One of the debates in survey research is around scales – “How much do you agree…”, “How satisfied are you…”, “How likely are you to…” – specifically, should the scale have a “neutral” point or not?
Scales should have a neutral point. Why?
- Surveys efficiently collect and organize people’s opinions. A good survey should “disappear into the background” – the respondent should be able to focus on content and their opinions, without having to worry about or figure out how to answer the questions they’re being asked. The more effort spent understanding the question, the less energy available to communicate opinions.
- Most people don’t have strong opinions about everything. There are many times when something just isn’t important to a respondent. Since a survey’s goal is to reflect their opinions, it should reflect when they don’t have a strong reaction to a statement.
- “Forcing” a choice of only positive or negative creates fake variation – biasing your data. Since some people don’t have an opinion either way on certain topics, creating a question where they are forced to choose positively or negatively creates random variation in the data – some respondents will choose slightly negative and some will choose slightly positive even though they don’t have an opinion at all. This random variation not only biases the data, it also makes your analysis inaccurate, creating false conclusions.
What do you think, should surveys have neutral points? Why or why not?