Z-scores help you understand whether the difference between two cells in a table is statistically significant. This is done by calculating the number of standard deviations between the values of both cells, and taking the underlying sample size of each column break into account. This helps you identify more statistically robust conclusion in terms of whether the two results are truly different. To help you to understand better, we use directional arrows to signify the differences.
To get started, click the Z-score button above the table. Now you will see cell highlighting as well as the z-score value appear directly below the cell % value.
Once toggled, you will see a z-score appear within every cell in your table, just below the % value.
The table below explains how to read the z-score ranges.
| Cell Colour | Z-Score Range | What it means? |
| Dark green | More than 1.962 | Highlighted cell is SIGNIFICANTLY higher than the reference cell (95% confidence level). By default, the reference cell is the adjacent cell with the Overall column. |
| Light green | Between 1.646 to 1.962 | Highlighted cell is SOMEWHAT higher than the reference cell (90% confidence level). |
| No highlight | Between 1.646 to -1.646 | There is no significant difference between the cross-break and the and reference cell. |
| Light red | Between -1.646 to -1.962 | Highlighted cell is SOMEWHAT lower than the reference cell (90% confidence level). |
| Dark red | Less than -1.962 | Highlighted cell is SIGNIFICANTLY lower than the reference cell (95% confidence level). |
You may also see some cells are highlighted either red or green. First, a highlighted cell regardless of colour means that that cell is significantly different compared to the reference cell (e.g. Overall column). And green highlight means that it's significant and higher than the reference cell, while red means it's significant and lower than the reference cell.
The image below will help you understand how to read your table when z-scores are toggled on.
Change root column for z-score calculation
By default, the z-score calculation will always compare each cell within your cross-breaks to the adjacent cell in the Overall column. This means that by default, you're always comparing to the total sample base. But you can change which column is used as the root for the z-score calculation by simply clicking the column header label.
In the screenshot below, we've changed the root column for the z-score calculation to the 16 to 24 age group column. This means that all other cells are now being compared to the 16 to 24 age group column instead of the Overall column.
