In a non-degenerate plasma, why are e-e collision negligible compared to e-ion for thermal conduction?

I'm trying to make some order of magnitude estimates of heat transfer in stars - to better understand 1) why conduction is said to be negligible (for non-degenerate matter) and 2) when convection happens and is dominant.

For thermal conduction, I have the expression that the flux of energy over a surface

$f \approx - \rho \langle v \rangle \cdot \lambda \cdot a k_B \frac{dT}{dx}$

For a mean-free path $\lambda$ and degrees of freedom over two '$a$'. I.e. an expression for the thermal conductivity.

Anyway, everything I read (e.g. See section 'Electron Conductivity'), says that the mean free path should be calculated for electron-ion collisions instead of electron-electron collisions. Can anyone explain this?

-