"Hindsight bias" could hide real lessons of Columbia accident report,expert says (Forwarded)

Research Communications
Ohio State University

Contact:
David Woods, (614) 292-1700;

9/2/03

"HINDSIGHT BIAS" COULD HIDE REAL LESSONS OF COLUMBIA ACCIDENT REPORT, EXPERT SAYS

COLUMBUS, Ohio -- A psychological effect known as "hindsight bias" might cause
people to misinterpret the conclusions of the Columbia Accident Investigation
Board (CAIB), according to an Ohio State University researcher who helped the
board during its probe.

The report makes important recommendations that could not only help NASA, but
also other organizations, according to David Woods, professor in the Institute
for Ergonomics and co-director of the Cognitive Systems Engineering Lab at Ohio
State. Woods is also a professor in the Department of Industrial, Welding, and
Systems Engineering.

Woods provided technical input on decision-making, organizational factors, and
hindsight bias to the CAIB during its investigation. His research on the factors
that contribute to human error is referenced in Chapter 7 of the report.

Woods gave an example of the dangers of hindsight bias, citing some text from
the CAIB report. "Often the first question people ask about the decision-making
leading up to the Columbia accident is, 'why did NASA continue flying the
Shuttle with a known problem …?' " he said.

The "known problem" refers to the dangers of debris damaging the Shuttle wing
during takeoff, which the board has identified as the physical cause of the
accident.

But as soon as the question is posed in this way, readers risk oversimplifying
the situation and the uncertainties people faced before the outcome was known,
Woods said.

"Because it is difficult for readers to disregard what is commonly called '20/20
hindsight,' they can misinterpret the report and label NASA as a bad
organization. As a result, the same difficulties that led to the Columbia
accident could go unrecognized in other organizations, too," he added.

Just as the CAIB worked hard to overcome hindsight bias and find the
organizational factors that led to the accident, Woods feels that readers of the
report will also need to overcome their own hindsight.

The accident's lessons apply to all organizations that have to balance safety
risks with pressure for efficiency. The key difficulty, according to Woods, is
this: the most critical time for organizations to uncover safety risks is the
one time when they can least afford diverting resources to do so -- when they
are under heavy production pressures.

In the case of NASA, the organization is pressured to fly the Shuttle on a
certain schedule.

Woods noted that helping organizations maintain high safety despite production
pressure is the topic of a newly emerging area of research known as resilience
engineering.

"We can't change the past, despite the tragedy. But the future is open to us.
Will the next accident report again describe how safety defenses eroded over
time in the face of production pressure? If we recognize that the CAIB's
analysis applies to all organizations, not just NASA, we can learn how to
balance acute pressures for efficiency with a chronic need for high safety so
that this pattern doesn't recur," he said.

With nearly 25 years of experience diagnosing the factors behind human error,
Woods has conducted extensive research on how people interact with computers to
make decisions in high-risk environments. His work has won awards for improving
the safety of automated cockpits. In addition, Woods is a member of the National
Research Council committee that is helping NASA and the Federal Aviation
Administration plan the country's next-generation air transportation systems.