Explanation of Index and Z-Scores

The Index indicates the degree of over/under representation of a variable category in your file versus the comparison base, where an Index of 100 means proportions are equal and an Index > 100 indicates an over representation of that category within your file. It is always best to use a comparison base that reflects your potential target market, to ensure that the Index is relevant.

The Z-Score is a measure of statistical validity of the result i.e. the probability that the Index result would have occurred by chance. A Z-Score > 2 usually indicates the result is statistically valid, and the higher the Z-Score, the higher the confidence in the result.

When reviewing and interpreting profiling results, it is therefore important to take account of the Z-Score as well as the Index. Z-Scores are heavily influenced by the size of the file being analysed. Where the file size is small the results are more likely to be inconclusive, reflected by low Z-scores.