The VRR shows the ability of the validation method to recover q-vectors, and is determined by
$$
VRR =\frac{\sum_{i=1}^{I}I(\mathbf{q}_{i}^{t} = \mathbf{q}_{i}^{s})}{I}
$$
where \(\mathbf{q}_{i}^{t}\) denotes the \(\mathbf{q}\)-vector of item \(i\) in the true Q-matrix (Q.true),
\(\mathbf{q}_{i}^{s}\) denotes the \(\mathbf{q}\)-vector of item \(i\) in the suggested Q-matrix(Q.sug),
and \(I(\cdot)\) is the indicator function.