Sign up, it's free!

How to avoid non-response bias

How to avoid non-response bias

Researchers have at least one common nightmare about sending surveys to their target audience. They all worry that not everyone in the list will fill in and submit their survey. We bet you have also done this, just like us.

This is not only a problem with online surveys. Some people won’t answer surveys that are conducted via phone, on paper, or even done face-to-face.

In many cases, this situation can create misleading and biased conclusions, which in the survey world is called “non-response bias”. Non-response bias occurs in statistical surveys if the answers of respondents differ from the potential answers of those who did not answer.

Let’s say you conduct a survey in a big company and ask questions about employees’ workload. However, managers who have a high workload are simply too busy to answer the survey. Therefore, you will miss their opinions and feedback – which might even be strategically more decisive.

Now you can imagine situations that such surveys can create. With a biased conclusion, you might end up fixing a problem which has never existed, or spend millions on a product which has no real buyer in the marketplace.

So, what’s the solution? How can you control non-responsive bias in your surveys? Here are some tips for you that should help.

1: Choose your sampling wisely

Having poor sampling is the first thing that creates non-response bias. When you target the wrong group, or try to reach more participants than are actually needed, you risk having many participants who don’t respond.

Try to focus on those participants that are potential respondents who will most likely respond to your surveys. This way, you will minimize the quantity of participants and raise the response rate of those whose feedback really matters for your survey.

2: Make a concise and brief survey

Usually, questionnaires with too many questions discourage people from taking time to give feedback. They might be interested in the subject, and even eager to let you know what they think, but nobody really has the time to answer too many questions. Due to the nature of online surfing, people actively avoid spending too much time on web pages.

So, if there are questions that don’t really matter, but you feel you have to ask them as a matter of convention, just skip them. Before writing each question, ask yourself, do I really need to know this and that… E.g. Do I really need to know their gender? Their city? If not, just skip it.

Tip:
And remember, some data gets collected automatically when you use SurveyLegend. For example, you always automatically get each participant’s geo-location (country, city) in your exported data; the time they started or finished your survey, and so forth… Therefore, if you think that data is enough for you, then make the work easier for your survey participants too!

3: Design gorgeous looking surveys

There is almost no need to make the assertion that a dull appearance automatically results in less participation. Surveys DO NOT have to look boring, just because they are surveys. Who came up with this idea?! Well, whatever… when you make surveys with SurveyLegend, they are already pre-designed to look awesome. So, don’t worry ;).

Note:
Just remember, we are all human and appreciate beauty. Perhaps you have a super serious corporate image and intend to project it to everyone all the time. But you may also risk getting non-response bias with that look!

4: Mind the mobile phones

Today, most surveys are taken on mobile devices (phones and tablets). Imagine what happens when you email your survey, or share it on Facebook. Nowadays, people check their emails and social media accounts mostly on their smartphones. If they want to participate in your survey, they’ll do it right away, but if the survey does not look good on the small screen of their phones, they simply leave. That’s it! And you’ll miss your chance to get their feedback.

Note:
Therefore, you should choose a survey tool that offers surveys that are compatible with mobile devices. But don’t worry, because we’ve got your back here as well! Surveys made with SurveyLegend are responsive and automatically adjust to any screen size.

5: Give your participants some motivation

This is a nice way of asking people to fill in your surveys, and it is getting very popular amongst survey creators. However, some survey experts have challenged this technique.

The incentive may be a money offer, a discount code on online purchases, or exclusive access to your premium services.

However, make sure that what you offer does not effect the type of people who respond. You should avoid attracting only a cross-section of your targeted sample to answer your survey. You should know your target group and provide them with incentives that are interesting for all of them.

Tip:
A good idea is to place the incentive (such as a discount code) at the Thank You Page of your online survey. So, only those who finish and “submit” the survey will be rewarded. Alternatively, you can ask for an email or postal address and send it out afterwards.

Also, to inform your respondents about these incentives right in the beginning and attract their attention, you can add a Welcome Page to the survey, and inform the respondents about what they will get as your sign of gratitude for taking their time.

6: Analyze the participants

Focus on the personal details (such as age, location, education, etc.), and compare those who have responded versus those who skipped your survey. Expert analysis of such data might reflect why some people do not participate in your survey. This can be considered in the final analysis of the collected data to limit non-response bias.

7: Analyze participants who have responded last

Studies suggest that those respondents who answer latest within the survey period, often reflect the views of those who did not respond.

8: Resurvey the Non-respondents

A follow-up survey or reminder usually evokes non-respondents to take action and provide you with their views.

We hope this article can help you deal with non-response bias. Now go ahead and make an online survey with SurveyLegend for free, and compare the new response rate with other online survey tools. Then, let us know what you think!

 

Good luck 😉

Kiarokh Moattar

Co-Founder, UX/UI Designer at SurveyLegend
I eventually grew up after painting on many walls, getting too many scars, watching loads of animated movies, taking care of lots of injured animals, and inventing crazy strategies to bypass the "dictatorship" of the adults, and got a B.A. in Psychology. Shortly after, I grew up a bit more and got two M.A. degrees in art & design.

Today, after growing up slightly more, I realize that I've been working with many companies and brilliant people, inventing new tech solutions, designing & coding cool stuff, making cute illustrations; while still truly enjoying, loving, and adoring the mother nature and all amazing cuddly creatures out there.

Leave a Reply