The credibility of exit polls and opinion polls can be measured in proportion to the closeness of results. Exit polls and opinion polls have not been accurate after predictions that Brexit would lose and Trump would lose. The Prediction of election outcome through exit polls, done before the event, has been a difficult task in developed countries, where citizens have to go through tough security measures to cast their votes.
The example of the Venezuelan recall referendum of 2004: The recall referendum of 2004 was to recall President Hugo Chávez. This case focuses on whether the exit poll that predicted that Hugo Chávez would be recalled had a significant impact in creating a wrong forecast and whether it should have been advised not to release its results.
In recent Bihar elections: The assembly election this year in Bihar states was a big challenge and as usual, the exit poll predictions were on the wrong track. The NDA candidates ended up winning with a huge margin.
US presidential elections 2016: Hillary Clinton was expected to win the US presidential elections in 2016. She led Donald Trump consistently in several polls and most of the bookmakers' odds. Ultimately, Trump won the US presidential elections in 2016.
Principles of statistics ignored: Most pollsters do not follow the statistical principles properly in designing, sampling, and analyzing their data. They do whatever they feel like doing.
Lacks holistic coverage: Exit polls in India are largely carried by private media houses either independently or in the coalition. They do not cover remote corners of the country and cover sensitive booths for their surveys.
No scope for Margin of error: The exit polls and opinion polls conducted for elections are a big deal in the nation. They decide the course of action by the government, opposition party, and public policies to boot. Missing out on some key factors like margin of error and survey takers can doom a poll to failure.
Selection bias: There is a lot of hype around it, people discuss various scenarios and predictions for hours together. But the truth is Exit Poll results are not at all likelier to be accurate than a normal survey. In fact, it is just a survey, with a few more steps to ensure that the sample set maintains the proportion of factors (gender, age, income) before they can be considered representative.
Issues in method of sampling: Almost all the exit pollsters used a method of sampling (sample selection) to identify voters who had come out to vote in the States. They had different ways of doing so. Non responses are not properly tackled and the aspect of ‘random’ sampling is not given adequate importance.
Exit pollsters face a lot of challenges when it comes to collecting data. While gathering data, they come across many problems which result in non-representative data collection. The main challenge faced by them is the inability of the voters to provide accurate information as they can’t make up their minds while casting their votes. There are three major issues faced by the exit pollsters during data collection:
Poor statistical sampling: One of the major challenges faced by exit pollsters is that they use a statistical sampling plan for collecting data from the population.
Cognitive processes: Cognitive processes refer to the mental steps we use to obtain, organize, and interpret information. The cognitive process begins with perception, or getting the information from the outside world (a safe Democratic seat exit pollster) into our minds.
Social-cultural processes: One of the greatest challenges faced by exit pollsters in India is social-cultural processes. Exit polls cannot be done alone because for various reasons, such as respondents are reluctant to speak to strangers or might not be aware of their right to refuse. Thus, a representative team is required to visit voters at home as well as take community interviews along with them.
