Blog

Introduction to Legal Analytics, Part 1

Legal analytics is changing not only the legal profession, but its relationship to other areas of the market as well. Legal analytics affects us all, because it is changing how individuals and companies use data to increase the efficiency of the decision-making process. This concerns every level of the market, from a simple civil case between two individuals, to the potential ramifications of a highly influential Supreme Court decision, to the solutions a large corporation may use as it adapts its strategies to compete in a changing market.

What Legal Analytics Is Not: Causal Inferences

An understanding of legal analytics begins with an understanding of causal inference. A causal inference is a conclusion that can be drawn from gathering data about an event’s occurrence. For example, a law firm might be looking into multiple workmen’s compensation claims in a large manufacturing corporation. There appears to be one factory within this corporation, though, that files a statistically significantly higher number of claims than all others. From this data, we might make a causal inference about sub-par conditions at this one factory, rather than making an uneducated guess about a company-wide lapse in policy.

But jumping the gun like this could be an uneducated guess as well. A causal inference is an incredibly useful tool in data management, but it is not the most powerful tool we have. Here’s where it might fail. In our example above, we made an inference based on a few pieces of information about a factory in a large corporation. But maybe there were factors we never even thought to consider. Do we know where the factory is located? Perhaps this factory is in a different country from the others. A difference in the government regulatory apparatus could be an unconsidered factor. Maybe there is an issue with the breakdown of machine parts in adverse climates. Maybe the factory is experimenting with a different management structure, or the managers themselves have chosen to re-prioritize efficiency over safety without keeping their superiors apprised.

How do we find the factors with the most impact? How do we separate correlation from causation? There could be a hundred facets to this factory’s problem, each with thousands and thousands of data points that all come together to paint a larger picture. For this reason, we need something more robust than a causal inference. We need to take out the guesswork. We need to make a predictive model.

That’s where legal analytics comes in.

This article is part of our 7-part Intro series. The others can be found here.