For my investigation, I am going to see if there is a similarity between the number of hours watched of television per week and their personal IQ for year 10's and 11's in key stage 4.

I will also use a method called 'Stratisfied sampling' and I will use the random numbers from which I obtain, to produce another graph to see if there is once again, any correlation.

Results

As one can see, on the scatter graph for the number of hours watched of television per week and their personal IQ for both the males and females, there was one result that was not typical/right data. Therefore, as I am trying to find typical data I removed this result and re-printed the graph along with a line of best fit and the correct formula. Due to the integer in front of the 'x' being so close to zero (y=0.029x), this identified that there was no slope and almost no correlation at all.

I then split up the data as I explained above in my method and I drew another graph separately for firstly males, and then females. Once again, as one can see from the formula on the graphs, there is no correlation or slope for either the males or females.

The type of graph

Males

Females

Male and Female

Equation from the graphs

y=0.0778x + 98.775

y=-0.0043x + 100.51

y=0.029x + 99.761

As one can see from the table, there is a difference with the slope of the trend line in the graphs. However, there is no or almost no correlation for any of the graphs.

I then wanted to find out some information about the distribution of the males and females in year 10 and 11. Therefore, I worked out the maximum, the minimum, the lower and upper quartiles of the data and the median. The results are as followed:

* Median = 100.5

* 1st Quartile = 95

This is for the graph on the females.

* Minimum Value = 11

* Maximum Value = 126

* 3rd Quartile = 105

* Median = 102

* 1st Quartile = 92

This is for the graph on the males.

* Minimum Value = 14

* Maximum Value = 131

* 3rd Quartile = 106.25

I then used these results to draw box plots. These can be seen later on in the project. As one can see from the box plots, the medians of both the boys and the girls are very close to one another (100.5 for the females and 102 for the males) with the interquartile range also being similar with 10 for the females and 14.25 for the males.

The maximum for the girls is 126 whilst that for the boys is 131. The minimum for both are very similar (11 and 14).

The range is almost the same with 115 for the females and 117 for the males, yet the interquartile range is 10 for the girls and 14.25 for the boys, meaning the boys data is much more closely bunched around the median and the girls are spread out.

I also used stratisfied sampling for males and females in year's 10 and 11. Here is my working out:

Number in year x 20 = ?

Total amount of 10+11

Therefore, for year 10: 200 x 20 = 11

370

Therefore, for year 11: 170 x 20 = 9

370

As a result of this, I will pick 11 from year 10 and 9 from year 11. To find the random numbers for year 10, I did shift Ran# x 200. The random numbers for year 10 are as follows:

) 80 7) 94

2) 44 8) 171

3) 3 9) 144

4) 67 10) 117

5) 140 11) 104

6) 21

To find the random numbers for year 11, I did shift Ran# x 170. The random numbers for year 11 are as follows:

. 165

2. 63

3. 80

4. 150

5. 56

6. 94

7. 123

8. 114

9. 75

I then drew a graph for these random numbers in both years 10 and 11 and compared the equations and lines of best fit. Once again, we see that there is almost no correlation and that there is no similarity between the amount of television watched per week and their IQ.

Conclusion

Throughout the project, I have gathered a lot of information from Mayfield School and used it for my maths coursework. Whilst completing my project, I drew up a lot of graphs and box plots using specific information to try and prove if the number of hours children spent watching television, affected their personal IQ.

Overall, there was absolutely no similarity between these two factors. However, all the equations from the graphs were 'y= 0. ___' but these numbers were not significant enough for there to be any positive or negative correlation.

Alistair Pell

1RMS