In the world of surveys, even seemingly minor oversights can lead to major distortions in data. Here are five common mistakes to be aware of when designing a survey to ensure accurate findings:
Single select instead of select all that apply – For multiple choice questions, most survey tools default to a single select option. That means the user can only select one answer choice. Sometimes, more than one answer choice applies, but the programmer forgot to change the question type to multi-select. This results in survey results skewed too low since users couldn’t select all the answer choices that apply to them.
Forgetting to include “None of the above” – Similar to Mistake #1, this error also results in misleading data – this time, over-inflated. If the question is mandatory, the user is forced to pick one of the answers to advance to the next question, even though it may not apply to them. For example, a company may ask, “Which of the following ways have you contacted customer service in the past 6 months? Email, chat, phone, etc.” But what if you HAVEN’T contacted customer service and they forgot to include “None of the above?” You’re forced to pick one of the answers, resulting in higher percentages than reality.
Leading Questions - These questions suggest a particular answer, which can bias the respondent's answer. For example, "How much do you love our new product?" assumes that the respondent loves the product. Again, wording like this can skew your data and provide over-inflated results.
Ambiguous Questions – Sometimes questions lack context or enough detail, and the respondent is forced to try to figure out what you’re really asking. Questions that are not clear can be interpreted in multiple ways, resulting in unreliable data. For example, "Do you use our product regularly?" might be interpreted differently by different respondents. A better question is “How often do you use our product?” followed by answer choices such as daily, weekly, monthly, and so on.
Double-barreled questions – These questions ask about two different things at once, making it difficult for respondents to provide a clear answer. For instance, "How satisfied are you with our customer service and pricing?" The respondent may find the customer service to be excellent, but the pricing too high. This question structure is confusing, and the respondent may only answer part of the question.
How can you avoid these common pitfalls? Test, test, and test again before publishing your survey. Think of someone you know well and take the survey from their perspective. This could be a parent, a grandparent, or a busy CEO. If you can get a friend, family member, or co-worker to take the survey, that’s even better. Someone who is not an industry expert is ideal to ensure you’re ready for the general population.
If you’d like more information on how to write an effective survey that gets you the data you need, contact us here https://unmutedci.com/contact-us/.
Comments