Master Plan Nashua Fire Rescue, NH
Figure 82. Total Response Time Continuum
leet iit
Response Time
Total Response Time
Historically, fire rescue service providers have used the performance measurement of average response to
describe the levels of performance. The average is a commonly used descriptive statistic, also called the
mean of a data set. Averages may not accurately reflect the performance for the entire data set because the
average can be significantly skewed by data outliers, especially in small data sets. One extremely good or
bad value can skew the “average” for the entire data set. Percentile measurements are a better measure of
performance since they show that most of the data set has achieved a particular level of performance. The
goth percentile means that 90% of responses were equal to or better than the performance identified, and
that the other 10% can be attributed to data outliers, inaccurate data, or situations outside of normal
operations that delayed performance. This can be compared to the desired performance objective to
determine the degree of success in achieving the goal.
An important consideration when evaluating fractile performance is that the results of each category are not
additive, meaning that the sum of two or more constituent metrics cannot be simply added together to find
the sum. This is because each dataset is discrete and, as such, must be observed individually, particularly
when data quality is an issue. If a metric, such as response time, possesses the majority of its data points,
while turnout time is not accurately documented, a significant difference can exist between the response
time calculated using the fractile descriptive and the sum of turnout time and travel time.
In evaluating the various response time components using the fractile analysis method, each component
must be evaluated and quantified separately, as the available data—and the quality of the data may vary
significantly.
Emergency Services
ESC Consulting International 122
