ಚುನಾವಣೋತ್ತರ ಸಮೀಕ್ಷೆಗಳ (Exit Poll) ಸಂಖ್ಯೆ ಎಷ್ಟು ಸರಿ? ಎಷ್ಟು ತಪ್ಪು? | How scientific are Exit Polls?


As soon as voting ended, Exit Poll numbers were released These numbers are being projected as if they are the final result. Since Sunday evening, Exit Polls have taken over all discussions, online and offline Every channel is showing a different outcome Predictions say Modi’s NDA government will come back to power, with seats ranging from as high as 354 to as low as 277. Trinamool Congress’ Mamata Banerjee DMKs Stalin Shashi Tharoor of the Congress Party and other Opposition leaders have naturally rejected the Exit Poll predictions But how have Exit Polls fared in the past? For example, in 2004, nearly all exit polls predicted that the Vajpayee-led NDA government would come back to power with as many as 280 seats. But in reality, NDA did not even cross 200 and Congress and its allies formed the UPA government Ten years later, in 2014, exit polls suggested that NDA would just cross the halfway mark But BJP alone got 282 seats and NDA 336 Predicting elections is a very complex statistical challenge and an expensive one Some major exit poll agencies in India are CVoter IPSOS Nielsen CSDS-Lokniti and Jan Ki Baat Each with its own method of arriving at the final number Problems in the survey begin with the very first step identifying which voters to interview Many exit polls have been criticised for not having sufficient sample size or for interviewing only upper-class, urban voters Equations of caste, class, religion, and the number of political parties and alliances make our elections very difficult to predict With interviews being crucial to the final outcome of Exit Polls An agency should be able to get truthful voters, and a sample size which is representative of the electorate to predict the vote share of a party with a degree of accuracy But projecting seat share based on vote share is where exit polls get even more complicated The elections of 2004 best highlight this problem. In 2004, compared to the 1999 elections, BJP’s vote share reduced by 1.6 % and the Congress vote share reduced by 1.8% But when translated to seats, BJP lost 44 seats and Congress gained 31 In an interview to Caravan Magazine in 2013, Yashwant Deshmukh, founder of CVoter said that exit polls can only predict vote share but people want to know the number of seats Exit poll results become even more questionable because most agencies do not publish their questionnaires, sample size, and methodology in the public domain. If exit polls are so unreliable, are they meant only for entertainment and TRPs? Yogendra Yadav, psephologist turned politician often criticises the media for reducing exit poll data to just the number of seats. He says that data from these surveys can be used by journalists to get a deeper understanding of why people vote in a certain way After knowing this, do you still trust exit polls? Turn off your TV and wait for the result on May 23.

About the author

Leave a Reply

Your email address will not be published. Required fields are marked *