AQA Large Data Set
What is the Large Data Set?
The Large Data Set in Exams
Password Protected File
Worksheets are password protected to prevent accidental damage.
Go to Home on the Excel ribbon, then Format Unprotect Sheet Password : LDS
Download here: AQA Large Data Set
Section 1 - Exploring the Large Data Set
Introduction
The study of real data in A-level Mathematics is essential in providing students with a view of the
relevance of statistical techniques.
The basic Vehicles LDS is available in the here.
Activity 1 – Getting to Know the Large Data Set
Activity 2 – Formulating Hypotheses
Activity 3 – Investigating Hypotheses
Activity 1 – Getting to Know the LDS
1. Read the information on the Introduction, Definition of field and Field values sheets so that you
know what data the LDS contains.
2. The following slides contain examples to help you to become familiar with the LDS.
3. The student activity, Activity 1, should last about an hour.
4. Answers are provided.
5. Little knowledge of Excel is required.
Example 1.1
Question Answer
Read the information on the worksheet called 1. PETROL
Field Values Sheets.
Answer the following questions: 2. PropulsionTypeId Code 8
1. Define PropulsionTypeId Code 1.
3. Coupe
2. What is the field value for Electric/Petrol?
3. Define BodyTypeId Code 5. 4. BodyTypeId Code 96
4. What is the field value for Multipurpose
Vehicle? 5. Company
5. Define KeeperTitleId Code 5.
6. KeeperTitleId Code 1
6. What is the field value for Male?
Example 1.2
Question Answer
1. How many diesel cars are there? 1. 1396
• Go to worksheet ‘Car Data’
• Go to Column C ‘PropulsionTypeId’
• Select ‘2’ from the filter arrow
• Highlight the column
• Look at the bottom right hand corner of excel
and read ‘counts’
2. What is the highest level of CO2 emissions? 2. 356 g/km
• Go to worksheet ‘car data’
• Go to column J ‘CO2’
• Filter to top 1 item
3. In North West London, how many Vauxhall’s 3. 17 cars
have an NOX emission of 0.013g/km? • Column E ‘GovRegion’: Filter to North West
• Column B ‘Make’: Filter to Vauxhall
• Column L ‘NOX’: Filter to 0.013
• Highlight column and ‘count’
Activity 1
Answers
***
Assumptions about Holiday Entitlement in the UK
Work weeks per year ~ 48 weeks
Holidays per year ~ 28 days
Activity 2 – Formulating Hypotheses
1. Setting up hypotheses in this way is not something that is assessed, but you do have to interpret the
results of statistical calculations and graphs in context.
2. Introducing the need to formulate hypotheses and questions carefully (as examiners do) is a way of
developing the high-level maths-specific literacy skills you need when answering many exam
questions. This will also ensure you are familiar with the context of the LDS.
3. You are expected to refer to prior knowledge of techniques from GCSE Maths or GCSE Statistics.
Example 2
Question Answer
Here is a poorly formulated hypothesis Better hypotheses:
“cars have lower emissions than they used to” - The mean CO2 emissions of cars registered in
2016 is lower than the mean CO2 emissions of
cars registered in 2002.
How can this be phrased in a more precise way so
that it could be investigated using the LDS?
- The median NOX emissions of cars registered in
HINT Improve the hypothesis by considering: 2002 is greater than the median NOX emissions
- What cars are being discussed? What is the of cars registered in 2016.
Make? What is the BodyType?
- What fuels are being discussed? E.g. does petrol
have lower emissions that diesel? - The mean CO emissions of Fords registered to
- What emissions are being discussed? Are you companies in 2016 is less than the mean CO
investigating CO, CO2, NOX? emissions of Fords registered to companies in
- What is the gender of the registered keeper? 2002.
- What summary statistics are you investigating?
E.g. mean, median, mode, maximum, minimum,
upper quartile, lower quartile, interquartile
range, standard deviation?
Activity 2
Question Answer
Here are some poorly formulated hypotheses . The mean engine size of cars with male registered
Suggest how they might be better formulated. You keepers is greater than the mean engine size of cars with
are not expected to investigate the hypotheses. female registered keepers.
