The correlation between height and weight.

Authors Avatar

Harriet Walden.

Maths investigation: Mayfield High School.

Mayfield is a fictitious high school but the data presented is based on a real school.

The data is secondary and I am reliant on Edexcel for its validity. There are a total of 31941 datum points, because of such a huge database I am going to select only some points to develop a statistical investigation.

 I have chosen to investigate the correlation between height and weight.

 I will sort the data into what is relevant to my specific investigation and what is not. I will hide the irrelevant data, which is everything except the year group, gender, height and weight.

 I expect the lower years to be shorter and lighter on average than those in the higher years. After grouping the data into year groups I will assign each pupil a number. I will take 30 stratified samples from the whole school. I have chosen to use stratified sampling because each year group is represented proportionally to their occurrence within the whole school. This means that I the results I obtain should show accurately the general trend in height and correlation throughout each year group. Without the sampling being stratified I could end up 30 samples all from the higher end of the school with none of the lower school represented.

I will use the formula:       number of pupils in year             x number of samples                                

                                      Total number of pupils in school

to find the number of samples I need to use for each year.

 I will then use my calculator to produce 30 random numbers. By putting the number of pupils in front of the random button on the calculator, the calculator will only produce numbers that are less the total number you put in. therefor all the numbers produced are relevant to a pupil.

  Any repeated numbers will be discounted.

 I will tabulate the 30 results I get and draw a scatter graph of the heights and weights from the 30 pupils.

 From the scatter graph I expect to see that as the height increases so does the weight.

 I expect on average boys to be taller and heavier and girls so I will then group all the original data into genders so I compare any correlation between girls and boys. I will still use 30 samples and still have each year group represented proportionally.

Join now!

 I will tabulate these results also and draw a scatter graph for each gender to compare the correlations. The easiest way to compare the correlations will be to draw a line of best-fit then work out the gradient. By using the same scale for each scatter graph I will be able to compare the lines of best fit more easily, I expect the line on the graph of the boys to be steeper as they are generally taller and heavier than girls.

 By putting the data in a tally chart I can calculate the cumulative frequency and therefor draw a ...

This is a preview of the whole essay