Statistics is concerned with the collection, presentation and analysis of information and data.
Once data has been collected, it can be shown in tables and graphs.
Care must be taken to ensure that each graph has a title, labelled axes appropriate scales units indicated
Pictograph or Pictogram
The frequency table below shows the estimated percentages of the different ethnic groups in New Zealand in the late 20th century.
|
Approximate Ethnic Composition of NZ |
Asian
|
|||
|
% of population |
5
|
Other ethnic groups made up less than 1% of the population
|
Pictographs represent the information with pictures. Ethnic Composition of New Zealand Scale: Each picture represents 5% |
 |
Mãori |
 |
European | |
 |
Pacific Islands | |
 |
Asian |
Sometimes the pictures will not be whole. e.g. In the graph above 7.5% would be represented by one and a half pictures
Column or Bar graph
Ungrouped data is usually small amounts of data where individual values are listed. e.g. 23, 45, 67, 89
If many of these values are the same the data can be place in a frequency table. These tables can be horizontal or vertical.
e.g. Test results (out of 10) of twenty people: 3, 5, 8, 6, 3, 7, 7, 8, 5, 6, 4, 3, 6, 8, 9, 1, 4, 6, 2, 10
This data, called a frequency distribution, can then be shown in a column or bar graph.
Pie graph
This is sometimes called a pie chart or sector graph.
A circle is divided into sectors. The angle of each sector represents the fraction each event is out of the total number of events.
Pie graphs often require calculations to work out the angle size for each sector and the use of protractors.
The frequency table below shows the estimated percentages of the different ethnic groups in New Zealand in the late 20th century.
|
Approximate Ethnic Composition of NZ |
Asian
|
|||
|
% of population |
5
|
Other ethnic groups made up less than 1% of the population
|
Maori needs an angle of 10% of 360 = 36°European needs an angle of 80% of 360 = 288° Pacific Islands needs an angle of 5% of 360 = 18° Asian needs an angle of 5% 0f 36 = 18° |
 |
Grouped Discrete Data
Grouped data is used when each value is likely to be different and values are placed in groups or classes.
e.g.Test results (%) of 30 people
This data is best shown in a bar graph.
Grouped Continuous Data
A histogram is a graph that is used to show the information from a frequency distribution with grouped continuous data.
It is similar to a column graph and the area of each column is proportional to the frequency of the score that it represents.
e.g Ages of workers in a company.
|
Note Because these are ages, |
|
IMPORTANT If any of the intervals in a frequency distribution are of a different size, the height of the bar above it, must be changed accordingly.
e.g. If the bottom interval in the table above was 55 − 74, twice as wide as the others, then the bar would be halved in height to 5.
A frequency polygon is created when the midpoints of the bars of a histogram are joined.
Cumulative Frequency Graph
A useful type of graph for finding, median, quartiles and percentiles is the cumulative frequency graph.
The grouped frequency table shows the times, in minutes, of the first 100 finishers in a 42 km marathon race.
|
Time (minutes) |
Frequency (number of runners) |
The frequency table on the left can produce the cumulative frequency table on the right.
Each term is the sum of all of the frequencies before it. In this case it tells how many runners finished in less than the time given. |
Time (minutes) |
Cumulative Frequency number < x |
|
120 -
|
6
|
120 |
0 | |
|
130 -
|
14
|
130
|
6 | |
|
140 -
|
20
|
140
|
6 + 14 = 20 | |
|
150 -
|
50
|
150
|
20 + 20 = 40 | |
|
160 -
|
64
|
160
|
40 + 50 = 90 | |
|
170 − 180
|
46
|
170
|
90 + 64 = 154 | |
|
200
|
180
|
154 + 46 = 200 |
From the cumulative frequency table a cumulative frequency graph can be drawn. Each point is joined with a straight line.
From the cumulative frequency graph, the percentiles can be found.
Median (50th percentile). As there are 200 runners, the median will be the 200 x 0.5 = 100th runner.
Draw a horizonal line from 100. This gives a median of 162 minutes.
Upper quartile (75th percentile). As there are 200 runners, the upper quartile will be the 200 x 0.75 = 150th runner.
Draw a horizonal line from 150. This gives a median of 169 minutes.
Lower quartile (25th percentile). As there are 200 runners, the lower quartile will be the 200 x 0.25 = 50th runner.
Draw a horizonal line from 50. This gives a median of 152 minutes.
The xth percentile As there are 200 runners, to find the xth percentile, multiply 200 by x and divide by 100, this will give where to draw the horizonal line.
Remember, if you find the 90th percentile, 90% of the data lies below this value.
Misleading Statistical Graphs
Graphs that are titled or labelled incorrectly or have uneven scales or no scales at all can be misleading. This can be done deliberately or by accident.
e.g. A company making breakfast cereals BreakHealth Company claims to be growing really quickly and uses this graph to show the growth in sales over the past 5 years.
This graph appears to show rapid growth because of its steepness but its slope is greater because of the fact that the sales scale starts at 5 million.
