General advice:
Please solve as much Probability sums as possible since they are miscellaneous in nature and
carries a lot of marks. Probability questions also come embedded within other topics such as
normal distribution, discrete random variable, box plots, histogram etc. so this topic is extremely
important.
Also, try to attempt the probability sums at the end so that you don’t waste too much time in the
beginning and confuse yourself.
Chapter 2,3,4: Histogram, Stem and leaf diagrams, box plots, mean.
Median, mode etc
Jan 2014
1. c) Poor basis for a decision. Brad is wrong or equivalent
Summer 2014
***2. e) No effect on the mean. The median will increase. The standard deviation will decrease.
6. d) No of people in a particular activity = probability of that particular activity* total people
Jan 2015
**2.d) Data is skewed. Data is not continuous
7. c) k is the upper quartile
d) P()= P(W<g) = 20/100
June 2015
8. a) Time is a continuous variable
Jan 16
***1.a)iii)
2. e) 5>d, so 5 is an outlier
***3.f) Gradient of textbooks is greater. So spend more advertising on textbooks.
** 6. c) (776+896)/2 = 836, which is the same as the mean.
***d) The standard deviation will reduce.
Jan 18
1.e)i) No effect on variance of class A since addition does not change the variance
ii) The mean will increase since the total score has increased
iii) The variance of the entire group will increase since the mean of class A is now further away
from the mean of class B
### IAL May 19
Explanation:
Quartile are partition values which cuts the data into equal parts
So in first quartile one of the data is moved left side to its right side so the Q1 moves to the right
side so that the left side will have same percentage on left and right side as before
It doesn't affect Q2 and Q3 cause the amount of data will remain same as before
Formulae for calculating frequency from histogram:
Frequency density = k*frequency/class width
You can use ‘k’ on any one side of the equation. But it must be consistent throughout.
Also, when you are drawing your own histogram, you can pick any value of ‘k’. Use k=1 for
convenience.
A particular sum:
Explanation of part e):
Start from the 1.1 class and move upto 1.5 class.Take the total frequency of the classes in
between. Since 1.6 is halfway from 1.5 to 1.7 class, take half the frequency of that class.
So, (12+11+8/2) / total frequency gives you the probability.
Summer 2016:
4. e) Mean>Median, so positive skew
So start from the 1.1 class and move upto 1.5 class
Now since you need upto 1.6, which is half the way from 1.5 to 1.7,
take half the frequency of that clChapter 5: Probability
Answer:
X=0 because B and C are mutually exclusive.
P(A)=p(A|C) as A and C are independent.You get one equation from this.
Sum of all probabilities inside the circles= 1-.06. You get a second equation from here.
Solve the two equations to get y and z.
#
Explanation for part f):
So there are 2 cases: Tom gets a 2 or a 5.
When Tom gets a 2, check the probability of greater than 2 for both die and pick the one with the highest
probability.
Do the same thing when Tom gets a five
So, Basically probability of Avisha winning the game = ½ * (getting more than 2) + ½* (getting more than
five)
Jan 19, Q1
Use condition for independent events
### There’s another probability sum which includes Sandra and Ting. Be sure to solve it.
Explanation:
So there are two formulae to calculate probability, one by Ting and one by Sandra.
Ting has the idea that the probability changes with each throw. So be careful when using Ting’s
model.
Summer 2014
4. c) Conditional probability using Normal distribution
**5. f)
Chapter 6: Correlation
Jan 15
**3. e) Since bread price increases, but milk price stays the same. Therefore the correlation will
decrease.
5. a) Resting heart rate,h, is being measured. So it is the response variable.
Jan 18
3. R is close to -1, there is strong negative correlation
5. c) linear coding does not have any effect on the correlation coefficient
f) As this is interpolation, the estimates are reliable
Jan 19
3. R is close to -1; there is strong negative correlation
5. c) Linear coding does not have any effect on the correlation coefficient
Summer 2017:
2.c) Q3-Q2 = Q2-Q1; symmetric or no skew
e) Value of r is close to zero, so no correlation
**f) Change in box plot, very important
5. b) r is close to 1, so supports the use of a linear model
**h) s=5 is >2 sd above the mean
**6.e)
Summer 2016
*1.c) As weight increases, the percentage of oil content decreases
**2.
June 2015:
7. d) r would be closer to 0
Chapter 7: Regression
Explanation to part c):
Product moment correlation negative dekhe inverse relationship
So, temperature decreases as height increases
Answer to part d) Same as in c because product moment correlation coefficient is unaffected by
coding.
Chapter 8: Discrete random variable
Summer 2014
● 3. c) F(4) = P(X<=4)
### June 2015
1. b) P(X=2) = F(2) - F(1)
*e) For each additional 1m^2 increase in floor size, the value of the house increase by …
### October 18, Q 5 a)
Explanation to part a):
Cumulative distribution is given. The last number on any cumulative distribution is always equal
to 1. Therefore:
(35/2)k = 1
Solve to find k.
Summer 2016
3. a) (Specifying the name of the distribution: Discrete uniform distribution
Chapter 9: Normal distribution
Mark scheme:
Explanation to part b:
It might be difficult to understand even with the working, hence a few words:
First convert to z-value.
You will find that the two values on either side of z are positive and negative of each other. This
suggests that the normal distribution is symmetric on either side of the two z-values. So I divided
them equally.
Good luck for the exam!
Last updated on: 21st Oct 2019, 1:31 pm
