What factors effect the Price depreciation of a car and how?

As a common teenager I am reaching the legal driving age and will soon require a car. As for most teenagers my age a new car is expensive so a second hand car would by more appropriate. For this project I will be looking at cars from two data sources:

- The Primary Source. I will be doing my own research by using an “Auto Freeway“ magazine for second hand cars and a “What Car?” magazine to find the prices of the cars when they are new. (Information Provided: Car Make, Car Model, Price When New. Price Second Hand, Age, Colour, Engine Size, Mileage)

- The Secondary Source. I will be provided with this data. (Information Provided: Car Make, Car Model, Price When New. Price Second Hand, Age, Colour, Engine Size, Mileage, MPG, Fuel Type, Service History, MOT, Tax, Owners, Insurance Group, Doors, Seats, Style, Gearbox, Central Locking, Air Conditioning, Airbags)

Since the magazine is not a totally reliable source I could not research as much information for the primary source as the secondary source. After I have 100 results of each source I will use systematic sampling by selecting every second car to cut my results into half from both sources. I will compare both sources. With my results I will calculate the Price Depreciation (%) showing by how much percent the car has decreased from the (Price When New).

I suspect that there will be several trends and patterns in the results. The lack of information in the primary source means that I can only make 3 successful hypotheses:

- The Older the car the higher the Price Depreciation
- The Higher the Mileage the higher the Price Depreciation
- The Engine Size would have a very little correlation if any, but if it does the correlation would be, the Bigger the Engine Size the Lower the Price Depreciation

To get my primary data I will be using a random sampling. To do this I will use the random button on my calculator. I will pres it twice: first time I will take the last two digits for the page number and then I will press it again and use the last 2 digits for the car number in the page. If there is no such page or car then I will repeat the procedure. I am using Random Sampling because since I don’t get told what order the cars are in. Due to this Cluster Sampling cannot be used because the group selected could be only for one make of cars. Convenience Sampling could have been used but is not as fair as other methods of sampling.

I occurred many different problems with my sampling:

- Some of the pages did not contain any cars.
- Most cars gave little detail. For this reason I could only make 3 hypotheses
- Some cars gave too little detail and were not included.

## Spearman’s Rank Correlation Coefficient

The spearman’s rank correlation coefficient will show me how well the data fits the trend. The closer it is to 1 or –1 the higher the correlation.

Using this data I can work out the Spearman’s rank correlation coefficient, to do this I will first need to rank all my results. Then work out the difference squared then use the formula:

rs = 1 – 6∑d²

n(n²-1)

I will use the results from the Age/Price Depreciation table from my primary results for this example:

rs = 1 – 6x3697

n(n²-1)