Limitations of sampling are: The calculator gives the same number twice.

##### Sample

My calculator did get the same number twice. Because it could bias my result I disregarded it and used the next random number as my sample.

The graph shows negative correlation. This generally shows older cars are cheaper. Some cars of the same age cost different prices, this can be affected by other hypotheses.

Sub hypotheses:

Some cars used prices are affected by their price when new. We can show this by working out the percent knocked off per year. We can display this in a table and graph.

The graph shows that the higher the taken percentage of original value the newer the car is. It is negative correlation. I was right in my assumption that car prices drop by percentage per year. The graph is nearer perfect correlation than my age and price graph, but it’s still not perfect, this can be affected by other hypotheses such as milage.

Looking at the graph I can see there are a couple of cars that don’t fit into the negative correlation. We can use measures of dispersion to rule out these ‘odd’ cars. Using range and standard deviation we can create a cumulative frequency graph to show the majority range of used car prices. By finding this range we can dismiss extraneous values from the sample.

This graph shows better negative correlation with just one extraneous value.

##### Conclusion

I have proved age affects the price of used cars. This graph shows the newer a car is the higher the percentage of original price. The price goes down by percentage every year.

P = -Sa + 80

M = -S

Y = Mx + C

P = Ma + C

Cherry Robinson 10/79 15/06/2007

## Maths coursework

Strand 2:

Testing hypotheses two: mileage

My second hypothesis is mileage. I need to prove mileage affects the price of used cars.

I need 30 samples again. I am going to use a stratified sampling method.

Limitations of sampling are: Luxury cars might have more mileage capacity than other cars. Diesel engines last longer.

I think my sample will be bias because there are less luxury cars than other cars. The database and samples are too small.

Final sample of 30:

## LUXURY

This is to find out the angle for my comparative pie charts.

## RANDOM

1= 0 - 10

2= 10 - 20

3= 20 - 30

4= 30 - 40

5= 40 – 50

6= 50 - 60

7= 60 - 70

8= 70 - 80

9= Delivery

10= Not Shown

## LUXURY

## RANDOM

This is to show correlation between mileage and the frequency.

Luxury

Median position = 15 + 1 / 2 = 8

40 < m < 50 this is the median interval. This is because the 8th number of frequency falls into this interval.

Range = 50 - 0 = 50

The range is in interval 50 < m < 60.

Mode = 20 < m < 30.

Mean = x = Efx / Ef = 12

The mean is interval 70 < m < 80

Lower quartile = n + 1 / 4 = 15 + 1 / 1 = 16, fourth number. Interval = 20 < m < 30

Upper quartile = 3(n + 1 / 4) = 48, twelfth number. Interval = 70 < m < 80

1:

## Random

Random:

Median position = 15 + 1 / 2 = 8

30 < m < 40 this is the median interval. This is because the 8th number of frequency falls into this interval.

Range = 50 - 0 = 50

The range is in interval 50 < m < 60.

Mode = 20 < m < 30.

Mean = x = Efx / Ef = 12

The mean is interval 60 <m < 70

Lower quartile = n + 1 / 4 = 15 + 1 / 1 = 16, fourth number. Interval = 30 < m < 40

Upper quartile = 3(n + 1 / 4) = 48, twelfth number. Interval = 60 < m < 70

1:

Random:

o square route Ef (x – x .) squared / Ef = 3206

Luxury:

o square route Ef (x – x .) squared / Ef = 1882

Random ois lowly dispersed.

Luxury is highly dispersed.

All this proves that Mileage affects the price of second hand cars.