Lokniti programme of the Centre for the Study of Developing Societies (CSDS), Delhi conducted post-poll surveys in Assam, Kerala, Tamil Nadu and West Bengal. The surveys were conducted March 28 through April 21 in Assam, April 7 through April 13 in Kerala, April 7 to April 20 in Tamil Nadu and March 28 through May 1 in West Bengal. In Assam, the survey was conducted among 3490 voters at 140 polling stations spread across 35 assembly constituencies, in Kerala among 3424 voters at 140 polling stations of 35 assembly constituencies, in Tamil Nadu among 4354 voters at 160 polling stations in 40 assembly constituencies and in West Bengal among 4223 voters at 200 polling stations spread across 50 constituencies. The sampling design adopted was multi-stage systematic random sampling. The Assembly Constituencies were randomly selected using the probability proportional to size method. Thereafter, four polling stations within each of the sampled constituencies were selected using the systematic random sampling method. Within each polling station, 40 voters were randomly sampled from the electoral roll using the systematic random sampling method. Of these 40, 25 interviews were targeted. The interviews of electors were conducted face-to-face at their homes after voting had taken place in their area. The questionnaire designed for conducting the interview was a standardized semi-structured one and was translated to the local language/s. In Assam, the interviews were conducted in Assamese and Bengali, in Kerala in Malayalam, in Tamil Nadu in Tamil and in West Bengal in Bengali. The interview duration across all States was about 35 minutes on average. While the interview targets were more or less achieved in Assam, Kerala and Tamil Nadu, in West Bengal the total number of interviews fell short of the target due to challenges/slowdown in fieldwork on account of a surge in Covid-19 cases during the latter phases.
In order to ensure representativeness and correct for underrepresentation of key demographics, the achieved raw sample has been weighted by gender, religion, locality, and caste group based on Census 2011 data. The final data sets have also been weighted by the actual vote shares secured by the major parties and fronts that contested the elections in each State. All analysis here has been presented on the weighted data sets.
In order to ensure the safety of the field investigators amidst the Covid-19 pandemic and also of those around them, field investigators were provided with masks and sanitizers for the fieldwork and it was ensured that the interviews were conducted following proper physical distancing and mask rules. Before going to the field FIs were made to sign an undertaking that they are not showing any of the Covid-19 symptoms and that they would take all the necessary precautions in the field.
THE TEAM THAT CONDUCTED THE SURVEY
In Assam, the survey was coordinated and supervised by Dr Dhruba Pratim Sharma (Gauhati University) and Dr Nurul Hassan (Kampur College, Nagaon), in Kerala by Dr Sajad Ibrahim, Mr. Kiran Raj and Mr. Abhishek P.S. (University of Kerala), in Tamil Nadu by Dr P. Ramajayam, D. Kirubanithi (Bharathidasan University, Tiruchirapalli), Dr Gladston Xavier and Paul K. Nathan (Loyola College, Chennai), Benisha Catherin (Program manager, Glo Foundation) and Mercy Pious (Independent development consultant), and in West Bengal by Dr Suprio Basu (University of Kalyani) and Dr Jyotiprasad Chatterjee (Barrackpore Rastraguru Surendranath College). Prof. Sanjay Kumar of Lokniti-CSDS, Delhi directed the survey.
The articles published in the Hindu based on these surveys:-
Local Factors Determine Electoral Outcomes in States, 4 May 2021, https://www.thehindu.com/opinion/op-ed/local-factors-determine-electoral-outcomes-in-states/article34475075.ece
What was the Methodology Used, 4 May 2021, https://www.thehindu.com/opinion/op-ed/what-was-the-methodology-used/article34475146.ece