What is a Survey?
A survey is a method of research that gathers the opinions, beliefs, and feelings of selected groups of individuals. Methods vary widely, but most of the common survey examples include interviews or directly asking for information through multiple choice questionnaires or polls. Survey research is one form of primary research, which is the first-hand gathering of data from its source.
How Surveys Work
The most important part of a survey is its questions. Questions should be created with the purpose of truly learning individuals' beliefs and opinions. Creators of surveys must be careful to strategically "tease" answers out of individuals, rather than always asking direct questions. This is done in order to get the most honest/accurate answers since it is understood that most individuals are unaware or hesitant of sharing their true opinions/feelings/beliefs.
Distribute survey questions
The most reliable surveys are distributed randomly amongst a sample size of at least 500 individuals. Generally, the bigger the sample size, the more statistically significant the survey is. Some surveys can be distributed completely randomly, while others should be distributed amongst certain populations based on varying demographics or interests.
Measure and evaluate survey results
Data is only as good as its interpretation. After questions have been answered, the statistics should be evaluated for patterns, unique or surprising numbers, or anything surprising or significant. Good data can usually tell a story, but it's up to the statistician/person evaluating the data to find it!
Survey Best Practices
Ensure that every single question has a purpose
A question should only belong in a survey if it is deemed necessary. The goal is to obtain important insights, so each question should have a clear purpose. The best way to do this is to identify what data you need first, then write your questions.
Keep questions short and simple
Survey respondents are easily frightened by long, complicated questions. Make sure your survey questions are easy to understand and follow a logical order. They should also be as concise as possible. Similarly, multiple choice answers should be kept short and clear.
Avoid biased or leading questions
Generally, it is wise to keep questions as unbiased as possible. Things like descriptive words can interject bias into questions, and may be construed as purposely leading the respondent to answer or sway a certain direction (which defeat the purpose of conducting the survey).