By: Dianne Meppen, BS
When talking to groups and even in meetings with clients, I often ask the question “If you were contacted for a survey, would you participate?” Generally a few hands go up acknowledging a willingness to take a survey, but – the vast majority of hands in the room stay down. My feelings aside, I am often reminded and kept keenly aware that their response to my question is the norm and not an exception. Simply put, not many people want to take a survey these days.
For many years the responses rates for household surveys has been on the decline. A national study conducted by Pew Research in 2012 estimated the response rate to be approximately 9 percent, down from 27 percent in 2000. What it means is that for every ten contacts, only one produces a completed interview.
Survey research is based on the idea that with scientific sampling methods, data collected from only a part of a population will represent the entire population. The reasons people don’t respond can be many – too busy, do not answer calls, lack of interest, privacy concerns, etc. The question researchers and those using the data face is this: with so many people refusing to participate, does the data still represent the population?
In an effort to counter the low response rates, researchers can use a variety of methods including things like making multiple contact attempts, sending letters in advance of the call, providing incentives to participants, extra interviewer training, callbacks, and using multiple data collection methods. At the same time, all of these methods increase the cost of surveying.
Although low response rates alone don’t always mean survey data is unreliable, it is a trend that is disconcerting. The next time you see survey results pop up in the news, on Facebook, or in a discussion ask: How many people did they sample and what type of people chose to respond? You may find that many surveys that claim to be representative of a population in reality are far from it.
Dianne Meppen is the director of survey research at the Kem C. Gardner Policy Institute.