Statistical data and its forms of presentation
The set of all objects to be studied is the population, and a randomly chosen part of it is a sample; when checking everything is expensive or would destroy the objects, we use a sample survey. The number of objects in the sample is its size n. Non-numerical data such as colours or brands are called qualitative, those measured by numbers quantitative. Writing quantitative data in increasing order gives the variation series; its distinct values xᵢ are the variants, the number of times each occurs nᵢ is the frequency, and ωᵢ = nᵢ/n is the relative frequency, with n₁ + … + nₖ = n and the relative frequencies adding up to 1. Joining the points (xᵢ, nᵢ) of a frequency table by a broken line gives a polygon. When there are many values that rarely repeat, we group them into intervals of equal length and draw a histogram of rectangles whose bases are the intervals and whose heights are the frequencies.
| xᵢ | nᵢ | wᵢ |
|---|---|---|
| 0 | 3 | 0,15 |
| 1 | 8 | 0,40 |
| 2 | 6 | 0,30 |
| 3 | 2 | 0,10 |
| 4 | 1 | 0,05 |
To find out whether pupils like the school canteen’s new menu, 30 of the 600 pupils are asked. If only pupils from a sports club are asked, the sample is biased; if 21 of 30 randomly chosen from the register say “yes”, the relative frequency is 21/30 = 0.7, i.e. 70 %. A random sample reflects the whole population better.
Opinion researchers draw conclusions from a sample, quality inspectors test part of a product to judge the whole batch, and medical researchers choose a group of patients to test a new treatment.
Choose a question (for example how many languages classmates speak), ask 10 people, tally the answers and write the relative frequencies. Think about whether asking only your friends would give a true picture of the whole class.
“Class survey”: ask at least 20 classmates how many hours they sleep per day, group the answers into 1-hour intervals, make a frequency table and draw a histogram. Which interval is the most frequent? Compare with another class.