by Kaetlyn Phillips, Data Services Librarian
Ever watched an episode of Family Feud? The show’s premise relies on contestants guessing answers to survey style questions. Surveys are an incredible way to quickly gather opinion data. As a result, we probably encounter surveys on a routine basis. For example, just this week I encountered the following:
a) While shopping three different stores prompted me to complete a user experience survey by using a website and code on the receipt. If I participated I would be entered in a draw for a gift card.
b) While watching YouTube I was asked to complete a short survey on ads I may have recently viewed.
c) While traveling through an airport there were numerous stations asking about the airport experience. This included satisfaction with going through security, cleanliness of the washrooms, and satisfaction with food services.
User experience surveys are only one type of survey you may experience. Other surveys include text based polls through news agencies, internet or phone based public opinion surveys, and Statistic Canada surveys. User experience polls or daily question polls are “quick polls” that measure the opinion of the participant. User experience data gathers your opinion for a specific transaction or single experience. New media quick polls often relate to a story from the day. Quick polls are great for gathering opinions in the moment as they usually take less than a minute to complete and often only feature a single simple question, but are not designed complexly enough to measure long term opinions on the subject. For example, you could rate a shopping experience as positive on Tuesday and negative on Wednesday due to multiple factors. Research surveys, public opinions surveys, and government surveys are much more structured than a quick poll, so we’ll tackle them in more detail next month!
Not everyone will participate in a survey, so sample sizes need to be considered. The sample is the number of participants needed to represent the survey’s targeted population. Quick polls often use volunteer sampling meaning participants have to choose to complete the survey by going to a website, scanning a QR code, or texting a response. These methods are anonymous and non-confrontational which reduces courtesy bias, when participants change their responses in order to be polite.
Volunteer sampling is efficient, but results must always be critically examined. For instance, volunteer sampling is prone to selection bias, in particular voluntary response bias, because the “silent majority” do not respond. People who have neutral experiences and opinions are just less likely to remember the encounter or are not motivated enough to complete the survey. People who have strong opinions are typically highly motivated to complete surveys as a way to express their feelings.
To increase motivation in neutral participants, quick polls may offer small incentives such as prizes or monetary compensation. However, the incentive has to be carefully chosen to increase motivation, but not encourage false or incomplete responses from participants just to get the incentive. However, as participants we should consider and think critically about what personal data we are giving up in order to get a quick poll incentive. In some cases demographic and contact data must be provided to get access to the survey, which could result in spam material. In other instances, completing the surveys may require making a store account or downloading an app which will track our data. Ultimately the choice is ours!