Abstract

The study assessed the methodology of the Citizen Satisfaction Index System (CSIS) project as implemented by a local resource institute (LRI). It analyzed the randomness and representativeness in the selection of respondents. The CSIS random protocols—sampling with probability proportional to size, starting-point, random-start, random-number generator, right-shoulder rule, Kish Grid, and randomized control numbers—ensured that the sample came from a cross section of the population and that everyone had an equal chance of being included. The methodology prevented nonresponse by establishing mechanisms for replacement and substitution. However, the design did not include prior information about the total population’s main variables structure to be compared with the sample structure, which was crucial in checking the validity of the sample plan design. Moreover, the methodology systematically excluded small and remote villages that have always been marginalized.