In constructing a frequency distribution, the approximate class width is computed as

▸

(largest data value - smallest data value)/number of classes

▸

(largest data value - smallest data value)/sample size

▸

(smallest data value - largest data value)/sample size

▸

largest data value/number of classes