Analysis of Variance (ANOVA)

The trick to effective analysis is to identify what is significant. It is not enough to look at simple reports and try to draw conclusions from them. Often, in glancing at a Section Summary, you notice differences that look interesting or you see what appears to be a meaningful relationship. But are these findings really important? Are the differences big enough to be “statistically significant”?

Statistical significance tells you if the differences you see are random, or if they are sufficiently large to justify further consideration. If the differences are random, it means the results are what would reasonably be expected to happen. If the differences are statistically significant, it means they are higher than could be expected to occur, indicating the potential influence of some non-random factor.

The ANOVA technique is another example of an advanced analytical tool; it identifies whether the differences are significant or meaningful, according to the profile of your employees or customers (e.g. are there differences in ‘intention to stay’ based on people’s location, salary or length of time with the organisation? Are those employees who are dissatisfied with their job different from others in any way? How do the views of customers in different locations vary? What are the differences by product?).