderate \(M\), so you can’t just scale up and use asymptotics. If the computation time isn’t \(O(N)\) or near offer, tests you can do with small samples are enormously useful