The Quality of Observational Data

I am deeply sensible of the honour that the members of the Statistics Section of the Indian Science Congress Association have done me by asking me to preside over the Section this year. It is customary on such occasions to review one of the fields in which one is working or to dwell on problems of general interest or to express oneself on the related philosophical questions. I propose to try a combination of these several alternatives and deal with a question which is the very basis of all our statistical activities and which has concerned me and also vexed me over the last many years. I wish to speak on the quality of observational data in statistics.