What do survey incentives do to your data quality?

© Nik MacMillan via Unsplash.com.

As a researcher, I am always concerned about getting enough responses to make my projects statistically accurate. While a 100% response is not absolutely necessary, 2% is not statistically useful.

Research results impact on decision making for a business and ultimately affect people, so you would obviously want to base these decisions on good quality data and statistically, accurate research results, explains Bischoff.

Biased dataset

I use a reputation measurement tool to measure the perceptions that stakeholders have about an organisation, be it a JSE listed company, major product retailer, association or institution. These perceptions make up a company’s reputation. Give a respondent an incentive and it may influence their perceptions and the feedback they provide, leading to a biased dataset.

Consider some of the factors that contribute to an increased response rate. Think for a moment, why would you answer a survey?

Firstly, people will answer a survey because they think it is important to them. Their input is valuable to the business being researched and will lead to improvement in some way that will also benefit them.

Secondly, people need to have certainty that data will be maintained properly. If someone would like to remain anonymous you need to assure them upfront that every answer that they provide will be confidential, and therefore there is absolutely no risk involved in participating.

When conducting reputation research a common stakeholder group that we usually reach out to our employees. By ensuring them that their answers are confidential, we can encourage them to provide their open and honest feedback about their workplace, giving us good quality data to analyse.

Communication plays an important role

Lastly, answering a survey should not be a complete time burden for a potential respondent. Once again be upfront with the respondent, explain that the survey will only take ten minutes of their time and make sure that they only have to spend ten minutes completing it.

If your survey then takes 20 minutes to complete, you are likely to have lost their interest as well as trust for any future surveys. When designing a survey carefully consider the average time a respondent will need to provide meaningful feedback without losing interest.

If they are going to rush through the survey to get the prize, an incentive is not the way to go to encourage responses.

I always highly recommend clients not to incentivise a survey, the value for the participant lies in the outcomes of a possible change that will benefit them after the survey is complete and recommendations implemented.

When embarking on a research study, communication, therefore, plays an important role. Make sure that you inform your target population about the survey, the purpose behind it, and importantly, how it may lead to improving something and the possible benefit to them. This might just be enough to encourage them to answer the survey.