ABSTRACT
The choice of the method of collecting data is an integral part of designing any indicator. It should be done in the light of the models and concepts one wishes to use. The constraints and possibilities of the collection process may make one decide to alter the concepts, or these constraints may affect them de facto. Different procedures for collecting data have inherent in them different degrees of reliability, and types of bias. Some produce greater detail or speed in getting results. Cost varies widely. Self reports, where individuals or organizations provide data on themselves according to some established format, provide a less expensive means of getting some of the information otherwise obtainable only through interviews. Observation is often at least one element in any information gathering process, and it undoubtedly should be as it permits the absorbing of different data or data in a different form than the other methods allow.
