Editorial

Litmus test for polls

By Nagaland Post | Publish Date: 5/20/2019 11:48:33 AM IST

 With exit polls projecting a resounding victory for Modi and NDA by netting over 300 seats in the 524-member Lok Sabha, the pollsters will be on the chopping block on May 23 when votes are actually counted. However, even if there is some margin of error, the big picture is that Modi has single-handedly toiled by addressing hundreds of rallies across the country during the campaign period. There are still sceptics who do not accept the accuracy of the exit polls given the fact that the record has not been flattering. This indicates that exit polls fail to predict the final result with reasonable accuracy. The exit polls went haywire in 2004 and 2009 with the exception of 2014 when a country-wide anti-incumbency was felt mainly due to the negative media publicity against the incumbent UPA-II which was spearheaded by the India Against Corruption (IAC) movement led by Anna Hazare and Arvind Kejriwal (who exploited the situation created to form the AAP). The pollsters or agencies involved are largely funded either by big political parties or media houses(most of the latter are also owned by politicians or people close with political parties). There is enough doubt on the techniques and sample sizes of many polling agencies. This arises due to the fact that elections serve as the ultimate means of power and money. Sample sizes are barely even 2% of the electorate. Even if they are 10% of the electorate, they cannot be wholly reliable. Even if the sampling is random and representative of the population, and the sample sizes are adequate, other kinds of errors exist, including ‘non-response biases’. For instance, in some states that went to the polls recently, it was reported that respondents often lied to pollsters, sometimes due to a ‘social desirability bias’. They lied because they wanted to make the ‘politically correct’ choice to the pollster, without admitting their support to some controversial leader or party. Also those interviewed may be afraid of admitting their support due to local political conditions, preferring to come across as riding the bandwagon. The most recent case being the nationwide poll conducted across Australia which had projected a resounding victory for the Labour party. However when votes were counted, it was a total reversal when the ruling Liberal party led by prime minister Scott Morrison turned the tide and won. It may also be recalled that in the US presidential election held in 2016 the Republican candidate Donald Trump wasn’t expected to win as per pollsters. He eventually won and proved all sceptics wrong, including the pollsters. According to the report, errors in predicting Lok Sabha election results are on the rise since 1998-99. Difference between actual election outcomes and the mean prediction is known as error. It was also noted that errors in poll predictions are difficult to explain as most elections in India are conducted without verifiable and authentic methodological protocols. This makes forecast extremely difficult in the context of Indian elections. the higher the number of parties, the more difficult the predictions become. The increase in the number of polling agencies which are not conducting the polls systematically and correctly which increased the chances of overall error. Most polling agencies do not disclose how they arrive at their prediction and their reputations would be tested when votes are counted.

Launched on December 3,1990. Nagaland Post is the first and highest circulated newspaper of Nagaland state. Nagaland Post is also the first newspaper in Nagaland to be published in multi-colour.

Desk:+91-3862-248 489, e-mail: npdesk@gmail.com Fax: +91-3862-248 500
Advt.:+91-3862-2482 267, e-mail:npostadvt@gmail.com

QUICK LINKS

SUNDAY POST

Join us on

© Nagaland Post 2018. All Rights are Reserved
Designed by : 4C Plus