Interpreting Data

Discrete vs. Continuous

STOP AND CHECK (PAGE 6)

● Examples of continuous data could be height, weight, time, or other

measurements. For example, the height of students or the time taken to run
100 metres.
● The Poisson and Binomial distributions use discrete data.

Expectation and Variance

STOP AND CHECK (PAGE 10)

● The expected value is the average (mean) value you’d get if you ran a lot of
trials of an experiment and recorded all the answers.
● Variance measures how spread out all the values are in a data set. A large
variance means all the values are very different.
● The standard deviation is the square root of the variance, so it also measures
how spread out the data is.

Interpreting Data
QUICK QUESTIONS (PAGE 15)

● E(X) = ∑x ✕ P(X = x)
𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦
● The probabilities are given by 𝑡𝑜𝑡𝑎𝑙
, so we can add those to the table:

1 Level 3 Statistics | Probability Concepts | © Inspiration Education Limited 2021. All rights reserved.
 Number of Frequency Proportion
Coffees

1 5 0.1

2 6 0.12

3 14 0.28

4 3 0.06

5 20 0.4

6 2 0.04

E(X) = (1 x 0.1) + (2 x 0.12) + (3 x 0.28) + (4 x 0.06) + (5 x 0.4) + (6 x 0.04)

E(X) = 3.66
The expected value of the number of coffees is 3.66.

● Var(X) = E(X2) – [E(X)]2

E(X2) = 12 x 0.1 + 22 x 0.12 + 32 x 0.28 + 42 x 0.06 + 52 x 0.4 + 62 x 0.04
E(X2) = 15.5

Var(X) = 15.5 – 3.662

Var(X) = 2.1044
The variance of the number of coffees is 2.1044.

● σ = (Var(X))2
σ = 2. 1044
σ = 1.451 (3 d.p.)
The standard deviation of the number of coffees is 1.451.

● Table:

Outcome Red White Black

Probability 4
8
3
8
1
8

2 Level 3 Statistics | Probability Concepts | © Inspiration Education Limited 2021. All rights reserved.
 Profit $1 (4-3=1) -$3 (they lose $2 (5-3 = 2)
the $3 they
paid)

To get the expected profit we multiply each profit by its probability and then
add those together:

4 3 1
(1 x 8
) + (-3 x 8
) + (2 x 8
) = -0.375

Since this number is negative, it represents a loss rather than a profit, so the
player’s expected loss is $0.375.

● Var(X) = E(X2) – [E(X)]2

E(X2) = 12 x + (-3)2 x + 22 x
4 3 1
8 8 8
= 4.375

(Remember to square -3 in brackets so that it comes out positive!)

Var(X) = 4.375 – 0.3752

Var(X) = 4.234 (3 d.p.)
The variance of the player’s profit or loss is 4.234.

● E[4X – Y] = 4 x E[X] + (– 1) x E[Y], using the rule:

E[aX + bY] = aE[X] + bE[Y]

Here a = 4 and b = -1
E[4X – Y] = 4 x E[X] – E[Y]
E[4X – Y] = 4 x 6.7 – 5.2
E[4X - Y] = 21.6
The expected value is 21.6.

● Var(4X – Y) = 42 x Var(X) + (-1)2 x Var(Y), using the rule:

Var(aX +bY) = a2 Var(X) + b2 Var(Y)

Var(4X - Y) = 42 x 2 + (-1)2 x 3.5

3 Level 3 Statistics | Probability Concepts | © Inspiration Education Limited 2021. All rights reserved.
 Var(4X - Y) = 35.5
The variance is 35.5.

● σ = (Var(4X - Y))2
σ = 35. 5
σ = 5.96 (2 d.p).
The standard deviation is 5.96.

The Distributions

Throwback to the Normal Distribution

STOP AND CHECK (PAGE 19)

● We standardise so that we can use the same formula to describe any normal
distribution, even if they have really different means and standard deviations.
● The standard normal curve has a mean of 0 and a standard deviation of 1.
● The four features of the normal distribution are:
○ It measures a continuous variable with no upper or lower limits on what
value it can take.
○ It has a mean which is also the most common value (or peak).
○ It has a bell-curve, symmetrical shape.
○ Values far from the mean are very unlikely.

The Binomial Distribution

STOP AND CHECK (PAGE 21)

● To use the Binomial distribution, we need to have:

○ A discrete variable.
○ Two outcomes for each trial: success or failure.
○ A fixed number of trials.
○ Independent trials (i.e., they don’t affect each other).
○ A constant probability of success.

