When you glance at the gas gauge in your car, there is a lot that you are taking for granted. What would be the impact of the gauge over-estimating fuel by 10 liters? What if the gauge actually confused things by measuring multiple fluids, not just gas? If (like me) you often cruise into a fill-up station within your last few liters of fuel, this would create a significant problem.
CEOs often talk about employees being their most important asset. Why then would we take the pulse of how employees are doing with an imprecise in-house employee survey? To save a few thousands dollars HR teams around the world have been creating their own “DIY” surveys, often with unexpected consequences. I understand the need to save money, but by creating the DIY survey we are not respecting the craft of creating valid, reliable instruments.
Here are 4 things to look out for if you are trying to improve your approach to employee surveys.
1. Does your survey reliably measure what you think it does?
The only way to know that your questions are valid is to do research with them. Through research, we are able to prove that data collected with the instrument has:
- validity (do these questions measure what they think they do?) and
- reliability (are the results comparable and repeatable over time?)
2. Are your overall survey scores distorting the truth?
Having an overall survey score helps us communicate results in a single measure across groups. It helps us quickly compare departments or business units to each other – or does it? By comparing the High-Performance Index(™) to several popular employee surveys, we have discovered that overall scores can be very misleading.
For example if a survey has 15 questions about being clear on expectations and only 3 about teamwork, the overall average score is heavily skewed toward clarity. This is an issue with heavily-researched “leading” engagement surveys, and there is a significant chance that your DIY survey from HR will also have flaws, unless you deliberately addressed distortion.
3. What is missing from your survey?
Another red flag in DIY surveys lurks within what hasn’t been asked. Many surveys are created with a specific intent in mind. Sometimes, this creates a bias toward data that validates a specific decision the organization would like to take. Biased survey design, whether innocent or intentional, will cause us to leave out areas of inquiry that could identify areas of concern or change the actions we decide to take.
Many DIY surveys for D&I (diversity & inclusion) underestimate the connection between D&I and achieving goals, focusing instead on elements of the human experience of work. In order to understand the full scope of your D&I concerns, we must also look at performance questions because they are very relevant to creating a better culture – and connect our D&I work directly to results! Our approach first looks at the big picture and then delves deeper into specific areas of further inquiry.
4. What will you do with the results?
Now that we have covered the survey design, let’s think about the results. Somebody needs to collate and interpret the results. This is a problem if you have asked several open text questions which will produce unstructured data. Vague or ambiguous questions make it really difficult to design an action plan. It is impossible to design actions to improve results if you asked ambiguous or open-ended questions like ‘do you feel included’ or ‘are you having a great employee experience’. Open-text data can be daunting to work with, so begin with the end in mind.Think about how you can decipher the results in order to take action.
We have painstakingly designed questions and poured through mountains of data to help give you clear results tied to practical actions you can take. If you don’t have thousands of hours to do the same, we suggest you use a validated instrument instead of going the DIY route. When you place your employee survey in the hands of experts, it shows that you truly respect the importance of precisely measuring the amount of gas that is in the tank.