The following study reports the results of a Randomised Item Response Technique (RIRT), which is an indirect questioning technique that combines Randomise Response with Item Response Theory. This study provides evidence the following: 1) Both population and group estimates of the prevalence of corrupt behaviour that citizens encounter when they interact with their local government; 2) Validates the results of the RIRT against the conventional direct questioning for the same set of sensitive questions; 3) Finally, it reports individual-level estimates of the latent trait of corrupt behaviour and compare them against individual-level estimates obtained from a direct questioning technique.