Multi-wave surveys allow you to analyse how attitudes, behaviours, or awareness change over time. Each time a survey is re-run, responses are grouped into a new wave, making it possible to compare results across different points in time.
This article focuses on analysing multi-wave results. For information on setting up or sending surveys across multiple waves, see Creating multi-wave surveys.
What is a wave?
A wave represents one instance of a survey being run.
For example, if you run the same survey in Q1, Q2, and Q3, each of those runs is treated as a separate wave. Waves allow you to:
analyse results from a specific point in time
compare how results change between runs
aggregate data across multiple runs for a more robust sample
Viewing results by wave
In the results dashboard, you can choose to:
view results for a single wave, or
aggregate results across multiple waves
Aggregating results combines responses from selected waves into one dataset. This increases sample size and can be useful when you want a more stable overall view rather than point-in-time comparisons.
Comparing waves
To compare waves directly, add Wave as a split and view this charts or crosstabs.
This allows you to:
view waves side by side
track trends over time
spot changes in responses between survey runs
When comparing waves in charts, you can further customise these visualisations. From Edit chart, you can:
change the chart type, for example switching to a line chart to better show trends over time
switch axes to adjust how waves and answer options are displayed
remove or hide specific answers
adjust visual settings to improve clarity
Customising charts can make changes between waves easier to spot, especially when presenting results to stakeholders. For more detail, see Editing charts on the results dashboard.
If you are using Crosstabs, you are also able to add waves as a stacked variable to dive even deeper into your results and compare multiple variables in once go.
Interpreting changes over time
When comparing waves, it’s important to interpret differences carefully. Small fluctuations, such as changes of 1–2 percentage points, are often within the margin of error and do not necessarily indicate a meaningful change. Larger or consistent shifts across multiple waves are more likely to reflect real movement.
Statistical significance can help highlight changes that are unlikely to be due to chance. By default, significance is shown relative to the total at a 95 percent confidence level, but this can be adjusted in chart options.
Creating boards for multi-wave reporting
Boards can be used to create clear, shareable views of multi-wave results.
For example, you might create:
a board for each wave (Q1, Q2, Q3)
a single board showing trends across waves
boards tailored for different stakeholders
Boards make it easier to present changes over time without rebuilding charts manually.
Using Compass for multi-wave analysis
Compass can support multi-wave analysis by helping you explore, organise, and summarise changes over time.
You can use Compass to:
create charts split by wave on boards
edit and refine charts as you explore trends
summarise key changes between waves
summarising key changes between waves
Compass is particularly useful when working with multiple waves, as it can help you quickly surface patterns and draft narratives that explain how results have evolved.
It’s important to always review Compass outputs alongside the underlying charts and data. Compass summaries should be used as a starting point for interpretation rather than a final conclusion.
Exporting multi-wave results
You can export results from multi-wave surveys for further analysis.
Options include:
exporting aggregated responses across selected waves (Excel or CSV)
exporting crosstabs that include wave comparisons
Things to keep in mind
Changes between waves may reflect audience differences, not just changes in sentiment or behaviour.
Consistent audience setup across waves improves comparability.
Aggregated views are more stable, while single-wave views are better for point-in-time analysis.