# Difference between quasi-likelihood estimating equations, IEE and GEE?

1. What is the difference between quasi-likelihood estimating equations, and GEE?

From a note

Quasi-score function is for independent, over-dispersed data (Poisson or binomial), while GEE1 is for correlated data.

But it seems to me that in the note, their equations both don't take into account the correlation between different responses. So how shall I understand them correctly?

2. How is independence estimating equations (IEE) different from the above two?

Thanks!

1. Two lines above your quotation: "Note that the dimension of $g_i$ varies from stratum to stratum and when $n_i=1$ for all $i$, $g$ reduces to the quasi-score function." The observations within a stratum is generally correlated, like in longitudinal data.