Runge-Kutta Methods: Modified Euler I
In the Runge Kutta methods, one uses intermediate values to calculate such midpoint derivatives
Key idea is that use an approximation for X(ti+0.5*h) as this is an argument of f which is multiplied by h. Thus error is (at least) one factor of h better than approximation
So if one wishes just to do one factor of h better than Euler, one can use Euler to estimate value of X(ti+0.5*h)