In the latest Budget, Navy/Marines have both placed initiatives and a larger share of their emphasis on learning. Senior leadership has recognized that if the Services are truly going to transition into a more distributed force, their Sailors and Marines need to be better decision makers.
There is a varying need and depth of decision-making instruction based on an individual’s rank, position within their community, and their operational environment. Because of this demand signal, leaders have called for implementation of a new component for strategy, decision science.
But isn’t decision-making already a part of our job description? What is this “decision science” how does it help the Navy and Marine Corps team? With this overarching question in mind, the Commandant’s new learning could not be a more timely, relevant, and essential doctrine for the Navy and Marine Corps to adapt to this new challenge.
The Chief of Naval Operations has called for incorporating “decision science” training into leadership development programs throughout the Navy to improve the service's understanding of human judgment and decision-making—such decision science training has only been experienced at the highest Captain and flag-levels.
Improved decision-making is a decisive advantage in stressful conditions and enables successful mission command, for example reacting to factors contributing to the decision in tactical operation. Much of it has to do with how quickly the decision has to be made, along with how much information is needed to make the decision.
If commanding officers received this lucrative "flag-level" training, one thing is certain the tools of decision science would be a beneficial tool for their judgments. So how can decision science aid decision-makers?
There is still much to be discovered, but it’s best to start with bringing structure to what decision science could mean and answering how the Navy got here.
Decision science focuses predominantly on choices in highly uncertain environments-- over arching themes cam be broken down into a collection of three core elements:
Analysis: how decisions ought to be made.
Description: how decisions actually are made.
Interventions: efforts to try and change the two
Decision science is not a combination of decision making operations research, and data science. It’s a fundamental shift in how to approach problems with probability tech.
And it is definitely not equal to data science alone. Data science is “the ability to take data — to be able to understand it, to process it, to extract value from it, to visualize it, to communicate it.” While the Navy and Marine Corps are heavily investing in operations research and data science, decision science has not experienced such prominent support and investments to date.
In addition to prescribed number crunching frameworks for decision making, leaders who are well-versed in decision science can structure decision environments to achieve better judgments. This method involves carefully examining how choices are structured, what incentive and feedback structures are in place, how value judgments are made and distinguished from scientific judgments, how uncertainty is expressed, and what types of path dependencies exist.
Though the curricula may need to be left with outside experts, it will be essential for both civilians and military members to work together to form frameworks to help implement decision science in the Navy.
What Is a Decision Science Framework?
Consider the following two problems:
Problem 1: Imagine an aircraft carrier has just completed a port visit in country where a new bug has been spreading. Several days after the visit, 600 Sailors catch the bug and are expected to be lost. There are two alternative treatment programs. If Program A is adopted, 200 Sailors will be saved. If Program B is adopted, there is a one-third probability that all 600 Sailors will be saved and a two-thirds probability that no Sailors will be saved. Which do you prefer, Program A or Program B?
Problem 2: Imagine an aircraft carrier has just completed a port visit in country where a bug has been spreading. Several days after the visit, 600 Sailors catch the bug and are expected to be lost. There are two alternative treatment programs. If Program C is adopted, 400 Sailors will be lost. If Program D is adopted, there is a one-third probability that nobody will be lost and a two-thirds probability that 600 Sailors will be lost. Which do you prefer, Program C or Program D?
This decision-making experiment has been replicated dozens of times. Most individuals responding to Problem 1 preferred Program A (the certain outcome of saving 200 Sailors) and most responding to Problem 2 preferred Program D (the gamble). However, Program A/C and B/D have the same outcomes.
In applying the decision science framework to this case, you can immediately see that making an accurate and informed decision requires reliable data on the expected lethality of the virus, and the probabilities of survival of the different treatment programs. Obviously, in the current environment, it is nearly impossible to know "expected values” of every situation.
However, applying decision science concepts enables leaders to critically and methodically work through decision-making situations in ways that mitigate the influence of biasing factors such as the framing of the information.
This is just one foundational way that decision science can be applied, so how does it apply to leaders within the Navy?
Despite dramatic technological gains the Navy has recently experienced numerous operational and administrative decision-making failures. While politics influenced these situations immensely, a number of problems were exacerbated by confirmation and sunk cost biases that hindered sound decision-making by members of the fleet.
But before we begin to implement or talk about AI, data science, or technology in assisting decision-making, service members need to recognize biases in our own decision-making process. Technology may have changed, but the way we approach decision making has relatively stayed the same – this is the value add with decision science.
Whether it’s a commander using probability tech-- what is the chance a desired outcome will materialize, on the bridge of the ship or a Marine leader operating in combat, the Navy and Marine Corps need to be more proactive at adopting the concepts of decision science to better equip their Sailors and Marines. The Marine Corps does teach decision-making to its officers at The Basic School and Infantry Officer’s Course, but overall there needs to be more reinforcement.
What do recent Navy Instructions reveal to the rest of the service? It was the first of its kind to bridge the applicability of decision science techniques to topics the Navy cares about. Several components of the document referencing decision making include:
Understanding the rise of global information systems, especially the role of data in decision making, is one of the forces that continues to shape our modern security environment
Establishing data-driven decisions as a foundation for achieving readiness in the warfighting enterprises by developing and maintaining authoritative and accessible data for decision-quality information.
