Why a histogram needs frequency density
The table gives information about the lengths, minutes, of phone calls.
- : frequency
- : frequency
- : frequency
- : frequency
Leila draws a histogram for this information. She puts frequency on the vertical axis, so her bars have heights , , and .
Explain why Leila's histogram gives a misleading picture of the data. [1]
- 1
The class widths are not equal: they are , , and minutes. In a histogram it is the AREA of a bar that shows the frequency.
- 2
B1 With frequency as the height, the bar for has the largest area () even though it has the fewest calls. The heights should be frequency densities.
Answer
The class widths are unequal, so the heights must be frequency density (frequency ÷ class width). With frequency as the height, the widest class, which has only calls, gets the biggest bar.
Examiner tip
Examiner tip: The key words are "unequal class widths" and "frequency density". An explanation that mentions only one of them, such as "she used the wrong axis", is not enough.
Common mistake: Saying "she should have drawn a bar chart" or "the bars should have gaps". The data is continuous, so a histogram with no gaps is right; it is the vertical axis that is wrong.