Writing survey questions isn’t difficult – but there is a right way to do it and a wrong way to do it.
In an effort to help you stay on the right side of the tracks, we’re sharing lessons that have been learned the hard way from countless employee surveys with employees from across the engaged spectrum.
Behold… these are Hive’s Survey Question 10 Commandments
1) Thou Shalt Nail the Basics
Like any piece of research, your employee survey should observe good research question practice. Start with the end in mind of what you hope to learn and achieve by conducting your employee survey. Your aim should be to create clear and simple questions that seek to deliver balance rather than bias.
Think about what you as the employer want to know, but also what your employees want the opportunity to talk about – keep it relevant.
Think about consistency and ensure that you make it easy for the participant to answer the questions. I have never really understood why the Best Companies employee survey ricochets between positive and negative questions, requiring the reader to keep re-interpreting the rating scale, with the inevitable confusion about whether questions have been answered correctly. Your aim is to have confidence in the results – not challenges about how people have interpreted the questions.
2) Thou Shalt Know Your Audience
Understand your audience and consider how questions will land. I remember asking a group of Trade employees whether they had a best friend at work – they all laughed (at me) and quickly ticked the ‘neither agree nor disagree’ box! Language is important and using language that people are comfortable and familiar with is a key part of making your survey accessible, inclusive and engaging.
3) Thou Shalt Avoid Loaded Questions
Avoid asking questions which reflect upon the participant’s knowledge or understanding as this will almost always evoke a positive response, regardless of the reality. There are better ways to test understanding than an employee survey, so consider what is important for you to know or understand. For example, consider rephrasing questions like ‘I am clear about the expectations of my role’ to ‘the company supports me to fully understand the expectations of my role’ or ‘my manager supports me to fully understand the expectations of my role’. The latter two questions enable you to both gauge whether someone has clarity of expectation and where you may focus any action that is required.
4) Thou Shalt Not Leave Questions Open to Interpretation
Sense check your survey for questions that people can misinterpret and where necessary include some explanation. This can often show itself when analysing and discussing the results with teams, in the worst case scenario rendering your results unactionable. I have experienced an example of this in a mid-size company where participants applied different interpretations of who they understood to be leaders, managers and line managers. In practical terms this meant that people weren’t bought into the actions that were identified as a result of the survey.
The following year, we included definitions alongside each of the questions. Customise Hive’s questions to fit with the language you use and use Hive’s sub-heading field to elaborate on anything necessary.
5) Thou Shalt Pay Attention to The Detail
Keep jargon to a minimum and use plain English as much as possible. Avoid complexity and technical terms. Where you have any concerns about the levels of literacy of your employees, a test run is a necessity. In circumstances like this, you can always incorporate a briefing session ahead of the survey in order to quickly run through the questions and explain the meaning or focus of each one. If necessary you can also deploy survey ‘buddies’ to support employees to complete questions.
6) Thou Shalt Stay Focussed
Like complexity, make sure each question has a clear focus and doesn’t inadvertently ask the participant two or more things. As an example a question like ‘I have opportunities for training, development and progression’, does not allow someone to have opportunity for one but not the other. Even questions like ‘my ideas are listened to and acted on’ may feel like they’re synonymous with each other, but doesn’t allow for someone to feel their manager listens to their ideas, but never implements any of them.
7) Thou Shalt Avoid Extreme Positivity
Some surveys try to be the change they want to see in the world (or company)! These surveys use language that reflects the organisation they aspire to be, with questions like “What three things most inspire you at work?”
There are two main drawbacks with questions like this. First, it assumes positive experiences are had by all. Second, some people who would answer a less extreme version, may opt out altogether as they may feel it doesn’t accurately portray their employment experience.
Whilst your survey should always reflect your employer brand, it should be more closely aligned with where you are now, than where you aspire to be and the language should be consistent with that. Survey content on its own will not move the engagement dial.
8) Thou Shalt Use Numbers
It is very helpful to be able to quantify responses, to enable you to track trends and movement over time. That’s why Hive uses a 1 to 10 scale. Yes and no answers make it difficult to understand the scale of improvement. I have often been challenged to take out the middle ground in employee surveys whether that be on a 5 point scale or 10 point scale, forcing participants to respond positively or negatively. However, my personal view is the middle ground is a rich seam to mine in terms of insights.
9) Thou Shalt Use Open Ended Questions Sometimes
If you are micro-surveying on a monthly basis or using Hive to run baseline surveys; and want to include open-ended questions to gain further insights, think about the position of them in the survey and therefore, how participants may approach them.
It is often useful to have an open question at the end of each theme or set of questions – this encourages participants to give ideas and suggestions around each theme. Catch all questions right at the end of the survey can be difficult to analyse as they can be somewhat wide-ranging or alternatively not completed as people rush to the finish line.
Most importantly, you should be clear at the outset about how you will analyse and feedback any open or qualitative questions.
10) Thou Shalt Avoid Vagueness
Whilst it is tempting to ask a general question as an overall litmus test, consider what you are really seeking to understand? Asking general questions about culture, engagement and alignment can often rely on all participants having a similar understanding of the concepts or terms. Only use overarching questions where you are happy for people to use their own definitions. For example, it can be useful to have an overall measure of employee engagement regardless of how individuals arrive at their measure.