Do online polls provide the same answers as telephone surveys?

There is no doubt that surveys can affect the result of a local or national election. It influences us, in different ways, according to our personality and beliefs. It may be the factor that will make us vote, or stay at home, if we feel our vote will count. But now that surveys are moving online, does it make a difference? It would seem so, according to two American universities, which have just switched from phone surveys to online polling.

Costs matter

There are a few reasons why online surveys are getting the upper-hand on traditional ones. The first one is the price to pay. It has become out-of-range for most organisations, due to high labour cost. The results also differ greatly. Although there have been very inaccurate ones in the past, it would seem that people feel more at ease, to speak their mind, online than to someone on the phone. If your association, company or group is looking to do a survey, and you would like to change to an online version, navigate to this website.

Two examples of major universities moving to online surveys

This year, two major players in the world of surveys, in America, have decided to make the move from phone surveys to online. Pew is particularly interesting as its study is international, and it is only its American portion that made the change. The Pew Research Center is handling a Global Attitude Survey. It is the largest study on global issues, on a multinational scale.

The second one, is the University of Michigan. They are known for their study on Income Dynamics. It is currently the longitudinal study that has been going on for the longest time, anywhere on earth. It is therefore interesting to have a closer look at what has changed for them, to understand better the differences between online and phone surveys.

A difference ranging between 0 to 18 percentage points.

It became quite clear that there is a large difference between answering a survey online and one on the phone. Normally, a 4% variation is allowed in a survey, as a margin of error. Therefore, to reach 18% indicates that the technique changes the results entirely. For example, when people were asked to rate their quality of life at home, people responded higher on the phone than online. When it came to political figures, their “very unfavourable” option was often the most chosen, and when they were asked questions about discriminations, people online identified gays, lesbians, blacks and Hispanics, more easily, as groups facing a high-level of it.