Conclusion
The New Human Science is a combination of the new understanding of how we make decisions, our human nature, and developing technologies. These new combined insights and tools impact the way we make decisions, who we select, and how we select and motivate people.
In this book, I have shown how the new human science and advanced analytics can aid decision making; the critical role collaboration plays in organizational success; where value is derived from; and how eliminating biases from selection, promotion, and incentive decisions will also make for more equitable and successful workplaces. More information and tools to assist with decision making can be found at DecisionAnalyticsInc.com.
Much more needs to be done on this topic. More tools need to be built, and more topics need to be explored, including the following.
Garbage In...
This is so ubiquitous a problem that it hardly bears mentioning: All of this depends on the quality of the data. If we start with corrupted data, what we do with it and the insights it provides are largely useless.
Our Argumentative Natures
The English version of law, like much of academia, is largely based on an adversarial model. Both sides develop their argument as thoroughly as possible and fight tooth and nail. Obviously, there are upsides to this approach. However, the one tremendous downside is that neither side is really all that interested in the truth of what actually happened; they just want their side to win. This has the very unfortunate result of placing both sides in a position in which they deep six facts that might not support their specific position. The use of advanced analytics to better get at the facts of the matter needs much more attention.
Advanced Analytics and Diagnosis of HCM Issues
In 1999, then 3-year-old Isabel Maude came down with a high fever, vomiting, and a skin rash. The doctors diagnosed chicken pox but failed to identify the much more serious condition that developed: necrotizing fasciitis (or what is more commonly referred to as the flesh-eating disease). It took a substantial amount of time for the physicians to get the diagnosis right, and that delay almost cost Isabel her life. It also resulted in a long series of plastic surgery for Isabel.
The problem was anchoring bias. The doctors diagnosing the illness were simply certain that they had gotten it right. Had the flesh-eating disease occurred to them, they could have almost certainly made a more accurate diagnosis. One positive that came out of this near tragedy was Isabel’s parents establishing the organization Isabel Health Care, which develops decision-support software for doctors. We need to develop an evidence-based deep Q&A expert system that can assist with determining human capital management (HCM) issues and solutions.
The Science (and Art) of Prediction
There are two ways to view the world: deterministically or probabilistically. It is difficult to argue that the world is entirely one or the other (although many hope we have some degree of free will). Though what we observe largely appears to be deterministic, we also see that the world seems to contain a healthy dose of randomness. The way products are arranged in a supermarket may not cause us to purchase goods, but it may increase the likelihood. We are not always going to choose the right employee, but if we apply some simple state-of-the-art decision mechanisms and techniques, we can increase the likelihood.
The Challenges with Being Empirically Declarative
These problems plague large-scale academic research because it is not clear whether we are really measuring employee engagement or effectiveness. It could well be something else that we have not identified that is actually causing the result we are seeing. Or it might be the proverbial cart-horse issue: Is the well-performing company pulling the engagement or effectiveness or vice versa? We can go some way toward reducing the impact of each of these problems, but it is difficult to eliminate them entirely.
One reason for many of the problems with these large-scale research projects is they would like to say something about generalizability. That is, do the effects observed apply to everyone? Each of these provides challenges to actually being able to declare unequivocally that the practices are causing the result. The unfortunate reality is that nearly all the academic research (including my own) suffers to some degree from one or more of these challenges. Fortunately, firms have a much better chance of conducting research into the impact of practice and policy choice for their specific situation, so generalizability is less of an issue. Driving forward with information on “what works” is what is important, and the answer largely depends on when and where.
Decision-Making Authority and Cooperation
The critical issues that I want to develop in relationship to the Challenger story are these: where the critical information resides, who has the authority to make decisions, and the role of cooperation. The issue here is that there are many within the organization who have critical information, and they need to be in a position where they can share that information with decision makers.
Sharing Control and Return Rights
So, how do you foster an environment of collaboration? The answer is that you share control rights and return rights with your employees. There is powerful incentive effects associated with transferring some of the rights of ownership to employees. Control rights are by definition the right of ownership that allows us to decide what we want to do with assets we own. Return rights are the rights to any revenue generated by those assets.
Individualization
Another potential benefit of advanced analytics is the ability to utilize technologies to tailor to the characteristics of the specific situation and the specific individual. There has long been a debate between the use of best practice and strategic choice. In essence, best practice research holds that there is a universal set of practices that everyone would benefit from using. Strategic choice holds that the situation dictates the specific policy and practice. I am firmly in the latter camp.
Have a look at the regression output using simple ordinary leased squared (OLS, a process of finding the line of best fit between data points by finding the difference between data points, allowing an estimate when the data is not available). Does anything strike you about the output? If you look closely at the line of best fit (the line running through the data points) either no one or very few actually fall directly on the line.
As we sequence our DNA, we can better individualize our healthcare and our nutrition. Similarly, advanced analytics will better allow us to get the right person in the right job and to determine what actually motivates that person as an individual.
Additional topics that need considerable thought include the following:
• Agent-based modeling: This could allow for a much more realistic model of how people actually respond to incentives or other interventions.
• Neuroeconomics: This emerging science has tremendous scope for better understanding decision making and how to do so more efficiently.
• Combinatorics: The branch of mathematics that provides insight into how rational decisions are made.