Abstract
Sturrock's theory of stochastic acceleration and Dupree's theory of strong plasma turbulence are here derived in a very simple way, showing that trapping cannot correctly be taken into account by this kind of theory. A comparison of Sturrock and Dupree's diffusion coefficients D shows that they differ significantly only in the domain where these theories fail, and where the highest values of D are reached. The need for a theory taking trapped orbits into account appears clearly. Such an attempt is the object of a sequel.