On the use of disequilibrium models in applied microeconomic research and the value of sample separation information