It doesn’t matter whether it’s empirical research for your dissertation or A/B testing for your product - creating a survey from scratch can be difficult and time-consuming. There are a lot of open questions: Which format should I choose? Which software do I use? How do I find participants? Do I need to pay them? And most importantly: Where do I even start?

This raises two of the most important aspects to consider right at the beginning of the process to create a survey:

WHAT do I want to find out from WHO?

a.   The big “WHAT” - a.k.a - What is your research question?

This may seem like an obvious point. Let’s say you are a marketing student investigating the role of Instagram in the fashion industry. You quickly draft a couple of easy questions about the use of social media and hope for the best.

It’s not that easy, though. Most thesis and therefore survey topics are quite generic, and you won’t be able to cover all their aspects in one survey. You, therefore, need to specify and narrow your topic along with your research question as early as possible. Sticking to the previous example: you might opt for a descriptive research question, such as ‘What percentage of your target audience uses Instagram to inspire their clothing purchase decisions?’. Alternatively, you might want to use a comparative question: ‘What’s the difference in effectiveness between traditional advertising methods compared to Instagram ads for clothing brands?’. Or, to end with, you might want to use a causal research question, which draws a connection between feeling fashionable and being active on Instagram.

Tip: Ask your survey supervisor whether your research question is relevant, sufficiently specific, and properly derived. Think about the exact data you would require to receive to answer the question and how to derive that information through a survey.

b.    The ‘WHO’ –  About which target group do you want to know more about?

This aspect might seem more straightforward to you. After all, you already know which question you are asking. But properly defining your respondents’ demographic profile is just as important for a robust research question. After all, answers across age, gender, and occupation can vary greatly. Looking at the Instagram example, outcomes and insights most likely depend substantially on the age group. Assuming that elderly people do not use social media to purchase clothes, your target group might need to focus on teenagers and young adults, who tend to use Instagram daily.

Tip: Make sure that your desired target group can be reached with the time and financial resources you have available!

Other practical considerations include the actual design of your questionnaire, the phrasing of the questions, and how to find survey respondents. All of these questions are answered by the linked blog posts - check them out now!

Click for the next part of the guide --> How do I create a high-quality and intuitive survey?