Question

How does the way a poll is conducted influence the answers people give? Can you provide a few examples so I can understand it better

Answer 1

The way how the question is framed is very important when we conduct a survey. Different types of bias can occur if the questions are not framed correctly. This is called response bias. There are different types of response bias.

1) Leading question: These are questions framed in such a way that it will lead the responder towards giving a particular answer. For example, if you are asked 'Do you support the government's economic policy which is intended to alleviate poverty?', you are more likely to say 'Yes' as answer.

2) Social desirability bias: These are the questions to which the responders are likely to lie because truth can often be considered socially undesirable. One such question is 'Do you always wear seat belt when you drive?'. It's more likely that people will say 'Yes' here even if they actually do not wear seat belt while driving.

3) Question order bias: This is the bias that occur as a result of order in which questions are asked. If someone is first asked about their political inclination and have responded as 'Republican', the questions that are followed will be answered from the Republican perspective. If the political affiliation is asked in the end, the person who actually does not support some policies of Republican government is more likely to say so even if he or she is a Republican.

Another bias associated with this is answer order bias in which people tend to pick the first few options more times than the ones in the end.

4) Limited options: This is the bias that occur because all possible options are not given with the question. When asked 'Do you support the government's foreign policy?', people who do not have an opinion or who are neutral are forced to say either 'Yes' or 'No' if those two are the only options given in the questionnaire.