Internal surveys: be right first time, every time
With so many plates spinning at any given time, it’s no surprise that even the most important tasks get pushed down the internal comms practitioner’s to-do list.
Although 95% of internal communications professionals say measurement is important, it’s the activity people spend the least time on.
This means that, when we do find time to commit time to internal research, we need to get it right.
During a recent meeting, a client of mine mentioned how surprised he was by the results of an internal survey he had conducted. A lot of the answers were so unexpected that he was convinced people had misunderstood the questions.
This made me think about how easy it is to get survey questions wrong. Consider how many times a day people ask you ambiguous, confusing or badly structured questions.
For example:
‘Do you do [activity] regularly?’ How often is ‘regularly’?
‘Can I help you with [task]?’ I don’t know. Can you?
‘Do you like [celebrity/TV show]’ How much insight will you really get from a closed question like this?
Of course, in the real world, most of us don’t stop to analyse a question’s grammatical structure before we answer (and those who do are likely to be struck off our Christmas card list).
If someone asks us a closed yes-or-no question to our face, we think nothing of volunteering more information than necessary.
In survey world, this isn’t always possible.
Ask a closed question and you’ll get a closed answer. Ask an ambiguous question and there’ll be no opportunity to correct someone mid-response.
Even considerate IC question masters who thoughtfully include text boxes for people to pour their thoughts into will invariably find that a yes-no question results in a yes-no reply.
These potential polling pitfalls mean we’re at risk of wasting our time and other people’s if we don’t get our survey questions right.
According to Elspeth Bradley, Research Director at Research By Design, internal communication professionals can use the following tips to keep their surveys on track.
Start by thinking about your key objective for the survey. What do you really want to find out?
Keep yes-no questions to a minimum – even if it’s just by including a scale of possible responses, as this will give more useful feedback. Make sure the scales you use are balanced, with a neutral mid-point.
Avoid asking two questions in one. This is confusing for recipients and you might not get back the answer you need.
Make sure questions are not ambiguous. If you’re not totally clear about your objective, you’re more likely to ask questions that are not on target.
When you’ve written your questions, test them on colleagues who aren’t involved with the survey. Ask them if they understood everything, or if there are any grey areas. Act on what they tell you.
Consider using an external research partner who can write questions for you and approach a project with zero bias (there will always be a degree of unconscious bias when data is collected internally).
