The Earth shrugged and 230,000 people died. The Earth shrugged and our frailty, our tenuous hold on life once again became apparent. On Sunday, December 26, 2004, a 9.1 magnitude earthquake occurred 100 miles off the coast of Sumatra, Indonesia. It was one of the deadliest natural disasters in history. The earthquake triggered tsunamis, massive ocean waves of power and destruction that traveled at 500 miles per hour, killing people in 15 countries, devastating towns and villages, resulting in the displacement of nearly 1.7 million individuals. The time difference from the moment of the earthquake until the tsunami hit was between 15 minutes and seven hours depending upon location. There were no early warning systems located in the Indian Ocean to detect the tsunami event or to warn those threatened by it.

Prevention is what early warning systems are about. Not that  an early warning system could prevent a tsunami, but rather, an early warning system that is indicative of an imminent event and allows people sufficient time to take prudent actions would reduce the loss of life and/or property. Early warning indicators are greatly desired by people and by organizations to help reduce uncertainty and deal with the risk inherent in their respective environments. Prevention is only possible with accurate and timely early warning indicators.

One location where people evacuated prior to the tsunami was on the island of Simeulue, located not far from the epicenter. On this island, folklore told of an earthquake and tsunami in 1907, and described the warning indicators. The islanders fled inland when the events they were witnessing matched their folklore, and they were able to reach higher ground prior to the tsunami reaching their location. I am a fan of folklore as it helps to convey the cultural richness of a society, but a reliance on folklore may not be the best that science can do to provide early warning indicators.

Some animals that experienced the tsunami were seemingly more sensitive than we humans were, with built-in systems making them uneasy and perhaps keeping them more in-tune with their natural surroundings. Just prior to the tsunami hitting, there were reports of elephants bellowing and running for higher ground, moving away from the coming waters. Dogs refused to go outdoors. Flamingos abandoned their low-lying breeding areas and zoo animals rushed into their shelters and could not be coaxed to come back out. A reliance on observing animal behavior, which is subject to all kinds of misinterpretation, may also not be the best that science can do to provide early warning indicators of oncoming events.

In the past, animals have been utilized as early warning indicators for humans. For example, miners used canaries as early warning indicators for a buildup of toxic gases within the mine. While they effectively served as early warning indicators for the miners, the canaries themselves did not do so well simply succumbing to the gases faster than the miners would.

Organizations, of course, come in all sorts of shapes and hues, and many of them (if not all) over the years, sometimes overtly and sometimes inadvertently, have developed warning indicators to which they pay attention. These warning indicators allow them to perform their tasks better, either responding more quickly or avoiding pitfalls and obstacles and increasing the likelihood of achieving organizational goals and success.

Indicators are currently being used by organizations in the attempt to prevent or warn of issues in diverse areas such as sales forecasting, cash flow crunches, credit risk, currency crises, economic outlook, famine, drought, other weather related events, such as tornadoes and hurricanes, disease resistance and potential terrorism targets. One database used as an indicator of propensity for political change is called the “Conflict and Peace Database.” This database catalogs world events surrounding conflict and peace in order to give researchers new insights into the precursor environments that lead to those outcomes. Having new and improved warning indicators is routinely sought, and there are markets for those that can build a better mousetrap. U.S. Patent #7291237 awarded November 6, 2007, for instance, was for a new indicator to be built into car tires warning of excessive wear, an early indicator of tire failure.

Warning indicators are a type of surrogate measure. We are surrounded by surrogate measures; they exist everywhere. A surrogate measure is when a particular measurable or observable condition is assumed meaningful about an underlying state or condition. They are often used when it is difficult, impossible or unknown how to measure the underlying condition directly. For instance, nutritional surveillance, counting calorie consumption, is used by aid organizations as an early warning indicator and a predictor of a coming famine. You rely on the gas gage on your car dashboard to give an accurate indication of the level of gas in the tank. You don’t visually inspect the level of gas in the tank with your own eyes because it would be difficult to do while traveling down the road and because visual inspection would not be as accurate as a well-designed gas gage regarding the amount of gas left in the tank.

Surrogate measures abound in the medical world—doctors examine patients for all kinds of signs and symptoms of underlying medical conditions that are not being directly observed. These initial signs may lead toward more tests, which may be more evolved or advanced surrogate measures to help further define ills. The shaky hand may direct the doctor to order a more advanced test (and expensive) for Parkinson’s; swollen ankles with excessive fluid may be an indicator of heart failure. The world of medicine today may have discovered the ultimate early warning indicator for human health, with DNA analysis providing the knowledge of disease risk for an individual. When does an indicator cross the line from being simply an indicator to being the answer of what is to happen? Maybe it doesn’t really matter as the ultimate goal is to constantly improve our warning indicators, our surrogate measures, until they are 100% accurate, 100% of the time.

One subset of surrogate measures is called unobtrusive measures. When surrogate measures are used to make assumptions about an underlying condition in such a way that the people affected may not realize that they are being measured, the measurements are called unobtrusive. An example would be looking at the wear and tear of floors in museums to determine the popularity of exhibits. In addition, the wear and tear around certain types of exhibits could be used as an indicator as to how staff should be deployed in the museum to handle expected crowds for newly installed similar exhibits. The assumption being made that those exhibits surrounded by excessive wear and tear are more popular than those holding up well (as opposed to the possibility of poor quality flooring in front of those exhibits).

Another aspect of unobtrusive measures has to do with the methodology of the measurement process itself. It is important that the way the measurements are obtained, the mechanics used to collect the predictive information be as unobtrusive as possible. You want to minimize the impact that the act of measurement will have on the measurement itself. For instance, if you were to observe a worker conducting his/her job, the very act of observation may affect how the worker carries out the job, causing the measurement to be biased. A worker upon noticing that they are being observed may decide to speed up in his/her work, showing off, instead of working at the normal pace. The worker may also decide to slow down if he/she thought you were involved in determining how fast the job should be performed, potentially affecting compensation. Asking people about their work environment in an inappropriate fashion can also affect how they might respond on an employee survey instrument. For instance, putting emotionally laden terms into what should be a neutral survey item stem or asking questions in a leading way could lead to bias or a skew in the survey results. Unbalanced or emotionally laden scales or response choices can have the same result. Employee survey results will be affected by the way the questions are asked—the goal is to minimize that impact.

A set of potentially bias-related surrogate measures are used when we observe people, their appearance, their dress, hair length, skin color, age and gender, etc., and use these observations to make assumptions regarding their underlying characteristics, their beliefs, expected behaviors or capabilities. These types of surrogate measures are usually very poor warning indicators— not scientifically based—and quite unrelated or correlated to the reality of beliefs and behaviors of others, which lead to inaccurate conclusions.

With the appropriate warning indicators in place, there is no such thing as a “sudden crisis,” only a lack of information caused by a reliance on poor indicators, the poor use of good indicators or a lack of appropriate analysis from the information provided by the indicators.