1. Men drive bigger cars than women. Solution
The mean mass of cars 1
with male registered keepers is
greater than the mean mass of cars with female
registered keepers.
2. Company cars are bigger polluters than other The mean CO2 emissions of cars registered in 2016 to a
cars, because companies don’t really care company are higher than the mean CO2 emissions of cars
about the environment. registered to male or female keepers in 2016.
Solution 2
The mean CO emissions of Fords registered in 2016 is
3. Fords are the best cars for emissions. lower than the mean CO emissions of each of the other 4
makes in the LDS for cars registered in 2016.
The mean emissions of Fords for CO2, CO and NOX are
lower than the corresponding means for the other makes
in the LDS.
Solution 3
Activity 3 – Investigating Hypotheses
1. This is a continuation of activity two.
2. We will take a hypothesis and investigate it using the LDS on Excel.
This activity will train students to use precise language to reach correct conclusions – echoing the
words used in the hypothesis.
Students are not expected to carry out a formal hypothesis test.
Example 3
Question Answer
Investigate the hypothesis below: - Using the filter arrows select the cars registered in
2002. Select column J. The mean CO2 emissions is
‘The mean CO2 emissions of cars registered in displayed in the status bar at the bottom of the
2016 is lower than the mean CO2 emissions of sheet as the ‘Average’. Repeat this for cars
cars registered in 2002’ registered in 2016.
- For 2002 the mean is approx. 171g/km.
- For 2016 the mean is approx. 120g/km.
- Make a comment to interpret this result.
- You may initially make imprecise comments such
as: It’s lower; The average has gone down.
- Make more precise statements such as:
The mean CO2 emissions for cars registered in
2016 is 120 g/km, which is much lower than the
mean CO2 emissions for cars registered in 2002
which is 171 g/km.
Interpret the following hypotheses
Activity 3
1. The median NOX emissions of cars registered in 2002 is greater than the median NOX emissions of
cars registered in 2016.
Column H ‘Year Registered’ Select 2002/2016 Highlight Column L ‘NOX’
Median NOX in 2002 = 0.123
Median NOX in 2016 = 0.033
Solution 1
2. The mean CO emissions of Fords registered to companies in 2016 is less than the mean CO
emissions of Fords registered to companies in 2002.
Column B ‘Ford’ Column F ‘Keeper Title ID 5’ for ‘Companies’ Column H ‘Year’ Select 2002/2016 Highlight Column K ‘CO’
Read ‘average
Mean CO 2016 = 0.335
Solution 2
Solution 3
Solution 4
5. The mean CO2 emissions of cars registered in 2016 to a company are higher than the mean CO2
emissions of cars registered to male or female keepers in 2016.
Column H ‘Year Registered’ Select 2016 Column F ‘Keeper Title ID 5’ for ‘Company’
Column J ‘CO2’ Highlight and read ‘average’ Mean = 117g/km
Solution 5
Column H ‘Year Registered’ Select 2016 Column F ‘Keeper Title ID’ Select ‘ID 1 and 2’ for ‘Male or Female’ Column J ‘CO2’
Highlight and read ‘average’ Mean = 124g/km
6. The mean CO emissions of Fords registered in 2016 is lower than the mean CO emissions of each of
the other 4 makes in the LDS for cars registered in 2016.
Column B ‘Make’ Ford Column H ‘Year Registered’ Select 2016 Column K ‘CO’ Highlight and read ‘average’
Mean = 0.35
Column B ‘Make’ Select all other makes except Ford Column H ‘Year Registered’ Select 2016 Column K ‘CO’ Highlight and
read ‘average’ Mean = 0.30 Solution 6
7. The mean emissions of Fords for CO2, CO and NOX are lower than the corresponding means for the
other makes in the LDS
Column B ‘Make’ Select Ford Highlight Columns for CO2, CO and NOX Read average Mean = 46.9
Column B ‘Make’ Select all other makes except Ford Highlight Columns for CO2, CO and NOX Read average
Mean = 44.3
Solution 7
Section 2 - Sampling
How to use Sampling Techniques
The LDS as a Sample
Activity 1
Q1
a. This will depend on your chosen sample.
