4.2 Presentation of data
Studying this course? Keep your progress in one place.
Key facts
- Classes are written as inequalities with no gaps or overlaps.
- Histograms here have equal class widths; heights are frequencies.
- Median at , quartiles at and , the th percentile at ; interpolate within a class.
- IQR ; outliers lie beyond from the quartiles.
- Compare a centre and a spread, in context. Symmetric box and whiskers suggest a normal model may fit; skewed data do not.
Frequency tables and classes
Use inequalities with no gaps or overlaps, so every value belongs to exactly one class.
A frequency distribution lists each value or class with its frequency: how many data items it contains.
For continuous data, write each class interval as an inequality, for example and . Every possible value belongs to exactly one class: 15 cm goes in .
Grouping loses the individual values, so anything calculated from a grouped table (median, quartiles) is an estimate.
| Correct | Gaps (wrong) | |
|---|---|---|
| First class | 10–15 | |
| Second class | 16–20 |
Correct
- First class:
- Second class:
Gaps (wrong)
- First class:
- 10–15
- Second class:
- 16–20
Which class does a length of exactly 20 cm go into?
Frequency histograms
Touching bars whose heights are the class frequencies show where the data cluster and whether they are skewed.
A frequency histogram with equal class widths has touching bars whose heights are the class frequencies. It shows the shape: where the data cluster, whether it is symmetric or skewed, and any gaps.
Unlike a bar chart for categories, a histogram has no gaps between bars, because the data are continuous.
Read it like a frequency table. The bars below give 4, 9, 15, 12, 7 and 3, so there are 50 items in total. Frequency density histograms with unequal widths are not required.
Using the histogram above, which class contains the median (the 25th value)?
Cumulative frequency and percentiles
Read the value at position , or from a cumulative frequency graph, or interpolate within a class.
The cumulative frequency up to a value is the number of items less than it. It is plotted at the upper class boundary. From it you can estimate the median, quartiles, percentiles, interquartile range () and range.
Without a graph, use linear interpolation. If the position lies items into a class of width containing items, starting at :
- Medianposition
- Quartilespositions and
- pth percentileposition
Worked example
200 apples: 90 have mass below 140 g and 60 are in the class . Estimate the median.
- 1
The median is the th value.
- 2
90 values lie below 140 g, so the 100th is items into the class.
- 3
, and , so the estimate is .
In a cumulative frequency graph of 80 values, at what position do you read the upper quartile?
Box and whisker diagrams
The five-number summary, with outliers marked separately.
A box and whisker diagram displays the five-number summary: minimum, , median, , maximum.
The box runs from to with a line at the median. Outliers are shown as separate points. The whisker stops at the most extreme value that is not an outlier.
You must give the features (quartiles, IQR, outlier boundaries) as well as draw the diagram.
Example: and , so and . The boundaries are and , so 230 is an outlier.
and . What are the outlier boundaries?
Comparing distributions and skew
Compare one centre and one spread in context, and use the median's position in the box to judge skew.
Compare one measure of centre (median) and one measure of spread (IQR or range), and interpret each in context: 'the median wait at B is 40 s shorter, so a typical customer is served faster'. Use the IQR when there are outliers, because the range is distorted.
If the median is near the middle of the box and the whiskers are similar, the data are symmetric and may be normally distributed. A median closer to means positively skewed; closer to means negatively skewed. Skewed data are not well modelled by a normal distribution.
| Symmetric | Positively skewed | Negatively skewed | |
|---|---|---|---|
| Median | Middle of the box | Closer to Q₁ | Closer to Q₃ |
| Tail | Similar on both sides | Longer to the right | Longer to the left |
| Normal model | May fit | Does not fit | Does not fit |
Symmetric
- Median:
- Middle of the box
- Tail:
- Similar on both sides
- Normal model:
- May fit
Positively skewed
- Median:
- Closer to Q₁
- Tail:
- Longer to the right
- Normal model:
- Does not fit
Negatively skewed
- Median:
- Closer to Q₃
- Tail:
- Longer to the left
- Normal model:
- Does not fit
In a box and whisker diagram the median is much closer to Q₁ than to Q₃. What is the skew?
Try an exam question
A data set has , median , and maximum . (a) Find the interquartile range. (b) Find the upper outlier boundary and state whether 70 is an outlier. (c) Describe the skew.
[4 marks]
- [1].
- [1]Upper boundary .
- [1], so 70 is an outlier.
- [1]The median is closer to than , so the distribution is positively skewed.
That's the notes covered.
Carry on to the next subtopic.
Try a question
Which class contains the median time?
1 mark, marked against the mark scheme
Which class contains the median time?
1 mark, marked against the mark scheme