Question 1
1 markTom grouped the masses, m grams, of 60 oranges into the classes \(100 < m \le 120\), \(120 < m \le 140\), \(140 < m \le 160\) and \(160 < m \le 180\). He drew a histogram with 'Frequency' on the vertical axis, using the frequency of each class as the height of its bar.
Tom then joins his last two classes together to make one class, \(140 < m \le 180\).
Explain why Tom can no longer use the frequency of each class as the height of its bar.
Model answer
Before, every class was 20 g wide, so heights equal to the frequencies kept the areas of the bars in proportion to the frequencies.
The new class \(140 < m \le 180\) is 40 g wide, twice as wide as the others. A bar as tall as its frequency would have twice the area it should, so that class would look bigger than it is.
With unequal class widths the heights must be frequency densities (frequency \(\div\) class width), so that the area of each bar represents its frequency.
Explanation
Frequency can be used for the heights only when every class has the same width, because then the areas of the bars are in proportion to the frequencies. Once one class is wider than the others, the heights must be frequency densities, frequency \(\div\) class width. A common error is to keep frequency on the vertical axis after classes have been joined or split.
Mark scheme
-
1 mark
B1 for the classes no longer all have the same width (the new class is 40 g wide, twice the width of the others), so the heights must be frequency densities (frequency \(\div\) class width) for the areas of the bars to represent the frequencies oe
Accept 'the class widths are now unequal' or 'the new class is wider than the others' alone, or 'the bar for 140 to 180 would have too big an area / would look as if it had too many oranges, because it is twice as wide'. B0 for 'the data is continuous', 'it is not accurate' or 'he must use frequency density' with no reason.