To compare different metrics, use the technique of stating them as standard deviations

If you standardize the values of variables that vary on their measurements, you can compare their relative magnitude. For example, say you want to match how your department stands on cases per lawyer against paralegal base pay. The two metrics exist on incommensurable scales. But you can express both metrics in standard deviations and thereby compare your positions. On cases per lawyer your department might be 1.2 standard deviations to the right (more cases than a benchmark group), while on paralegal base comp you are only 0.8 standard deviations to the right (not as much higher). You can now say the degree to which one metric compares to the other using the shared terminology of standard deviations.

This technique is shown to great effect in an excellent chapter in Laura Empson, ed., Managing The Modern Law Firm: New Challenges New Perspectives (Oxford Univ. Press 2007) at 113-114. The authors translate several variables of what they call embeddedness into a single chart on which are shown lines of different slopes. There are lines for Board membership, cost of goods sold, repeated ties, and relational status — all related to partner billing rates. Since each of the variables has different units it is not possible to directly compare the sizes of their effects. However, by standardizing the variables and expressing them in terms of standard deviations it is straightforward to see the different effects they have on the dependent variable, partner billing rates.

We welcome comments

Your email address will not be published. Required fields are marked *