Social media surveys see a rise in popularity

When senior Megan Huber needed student opinions on a new chair and desk being implemented on the university campus for her leadership class, she sent a link to an online survey to all of her Facebook friends.

She said she uses online surveys to get respondents when she needs student opinions for class assignments.

“They’re quick and easy,” Huber said. “[Online] surveys just make the most sense to get quick information, rather than trying to conduct personal interviews.”

Social media is one of the easiest and fastest ways to reach a larger population, she said.

“I send a link to the surveys to all my friends on Facebook through an event invite,” she said. “Usually it’s a group project, so if we all send it to all of our friends, we can usually reach about 1,500 people.”

Sophomore Caitlyn Goodhue stated in an email message that she uses online surveys in her communication research methods class because they can be sent to people living in different areas.

“Online surveys are useful because they allow someone to reach a large number and range of people,” Goodhue said. “For example, if someone wanted to poll friends or people they knew in other parts of the country or abroad.”

Despite their potential to reach a wide range of participants, Paul Brewer, associate director of research at the Center for Political Communication, said distributors of online polls are often unaware that these sampling procedures can lead to skewed results.

“Their samples are not representative. They are not generalizable to the broader population,” Brewer said. “You can’t say whether the results from your survey tell you anything about the people who weren’t in the survey.”

Brewer said reputable polling is conducted through probability sampling, which means every member of the population being researched has a chance to participate in the survey.    
“There are some organizations who try to do representative online surveys,” he said. “But your garden variety online survey doesn’t go through that trouble.”

Brewer said Internet users taking the same online poll tend to have similar interests and generalize the sample population.

“If you post a link to an online survey to a Facebook page, you are going to get people who were looking at that Facebook profile, or people who happen to be looking at that website, which probably aren’t going to be typical of the broader population,” he said.

Freshman Emma Glickstein said she isn’t bothered by online survey requests through Facebook because she hasn’t received many this year. However, she said when she does receive the surveys, she doesn’t put much effort into them.

“Sometimes I barely read the question,” Glickstein said. “Also, I just answer what seems appropriate or what most people would answer.”

Brewer said he while online surveys are very effective in reaching a large number of participants, the biased population sampling cannot be ignored.

Ed Ratledge, director of the Center for Applied Demography and Survey Research said people will continue to use online surveys despite their flaws.

“They’re so easy to do,” Ratledge said. “Our traditional methods of surveying can cause up to 30 to 40 dollars an interview as opposed to zero.”

Huber said when analyzing the results of an online poll, it is important to take the unrepresentative sample into consideration.

“Surveys are relatively effective for school projects,” she said. “Teachers are pretty understanding of the fact that your results may be biased toward college students and their points of view.”

Huber also said she agreed online surveys may not be appropriate in situations outside of the classroom.

“If I were doing a real-world research project, I would have to try to find a different method, or else find a better way to distribute the survey other than Facebook,” she said.

Ratledge said many online surveys do not require respondents to complete the entire survey.

“It has to do with the privacy rights that an individual actually has,” he said. “It is done in order to eliminate uneducated opinions. If you don’t have an opinion, we don’t necessarily want you addressing the issue.”

Although conventional phone and in person surveys are more expensive, Ratledge said he advocates for more traditional methods of research when attempting to draw conclusions or make predictions.

“All you can report from an online survey is, ‘here is what we asked and this is what they said,’” Ratledge said. “Just don’t tell me that this is the way the entire population would respond.”