This is a question taken from p.95 of Hartley and Hawkes: Rings, modules and Linear Algebra.
Firstly is the question correct or should it read $J \cong K$? This would seem more natural to me.
Assuming the question is correct then one of the directions is trivial so here's my attempt at the other direction.
If $R/J \cong R/K$ then there exists homomorphisms $\psi, \phi$ mapping $R \to N$ with $\ker(\psi) = J$ and $\ker(\phi) = K$ then $\text{im}(\psi) \cong \text{im}(\phi)$. Then I would have to show some how that this implies $J = K$.
I cannot even show that it implies $J \cong K$.
Many thanks
Notice that $(j + J)(r + J) = 0 + J$ for $j \in J$ and that $\phi$ is a module homomorphism.
Since $\phi$ is surjective there exists $r+J$ such that $\phi(r+J) = 1 + K$.
We have $0 + K = \phi(0 + J) = \phi((j+J)(r+J)) = j\cdot\phi(r+J)=j(1+K) = j+K$.
So $j \in K$ and thus $J \subset K$. By symmetry, using $\psi$, we get $J = K$.