Go to ‘Sampling’ Tab Propulsion Type ID 2
Sample Size 50 Sample Field ‘part’ Click on
‘select sample’
All cars have data on their particulate emissions.
Solution 1
b. Median Particulate Emissions:
Go to ‘Sampling Statistics’ tab Median = 0
Q2
a. Answer will depend on your chosen sample.
Solution 2a
b. Answer will depend on your chosen sample.
Sample Population
100 Petrol Powered Ford’s All Petrol Powered Ford’s
unbiased estimator for population variance
AS Book Key Point 22.1
Solution 2b
from Car Data tab
Q3
Solution 3
Q4
Solution 4
Q5
Solution 5a
Solution 5b
Q6
Solution 6
Section 3 - Data Representation and Interpretation
Q1
Solution 1a
Solution 1b
Q2
Solution 2
Q3
Solution 3a
Solution 3b
Q4
Solution 4
Section 4 - Probability
Example M1.1
Example M1.2
number of diesel cars
Solution
P(
Example M2.1
Make
Make Frequency Percentage
BMW 551 14%
FORD 1089 28%
TOYOTA 275 7%
VAUXHALL 1069 28%
VOLKSWAGEN 843 22%
BMW FORD TOYOTA
VAUXHALL VOLKSWAGEN
PropulsionTypeId
PropulsionTypeId
Code Description Frequency Percentage
1 Petrol 2311 60%
2 Diesel 1396 36%
3 Electric 1 0%
7 Gas/Petrol 1 0%
8 Electric/Petrol 118 3%
Petrol Diesel Electric
Gas/Petrol Electric/Petrol
KeeperTitleId
KeeperTitleId
Code Description Frequency Percentage
1 Male 1481 39%
2 Female 878 23%
3 (Not used) 0 0%
4 Unknown (Rev, Dr etc.) 117 3%
5 Company 1351 35%
Male Female
(Not used) Unknown (Rev, Dr etc.)
Company
BodyTypeId
BodyTypeId
Code Description Frequency Percentage
1 2 door saloon 2 0%
2 4 door saloon 225 6%
3 Saloon 27 1%
4 Convertible 60 2%
5 Coupe 141 4%
6 Estate 454 12%
13 3 Door Hatchback 475 12%
14 5 Door Hatchback 2091 55%
2 door saloon 4 door saloon Saloon
96 Multi Purpose Vehicle 352 9% Convertible
3 Door Hatchback
Coupe
5 Door Hatchback
Estate
Multi Purpose Vehicle
Year Registered
Region Frequency Percentage Percent
Year Frequency age
London 998 26%
2002 1299 34%
North West 1164 30% 2016 2528 66%
South West 1665 44%
Region Year Registered
London North West South West 2002 2016
Example M3.1
Solution
Example M3.2
Solution
Q1
Total
1298
2528
Total 998 1164 1664 3826
1183
Solution 1a
3826
450
Solution 1b
3826
367
Solution 1c
1298
Q2
Total
2311
1396
120
Total 2566 352 909 3827
118 697
Solution
3827 2a Solution
1396 2b
2311 ( ) 352 2311 352
P ( A)= P B= P ( A) × P ( B)= × =0.5554 . .
3827 3827 3827 3827
Solution 2c
118 NOT INDEPENDENT because
P ( A ∩ B)= =0.0308 . .
3827
Section 5 - Statistical Distributions
Q1
Solution 1
Q2
Calculator
Menu #7
Binomial CD
Variable
Solution 2
Q3
Solution 3
Q1
Solution 1
Q2
Solution 2
Q3
Solution 3
Q4
Solution 4
Q5
Solution 5
Q6
Solution 6
Section 6 - Hypothesis Testing
Q1
Solution 1
Q2
Solution 2
Q3
Solution 3
Q4
Solution 4
Q5
Solution 5
Read this document on the ‘Summer of LDS’
Quizlet
You may want to create something similar for yourself.
Exam Questions
9 questions | 47marks | 55minutes