Survey methodology

IN the post-poll survey, CSDS investigators went back to the very same locations which were covered and respondents who were approached in the pre-poll survey. The method of selection of respondents and their profile were described as part of the report on the pre-poll survey (Frontline, December 20, 2002). Given the time constraints, the investigators could interview 1,405 respondents. The sample under-represented Muslims and Scheduled Tribes and the raw score would have yielded an under-estimation of the Congress vote. The vote figures reported here have been adjusted in the light of the final results.

The survey was carried out in 76 locations in 26 Assembly constituencies, since election in one of the sampled constituencies was countermanded. Respondents were interviewed to find out whether they had voted or not. The post-poll survey was carried out between the evening of December 12, soon after polling closed, and Deccember 14, the day before counting.

The survey was designed by a team including Professor D.L. Sheth, Professor V.B. Singh, Yogendra Yadav and Sanjay Kumar of the CSDS and Professor Ghanshyam Shah of the Jawaharlal Nehru University in collaboration with the State coordinator, Professor P.M. Patel of M.S. University, Baroda. Professor Patel also supervised the field-work in association with four regional coordinators: Dr. Hemixa Rao, Gaurang Jani, Dr. Gajendra B. Shukla and Dr. Bhanu M. Parmar.

The CSDS would like to thank Achyut Yagnik and Professor Philip Oldenburg for their help with analysis and interpretation. A team of 85 trained investigators drawn from the various universities in Gujarat carried out the survey. The data were processed by Kanchan Malhotra, Himanshu Bhattacharya, H.A.Q.A. Hilal, Ritu Rao, Dhananjay and Bhaskar Jha at the CSDS. The survey was co-sponsored by Frontline and NDTV.