Using quantitative techniques, data-driven analysis, and other research to catalyze Navy leadership development by the end of 2020 and using science-based practices and training to support leader development and better decision making.
The report states:
“In the Decision Science Environment, where leaders have to deal not only with incomplete data but also with analysis and decision making in a world that involves overwhelming data, essential elements of future success include: ability to evaluate information, reason strategically, act decisively, possess good judgment, creativity, and excellent analytic and problem-solving skills.
Navy force must be more proficient by improving strategic thinking, increasing geopolitical awareness, building key technical and professional capabilities, and deepening our understanding of the conditions in which military force can be used effectively.
With these goals articulated and underpinned by the right measures, the Navy is on the right track. Updating training programs to leverage recent advances in the field or decision science will be key in preparing Sailors and Marines for combat in the future.
Decision Making Science is Essential for Distributed Military Operations where informed decisions requires foundational understanding of the human influence in decision making. Advances in weapons systems, Command, Control, Communications, Computers, ISR, and Targeting, data analytics, machine learning, and artificial intelligence have accelerated the pace of warfare, forcing humans to make rapid decisions under conditions of risk and uncertainty.
In order to maintain an edge in decision speed and quality, we will need to augment human cognition with machine driven analysis. The return to great power competition has necessitated the rapid development and evolution of the methods and tools by which war is conducted, and at the same time has re-emphasized the importance for understanding the timeless nature of war.
Despite technological advances, modern equipment relies on dated and vulnerable infrastructure that can be compromised by forms of enemy interference – potentially leaving operating forces in an environment denied of GPS data, communications, situational awareness, or command and control capabilities.
An operating environment that is uncertain, complex, disorderly, and fluid requires Commanding Officers and their subordinates to make sound decisions, act independently, and conduct combat operations within the commander’s mission and vision – working in a dynamic risk environment without the ability to seek advice and guidance from superiors.
The ability of each Sailor and Marine to be critical, reason strategically, and exercise good judgment and decision-making under conditions of risk and uncertainty is a common strand uniting both the conduct of naval operations on the leading edge of technology and in a technology-compromised environment.
“So What”? Is Every Marine Rifleman going to be a Decision Scientist? Every Sailor firefighter a Decision Scientist?
While all individuals benefit from understanding the concepts of decision science, not everyone needs to be a decision scientist—much can be left to experts. However, the Navy needs to train more than just flag officers, it must train the rest of the force on such a critical piece of a Sailor and Marine’s job description, decision-making.
Expert supervision will be required to develop and initially administer training force wide. Leaders need to understand enough to require a thorough analysis of decisions and ask the right questions – especially in distributed maritime operations.
Navy does have the skill to provide this analysis – they just do not presently have a structure to make it work. The demand signal is growing but it will take time to develop the required expertise at the unit level, the sooner this happens the better.
There currently is not a clear pathway to implementation of these concepts in both the Navy and the Marine Corps and it remains to be seen how leadership will incorporate the field of decision science into the force.
Leaders must incorporate decision science into their organizations to meet this demand, and more importantly, enable access to the tools to be better decision-makers.
Summary of the Intelligence Community’s Analytic Tradecraft Standards
1. Sourcing
Provide at least basic descriptions of the sources of information that support each analytic conclusion e.g., “according to a senior official with firsthand access, ; identify key sources that are most important for each analytic conclusion; be transparent about the quality of available sources.
2. Uncertainty
Explain the level of uncertainty for each analytic conclusion; use approved terms e.g., “likely” or “very likely” to express the likelihood that the assessed event or outcome will occur; express a confidence level based on the quality of the overall analytic argument.
3. Distinctions
Distinguish between underlying evidence, assumptions, and analytic conclusions; consistently use “signaling language” to alert decision-makers as to the type of information they are reading; be transparent about key assumptions that underpin each analytic conclusion.
4. Alternatives
Identify at least one plausible alternative for every major analytic conclusion to mitigate surprise or alert decision-makers of low-probability/high-impact situations; identify indicators that, if detected, would alert decision-makers that an alternative conclusion is becoming more likely.
5. Relevance
Ensure analytic products are tailored to the needs of decision-makers; provide deeper insights to decision-makers by addressing second- and third-order impacts; conduct opportunity analysis as appropriate.
6. Argumentation
Prominently present a main analytic conclusion up front; subordinate analytic conclusions should support the main conclusion; skillfully combine evidence, logic, assumptions, and information gaps to support analytic conclusions.
7. Analytic Line
Be transparent about how analytic conclusions are consistent with or different from previously published analysis; alert decision-makers if there are significant analytic differences between two or more intelligence organizations.
8. Accuracy
Ensure clarity of message in all analytic products; do not avoid difficult analytic conclusions in order to minimize the risk of being wrong
9. Express absolute, rather than relative, probabilities
Many Reports contain language for example, “likely” instead of “more likely”; the agency also requires its analysts to assess events, actions, or behaviors instead of cognitive states or beliefs.
10. Visualization
Use visual information to clarify, complement, or enhance the presentation of analysis.