... number of people. Also, random sampling will be better as it will be easier to avoid bias, because with stratified you may end up selecting all the pupils who have improved the most even if you don't really mean to. To select the random people I filtered the data in excel by selecting first the males from every set, and then the females. Then, using the random function on my calculator, I selected 2 boys and 2 girls from each set. I thought this would be a good number because it reflects the whole year group. Altogether I have 52 pupils for my investigation: 2 boys and 2 girls from each set out of the 13 sets. This gives a fair representation of both sexes. I used the same number of boys and girls, because there is roughly around the same number of boys and girls in the year group ...

See this essay or coursework document in full right now