Delhi Election 2013: Three Surveys, Three Different Results

By on September 28, 2013


The Delhi Election Survey is carried out in all Delhi constituencies at the time of each Delhi Assembly Election. The data collection and the sampling procedure raise doubts about the data  quality used for the survey.

Three Surveys, all taken on the same days, came to different conclusions. The Cicero Associates opinion poll conducted found vote share for Aam Aadmi Party 32 percent, Congress 24 percent and BJP 23 percent. The C Fore found vote share for Aam Aadmi Party 21 percent, Congress 34 percent and BJP 32 percent. The C-Voter poll gave Aam Aadmi Party 16 percent, Congress 34 percent and BJP 38 percent.

delhi election survey

The differences in the surveys can be accounted for several ways. The sampling error is looking the best possibility for different results. The sample size of Cicero was 20,000 candidates, C-Fore was 15,000 candidates and C-voter was 39,000 candidates in four states. The sample sizes was not large enough to overcome appropriate skepticism about their conclusions. The surveys may not be polling those who actually vote, or people are no longer responding to pre-polls truthfully. The surveys conducted suggest that something systemically wrong was taking place in their sampling data.

In the Cicero survey, 41 percent of voters chose Arvind Kejriwal as their CM candidate while 20 percent opted for Congress leader Sheila Dikshit, and only 14 percent in favour of BJP leader Vijay Goel. In the C Fore survey that covered 14,689 voters, 22% chose Kejriwal as their CM candidate while 29% opted for Diskhit and BJP’s Vijay Goel won the approval of 18%. In the C Voter survey, the preferred Chief Ministerial candidate projects a vote share of 17  percent for Arvind Kejriwal(Aam Aadmi Party), 38 percent for Sheila Dikshit(Congress), and 18 percent for Vijay Goel(BJP).

All the three surveys only agree that Vijay Goel is the least preferred choice of Delhites.

delhi 2013 survey

Opinion Polls done months before an election don’t turn out to have been very predictive of real outcomes. Let us have a look at the C-Voter survey 2008.  The survey said opposition BJP is expected to bag between 36 and 42 seats for the Delhi Assembly in a House of 70. The ruling Congress may get 28 to 34 seats, but actual result doesn’t match the projection.  This proves that the public is not actually swayed by the results of opinion polls.  In general, the public appeared to vote for its favourite candidates even if the odds were against them. The people did not seem to be swung in significant numbers one way or another by opinion survey data.

Delhi election

Opinion Polls can be manipulated to give a false picture of public opinion.The errors in opinion polls can be produced, intentionally or by accident. The error could be the actual wording of the questions in the interview, or the context in which the question has been asked.

Since polls will continue to be used as indicators of public opinion, the surveys should be doing more sampling to their reporting of the results.Before analyzing polls, check out whether the pre- polls conducted are in the welfare of the general public, or the media gather data furthering the interests of some particular parties.

Leave a Reply

Your email address will not be published.

Time limit is exhausted. Please reload the CAPTCHA.