A statistical test, a two-group mean-comparison, can show whether differences are meaningful between benchmarks

With the data from the ongoing survey of General Counsel Metrics, it will be possible to tell whether departments in the UK, for example, differ in their fundamental metrics from departments in France. Certainly the metrics will differ, such as internal spend as a percentage of all legal spend. But how can we know if the difference is big enough to deserve commentary? Might it simply be an artifact of the particular participants or random fluctuation in values?

A statistical test called the two-group mean-comparison can look at the average of both countries and calculate the standard deviation of the metric in both. It then calculates the 95 percent confidence interval, which means the likelihood that 95 times out of a hundred the difference is real. This final step calculates that so-called p-value. The lower the p-value, the more likely the difference is significant from the statistical standpoint. Whether the difference is meaningful in the real world is an entirely different question.