|
[Top of Page]
RATE RELIABILITY
Vital statistics are easily influenced by random variation and single-year rates can fluctuate from
year to year. A multiple-year rate such as a 5- or 10- year average of single-year rates would be
more accurate in formulating conclusions on vital events. For example, the infant death rate for
Kansas has ranged from 8.8 to 8.2 infant deaths per 1,000 live births in 1992 and 1996,
respectively. The 1992-1996 5-year infant death rate for Kansas was 8.0 infant deaths per 1,000
live births. The 5-year rate smooths some of the variation in single-year rates and would be a
more reliable indicator of infant death rates in Kansas.
[Top of Page]
Rates based on a small or large number of events in a sparsely populated area can vary widely.
To exemplify the variation that may occur with a small number of events, in 1996, Wallace
County was the least populated county in Kansas with 1,812 residents and Sedgwick County was
the largest with 422,437 residents. With 18 deaths occurring in Wallace County in 1996, the
crude death rate would be 9.9 deaths per 1,000 population, whereas, 3,442 deaths occurring in
Sedgwick County represents a crude death rate of 8.1 deaths per 1,000 population. If 5 more
deaths occurred in each county (e.g., multiple-death accident) Wallace County's crude death rate
would increase to 12.7 deaths per 1,000 population. Sedgwick County's crude death rate would
increase by only a few hundredths and with rounding, still remain 8.1 deaths per 1,000 population.
Therefore, one must use caution when comparing rates of vital events between counties of
significantly different population sizes to avoid misleading conclusions.
[Top of Page]
|