Statistics and the mean
One set of data runs through this whole lesson: the maths marks of 30 students, out of 100. We use it to see what statistics is, how data is written, and where the formula for the mean comes from.
Every box marked Try it can be clicked or changed.
By the end of class you will know why this works.
What is statistics?
Statistics is the branch of mathematics that deals with the collection, organisation, presentation, analysis and interpretation of data.
Data are facts or figures collected for a definite purpose: the marks of a class, the runs scored in a series, the rainfall in each month, the price of tomatoes every week.
Statistics (plural)
The numerical facts themselves.
“Our school’s exam statistics show that 46 of 50 students passed.”
Statistics (singular)
The subject: the methods used to collect, study and draw conclusions from data.
“Statistics is used in every science.”
The word comes from the Latin status and the Italian statista, both connected with the state. Early statistics was information kept by kings and governments about their people, land, and taxes.
The five stages of statistics, using our class’s marks
Raw, ungrouped and grouped data
The same 30 marks can be written in three ways. Switch between them and notice what we keep and what we lose each time.
Ungrouped data
Every individual value is kept. It may be a plain list (raw data), or a table showing each value with its frequency, the number of times it occurs.
We can still say: 4 students scored exactly 40.
Grouped data
Values are put into class intervals. We know how many values fall in each class, but not the exact values. This is used when there are many observations spread over a wide range.
We can only say: 7 students scored between 40 and 55.
Words you need for grouped data
Take the class 40 – 55 from the grouped table above.
- Class interval
- 40 – 55
- Lower limit
- 40
- Upper limit
- 55
- Class size (width)
- 55 − 40 = 15
- Class mark (mid-point)
- (40 + 55) ÷ 2 = 47.5
- Frequency
- 7 students
A mark of exactly 55 is counted in 55 – 70, not in 40 – 55. The upper limit belongs to the next class.
What is the mean?
Mean, median and mode are the three measures of central tendency: single numbers that describe the centre of the data. Today we study the mean.
The mean (arithmetic mean, or average) is the value each observation would have if the total were shared out equally.
Five friends have some chocolates each. If they put them together and share equally, how many does each get?
Mean of ungrouped data
Step A · Writing a long sum with Σ
If there are n observations, call them x1, x2, x3, …, xn. The small number is the position: x3 means “the 3rd observation”.
In xi, the letter i is a counter. It takes the values 1, 2, 3, … one by one. The Greek capital letter Σ (sigma) means “add up all of these”.
The first five students in our list scored these marks.
Step B · When values repeat, Σxi becomes Σfixi
In our 30 marks, 40 appears 4 times. Adding 40 + 40 + 40 + 40 is the same as 4 × 40. So we write each different mark once as xi, and how many times it occurs as its frequency fi.
The number of observations is also the sum of the frequencies: n = 1 + 1 + 3 + 4 + … = Σfi = 30. Put both into the mean formula:
Careful: here i counts the rows of the table (k = 13 different marks), not the 30 students.
| Marks xi | Students fi | fixi |
|---|---|---|
| Total | Σfi = 30 | Σfixi = 1779 |
Using the table:
This is the true average of the class, because every student’s actual mark was used. Remember 59.3; we will compare with it soon.
Mean of grouped data: the direct method
The problem
Suppose we are given only the grouped table, not the 30 marks. We still need Σfixi, the sum of all the marks. But look at one class:
Each of those 7 marks could be anything from 40 up to 54. We do not know them, so we cannot add them exactly.
The assumption: every value sits at the class mark
We assume that the observations in each class are centred around its mid-point. So every value in a class is taken to be equal to its class mark:
xi = lower limit + upper limit2
Since we secretly still have the raw marks, we can check how good this guess is.
So the sum of the values in the i-th class is taken as fi × xi. Now the same steps as before give the formula:
Same formula as ungrouped data. Only the meaning of xᵢ changes: now it is the class mark.
The direct method, step by step
| Class interval | Frequency fi | Class mark xi | fixi |
|---|
Why did we get 62 and not 59.3?
The exact mean from the raw marks is 59.3, but the grouped table gave 62. The difference comes from the class mark assumption. In this data, most students’ marks sit below the middle of their class (four students scored exactly 40, at the very start of 40 – 55). The class mark is a little higher than them, so the estimate comes out a little high.
If we group the same marks in a different way, we get a slightly different estimate each time:
When all we have is a grouped table, we cannot know the actual marks, so the class mark is the best single guess. The mean of grouped data is an approximation, and it is usually close to the true mean.
Your turn
Find the mean by the direct method
Pick a problem or type your own table. Work out the mean in your notebook, enter it, and check. Open the working only after you try.
| Lower limit | Upper limit | Frequency fi | Class mark xi | fixi | Remove |
|---|---|---|---|---|---|
| Total | |||||
Quick check
Answer these with your class.
Summary: one formula, three meanings of xi
| Data given as | Mean | xi means | i counts |
|---|---|---|---|
| A list of values (raw) | = Σxin | the i-th observation | each observation, 1 to n |
| Ungrouped frequency table | = ΣfixiΣfi | a value in the table | each row (each different value) |
| Grouped frequency table | = ΣfixiΣfi | the class mark of the i-th class | each class |
Next class: the assumed mean method and the step-deviation method, which make the same calculation easier with big numbers.