With the political season in full swing, every other day I am reading or hearing about a survey about the candidates or public opinion about issues. In my next few blogs, I’m going to go back to talking about the basics of how to interpret survey results.
On- line surveys have become extremely popular. There are services that have made it easy to do a survey and obtain the results quickly.
If you’ve read my blogs, you’ve learned there are big differences between a statistically-valid random sample survey process, where the data collected represents the views of the total population being sampled (within a certain error range) and a self-select survey, like many that are out there now, where the people who respond make a choice to do so.
Self-select surveys provide good information about peoples’ opinions regarding what’s important and what they appreciate about the product or service a business provides and about some issues. They can also be a quick way of identifying negative concerns, because that’s often what motivates someone to respond to a survey.
So both random sample and self-select surveys are valuable. But, it’s a good idea to understand how a survey has been conducted when interpreting the data collected.
If it’s a self-select survey, data is only accurate for the people who completed and returned a questionnaire. A random sample survey must follow specific procedures, including sample selection and numbers sampled, in order to provide data that is relatively accurate for an entire population.