If five intervals are used, we would have
w¼
67
5
¼ 13 : 4
and if 15 intervals are used, we would have
w¼
67
15
¼ 4 : 5
Between these two values, 4.5 and 13.6, there are two convenient (or
conventional) numbers: 5 and 10. Since the sample size of 57 is not large,
a width of 10 should be an apparent choice because it results in fewer
intervals (the usual concept of ‘‘large’’ is ‘‘100 or more’’).
- Since the smallest number is 12, we may begin our first interval at 10. The
considerations discussed so far lead to the following seven intervals:
10–19
20–29
30–39
40–49
50–59
60–69
70–79
- Determining the frequencies or the number of values or measurements
for each interval is merely a matter of examining the values one by one
and of placing a tally mark beside the appropriate interval. When we do
this we have the frequency distribution of the weights of the 57 children
(Table 2.2). The temporary column of tallies should be deleted from the
final table. - An optional but recommended step in the formulation of a frequency
distribution is to present the proportion orrelative frequencyin addition
to frequency for each interval. These proportions, defined by
relative frequency¼
frequency
total number of observations
are shown in Table 2.2 and would be very useful if we need to compare
two data sets of di¤erent sizes.
60 DESCRIPTIVE METHODS FOR CONTINUOUS DATA