Creating good questions

Use Questions, Avoid Statements

At Brilliant, we use questions instead of statements. For example:

The reason we use questions is that statements implicitly assume something should be a certain way. By using questions, we avoid placing a value judgment on a particular behavior. Additionally, a statement can be easier to agree with than to disagree with, which may influence the results.

One Aspect at a Time

To avoid confusion among respondents, it’s important to ask about only one aspect at a time. For example:

 

When multiple aspects are included in the same question, such as "listening and giving feedback," it becomes difficult for the respondent to take a clear stance. One area may be perceived positively, while the other negatively. In this case, it's better to split the question or choose the most important area for your needs.

Consistent Language and Terminology

A survey often contains multiple questions, and maintaining consistency helps respondents feel familiar and reduces the risk of misinterpretation. For example, consistently using either "company" or "organisation" throughout the survey creates a cohesive experience. Similarly, choosing "workgroup" instead of "team" ensures clarity. Use terms and expressions that align with your identity and language.

Avoid Similar Questions

Questions that are too similar can confuse respondents. If they receive a new question that closely resembles a previous one, they may feel like they have already answered it. This can be frustrating and, in the worst case, lead to respondents dropping out of the survey.

Avoid Negatively Worded Questions

Negatively phrased questions are harder to understand and may confuse respondents, leading to misinterpretations. For example:

Short Questions Are Easier to Understand and Answer Quickly

Long sentences are often harder to grasp than shorter ones. The recommendation is to keep question texts short. Additionally, placing the most important concept at the beginning of the question often improves clarity.

Unambiguous questions are interpreted consistently

It’s important that all respondents interpret the questions the same way. Use simple language and avoid complex terms that could be understood differently.

Questions must match the response scale

At Brilliant, we use different response scales for various purposes, but our standard and most commonly used scale is a 1–5 rating scale. For example:

1: No, needs improvement
5: Yes, definitely

Ensure that the response options logically match the questions.

Negative Responses on the Left, Positive on the Right

Brilliant’s surveys use a color-coded response scale where positive responses (5) are displayed in green on the right, while negative responses (1) are in red on the left. This influences how questions should be phrased. For example:

Here, answering "Yes" means such jokes occur, but since "Yes" is a positive response, the result would be shown in green. Instead, it's better to phrase the question like this:

Now, the positive response is placed on the right, ensuring that a "Yes" result is positive and displayed in green, while the negative response is on the left and shown in red.

Preview, Test, and Request Feedback Before Use

Get help from your colleagues and test your questions to ensure they work as intended and are interpreted correctly.