An increased focus on data-driven decision-making, and a need to pivot and adapt quickly, means many organisations want to survey staff much more frequently than ever before. However, with too many surveys in a short period of time, survey fatigue can set in, leading staff to become less interested and less likely to complete your surveys. This can reduce data quality and create cynicism from staff about their effectiveness.
Using a sampling approach over multiple surveys is one strategy that can mitigate the impact of survey fatigue. It reduces the number of staff required to complete a survey, saving staff and company time. Sampling can also be a viable option for when you are considering shorter pulse surveys, or when you are interested in running a survey in a relatively short period of time.
In the context of a staff survey, ‘sampling’ simply means surveying some staff as opposed to all staff in an organisation. Some examples of different sampling approaches include:
This is the easiest approach. This involves selecting a random sample of staff from the whole organisation to be surveyed.
This is a slightly more complex approach that involves first identifying the relevant subgroups to be surveyed. These sub-groups will usually be the different departments in your organisation [1]. Once the subgroups have been identified, this process involves collecting a random and proportional sample from each subgroup. Overall, this approach is a great way to ensure that your sample is representative of all departments.
This is like the stratified random approach, this involves first identifying your relevant subgroups. However, rather than collecting a random sample from each subgroup, a random subgroup/s is selected and all staff from that subgroup is surveyed instead. For example, if the ‘HR’ department is randomly selected, all HR staff will be surveyed, instead of surveying a portion of staff from HR, Finance, Operations, etc. This approach can be easier than stratified sampling, as less effort is required to administer and randomly pick out participants to be surveyed. It can also be particularly useful for when an organisation is interested in tracking the progress of a specific department.
These three sampling approaches are particularly useful if you have access to a full staff database, as you can easily draw random participants or actively pick out certain demographics of staff to be included in your sample.
Note, there are also a range of different sampling approaches which do not involve randomisation or drawing from a specific group/demographic. For example, convenience sampling involves asking for volunteers to complete the survey, whereas snowball sampling will ask existing survey participants to refer others to complete the survey. These approaches could be more useful in different contexts, such as running a customer or service quality survey.
Perhaps the biggest downside to sampling is that it provides an opportunity for only some staff to participate in the survey itself. As surveys are a great way to give staff an opportunity to provide feedback and to be heard [2], using a sampling approach can unfortunately cause some staff to feel left out or excluded from the process. In situations where there is low trust from staff, this approach could also cause some staff to doubt the survey process and have suspicions of a deliberate process to include/exclude the opinions of certain staff (therefore raising concerns around the validity of data and leading to lower widespread acceptance of the survey results [3]). Before you consider sampling, check whether staff really are fatigued, or whether they actually view your surveys as a welcome opportunity to provide feedback.
Another disadvantage of using a sampling approach is that it also increases the level of variation in your survey results. Surveying a sample will naturally provide less accurate data than if you were to survey the whole organisation. The smaller the sample, the even greater the uncertainty around the results. As such, you should first consider whether your organisation is large enough to provide a reliable sample.
Overall, using a sampling approach can be useful, especially if your organisation is looking to save time and reduce the impacts of survey fatigue. However, some caution is warranted, as there are also implications arising from small samples and exclusive practices.
Interested in learning more about how to improve your staff surveys? Contact our expert consultants at Xref Engage to discuss the best approach for your organisation.
Uncover data-driven, actionable insights with automated reference, employee engagement and exit surveys.
