We introduce a framework for parameter estimation of microkinetic models via injecting data, collected in optimal regions of the "entire" experimental operating space, in models. We demonstrate this framework by combining differential conversion experimental data without transport limitations and models for ammonia decomposition on Ru/gamma-Al2O3 for hydrogen production. Experiments indicate that there is significant H-2 inhibition even at low ammonia conversions. Statistical model discrimination techniques indicate that multiple microkinetic parameter sets are able to describe quantitatively the experimental data, and some global rate expressions are also adequate. Among the best microkinetic models, nitrogen adsorption/desorption and the NHx dehydrogenation reactions are the kinetically significant reactions. It is found that macroscopic data (conversion) are insufficient for complete model discrimination; microscopic scale data are proposed for further model discrimination.