Data Should Drive Decisions: It can help counter conscious or unconscious bias

Friday, August 25, 2017 - 16:31

 

Day after day, managers confront operational problems, think about them and make choices about what to do or not to do. In other words, they decide something. But they don’t always take into account the data available to them when they make those decisions.

For law departments and law firms, analyses and visualizations of numbers can help managers make better decisions. If they collect metrics and weave them into their pondering, the outcome is often sounder and easier to explain. Let me demonstrate what I mean by starting with a plausible management decision and then outlining five situations where the contribution of data could improve the quality of that decision.

Assume that two or three managers in a legal group have met to figure out whether to hire another paralegal. Further, assume that our managers disagree on the appropriateness of adding paralegals. Their decision will improve if the managers have relevant data, analyze it effectively and apply it to their problem. How can numbers help them objectively think through the problem and find potential solutions?

 

Sidestep Cognitive Fallacies

Data that has been accurately and comprehensively collected can counteract many of the cognitive biases that afflict managers. Nobel laureate Daniel Kahneman has studied and illuminated many cognitive fallacies. In his book “Thinking, Fast and Slow,” Kahneman explains many of the quirks in our reasoning that deviate significantly from classical notions of rationality. Unfortunately, we are usually unaware of the gremlins in our minds that attack what we believe to be our clear-headed, balanced evaluations. Consider four cognitive fallacies and how data might help correct them.

  1. Framing: The managers’ thinking might be skewed by “framing.” For example, one of them may start arguing for more paralegals by asserting that “we should have six paralegals, at least one for every two lawyers.” That preliminary high demand will likely frame how other managers consider the decision. It warps the mind frames of the other two because it creates a baseline, a benchmark, a seemingly pivotal number to negotiate around. Yet it might be baseless. An antidote to framing could be benchmark data on paralegals per lawyer in firms or law departments.
  2. Salience: The bias of “salience” also warps rationality because it elevates a notable example, which may not be representative or appropriate at all, but which undeservedly shifts impressions of what is typical. A slew of counterexamples might be forgotten when a manager proclaims, “I just read that GE has one paralegal for every lawyer!” We shouldn’t necessarily focus on the latest example or most famous one just because it comes to mind easily. To blunt its potential impact, someone could gather articles that report average lawyer/paralegal ratios based on surveys of many companies.
  3. Confirmation bias: Data also can ameliorate “confirmation bias.” An example of this phenomenon is a manager who has consistently had good paralegals and ignores evidence of incompetent ones. It blinds people to examples that are contrary to their opinions or expectations. Perhaps the mixed evaluations that lawyers in the managers’ group make of paralegals would be data that challenges the one-sided, rosy view.
  4. Risk aversion: A fourth mental frailty, “risk aversion,” tilts our reasoning to fear losses more than favor gains. Some studies suggest that pluses must be twice minuses in our internal calculations for people to press forward. The downsides of making a call one way or the other (“We might hire a dud and never be able to get rid of them!”) loom larger than the countervailing upsides. On the other hand, a manager who argues for more paralegals might be risk averse because the group never wants to invite a crisis due to lack of resources.

Good data can combat each of these mental distortions as well as others that psychologists have catalogued. Facts and figures foil figments and fictions.

 

Uncover and Query Empirical Assumptions

When people make decisions, they often neglect to articulate the factual assumptions on which they base their conclusions. Worse, they may not even realize that they have been motivated by unstated (and usually untested) beliefs about how common something is or how much there is of something measurable. For example, one manager accepts on faith that lawyers will delegate work to paralegals, while another trusts without verifying that it will be easy to find, hire and retain capable paralegals.

If underlying assumptions such as these are not surfaced, and if there is no data to undermine or support their correctness, managers are apt to make weaker decisions than they should. Apposite data can be used to bore in and counterbalance statements and assumptions that might otherwise be accepted unquestioningly. In our scenario, that might include data collected through interviewing lawyers about how often they delegate tasks to paralegals – and how much more they might delegate. It could also include data from other legal groups that hire contract lawyers, temps or other staff instead of hiring full-time paralegals.

 

Disrupt Entrenched Values

As data becomes available for decision-makers, they should incorporate it and change how they view the probability of being correct. Around 250 years ago, Rev. Thomas Bayes first described what is now called Bayesian probability analysis. With a Bayesian view, the arrival of new data changes what are called “priors” and helps make predicted outcomes more accurate. Perhaps one of the managers believes at her core that lawyers are superior to paralegals, which is her prior.

New findings alter how Bayesians view the world; they should cause thoughtful people to reconsider their animating beliefs. Those values might be modified, for example, by the arrival of client-satisfaction survey praise for paralegals relative to lawyers.

 

Delay Premature Conclusions

When making a decision, it is crucial not to seize upon the first plausible solution to the problem. Rather, we should keep exploring alternative possibilities. For example, it might be possible to modulate the flow of work coming in from clients, which would change the equation for adding new paralegals. Or you can gather new data by drilling to discriminate more finely among the data you already have (are certified paralegals more desirable than uncertified paralegals?).

Data helps generate new possibilities to address a problem or to encourage managers to think about the problem longer. Regarding paralegals, it could be useful to find out how the current paralegals allocate their time among various tasks. Or exit interviews may have created a database of germane metrics.

 

Resist Peer Pressure

In the context of a group decision, solid data can serve as a talisman against the blandishments of a rhetorician, the rigidity of a zealot or the power of a senior executive. In the face of credible and on-point figures, it is harder to smooth-talk a contrary view, stick to unreasonable guns or ram a decision down a committee’s throat. “Yes, you’ve repeatedly said we don’t have cube space, but we’ve hotelled a dozen people for two years!” A manager who disagrees with the other two managers might be more inclined to buck them if some metrics back up the point.

This is not to claim that good data always means good decisions. It is to claim that operational data can help steer managers to reach a good decision more frequently than may now be appreciated. To be sure, the toughest decisions tend not to have decisive metrics. But even the gnarliest decisions – those that entangle personalities, traditions and long-range projections, or involve visceral disagreements over fundamental values – can benefit from whatever dollops of data are available.