Pareto analysis is a prioritization technique that identifies the most significant items among many. This technique employs the 80-20 rule, which states that about 80 percent of the problems or effects are produced by about 20 percent of the causes.

Brief summary of characteristics

• Used as a risk assessment technique at any level, from activity level to system level
• Yields broad, quantitative results that are graphically depicted on simple bar charts
• Depending on the information analyzed, generally requires some form of data tracking (e.g., monitoring the number of accidents caused by piloting)

Applicable to any activity or operating system

Most common uses

• Most often used to rank activity or system accidents
• Can be used to rank the causes that contribute to accidents
• Also used to evaluate the risk improvement that results from activity or system modifications with "before" and "after" data

The following graph is an example of the final results from a Pareto analysis.

Pareto Graph of Propulsion System Problems

Limitations of Pareto Analysis

Although Pareto analysis is highly effective in identifying the most significant contributors to activity or system problems, this technique has three limitations:

Focuses only on the past. Pareto analysis develops risk-related characteristics for an activity or system based solely on the numbers and types of problems encountered in the past. While Pareto analysis offers a valuable look at key contributors to past problems, the exclusive reliance on historical data can be misleading in the following ways:

1. The data under-represent events that, luckily, have not happened yet or have occurred rarely but that, statistically, are just as likely as events that have occurred more frequently. This can skew decisions and resource allocations, especially when a relatively small total number of problems has occurred for individual components or types of components.
2. Recent changes in operating practices, maintenance plans, equipment configurations, etc., may invalidate historical trends, or at least reduce their accuracy. This situation can also skew decisions and resource allocations, both when relatively recent changes have not been in place long enough to affect the data or when data are analyzed over extremely long time intervals during which numerous changes have been made.

Variability in levels of risk assessment resolution. Deciding how to group elements of an activity or system for a Pareto analysis is an inherently subjective exercise. It produces significant variability in (1) the time required to perform the analysis and (2) the level of resolution in the results. Grouping elements at too high a level may mask significant variations among elements in each group. On the other hand, grouping elements at too low a level may falsely indicate relative importances of individual components.

Dependent on availability and applicability of data. The quality of Pareto analyses is completely dependent on the availability of relevant and reliable data for the activity or system being analyzed. A diligent focus on collecting meaningful data is critical to a successful Pareto analysis.