Standardized Residuals calculation is an effective method of brand equity analysis which helps to measure associations between brands and statements (often image attributes) in order to identify brand profiles.
The main advantage of this method is that it reveals relative rather than absolute brand image by taking into account the “effect of large brands”, that is when consumers tend to choose a larger number of statements for market leaders than for smaller brands.
Calculation principles
A residual is the difference between the observed value Y and the expected value Х. Each residual is adjusted by the standard deviation of all the residuals:
Adjusted residual = (Y–X) / √ (X * (1 - row total proportion) * (1- column total proportion))
The adjustment allows for the comparison of different brands and statements, as the adjusted residuals are distributed normally.
How to calculate residuals
First you need to build a Grid Report with Vert% or Count metrics: list of attributes should be in rows and brands in columns.
Switch to Chart mode;
Toggle the Comparison tab and select the ‘Residuals’ comparison type at the Chart setting menu;
Set the residuals settings:
Select a critical (threshold) value to indicate significant residuals;
Select an indicator for visualization ('chart' option is in the example);
Select the colors to indicate significant positive/ negative residuals.
Significantly higher residuals indicate a positive association between equity and brand. Significantly lower residuals indicate a negative association.
Then select an appropriate chart type to represent the residuals. We recommend using a Grid chart, but you can also try other options.