4 Level 3 Statistics | Probability Concepts | © Inspiration Education Limited 2021. All rights reserved.
The Poisson Distribution
STOP AND CHECK (PAGE 22)

● To use the Poisson distribution, you need these things to be true:

○ Events happen randomly.
○ Events happen independently (don’t affect each other).
○ Events can’t happen at the same time or in the exact same place
○ Events happen at a constant rate, that is, the smaller the interval, the
less events are likely to occur.

The Uniform Distribution

STOP AND CHECK (PAGE 24)

● You can tell a question is a uniform distribution question when it has:

○ A continuous variable.
○ Maximum and minimum values for the variable.
○ All values are equally likely.
● The formula for the uniform distribution is the area of a rectangle formula,
base ✕ height. To find the height, we use:

1
Height = 𝑚𝑎𝑥𝑖𝑚𝑢𝑚 − 𝑚𝑖𝑛𝑖𝑚𝑢𝑚

The Triangular Distribution

STOP AND CHECK (PAGE 27)

● To use the triangular distribution, we need:

○ A continuous variable.
○ Not all values are equally likely (unlike the uniform distribution).
○ A most likely value.
○ Maximum and minimum values for the variable.
● The mode is the most likely value to occur, and it is where the peak of the
triangle will be.

5 Level 3 Statistics | Probability Concepts | © Inspiration Education Limited 2021. All rights reserved.
The Distributions
QUICK QUESTIONS (PAGE 27)

● This is a normal distribution because we have continuous data with no upper

or lower bounds. We also know this because we’ve been given a mean and
standard deviation. The parameters for this distribution are:
○ µ = 320
○ σ = 30

● This is a uniform distribution because we have continuous data with upper

and lower limits but we don’t have a most likely value. Its parameters are:
○ Minimum = 8
○ Maximum = 20

● This is a triangular distribution because this time we have continuous data

with upper and lower limits and we have a most likely value. The parameters
are:
○ Minimum = 2
○ Maximum = 18
○ Mode = 6

● This is a binomial distribution because we have discrete data, with a set

number of trials (staff members who are given cakes), independent trials
(random decision), two outcomes (over 50 or under 50), and a fixed
probability of success (being over 50). The parameters for this distribution are:
○ p = 10
○ p = 0.45 or 45%

● This is a Poisson distribution because we have discrete data, with a set rate of
receiving texts, and we can assume the texts are random and independent.
The parameter is:
○ λ = 4 (per hour)

6 Level 3 Statistics | Probability Concepts | © Inspiration Education Limited 2021. All rights reserved.
Solving Distribution Problems

Solving the Normal Distribution

STOP AND CHECK (PAGE 31)

● To find the probability that x is less than a value, go into Ncd in your calculator,
and input:
○ Lower as -999999999
○ Upper as the value you’re looking at.
○ The mean (µ) and standard deviation (σ) you’ve been given.
● If you need to find two separate probabilities, do exactly what you did above,
but set the upper as +99999999 when you’re finding the ‘x is more than a
value’ part. Then just add the two answers together!

Continuity Corrections
STOP AND CHECK (PAGE 32)

● You use a continuity correction when the data has been rounded.

Binomial by Calculator
STOP AND CHECK (PAGE 34)

● Bpd is what you use when you’re trying to find the probability that x equals a
value, and Bcd is what you use when you’re finding the probability that x is less
than or equal to a value.

7 Level 3 Statistics | Probability Concepts | © Inspiration Education Limited 2021. All rights reserved.
Poisson by Calculator
STOP AND CHECK (PAGE 35)

● Exactly like in the Binomial, Ppd is what you use when you’re trying to find the
probability that x equals a value, and Pcd is what you use when you’re finding
the probability that x is less than or equal to a value.

Solving Distribution Problems

QUICK QUESTIONS (PAGE 40)

● This is a Poisson distribution (as it is discrete data without trials and with a
‘rate’ style parameter), where λ = 2.
We want to find P(X = 0), so we go to Ppd in the calculator and set x = 0 and µ
= 2.
The answer comes out as P(X = 0) = 0.1353, so there is only a 0.1353 chance
that a page will not have to be reprinted.

● This is a Binomial distribution (since it’s discrete data with a set number of
trials, 3 jobs). We want to find P(X ≥ 2), so we need to find 1 – P(X≤2), since the
calculator can only do the calculation this way around.
In Bcd, we input:
○ x=2
○ Numtrial = 3
○ p = 0.05
P(X ≤ 2) = 0.999875
1- P(X≤2) = 0.000125
So the probability that Bobby has to write a report is 0.0001 (to 4d.p.).

● This is a uniform distribution because it is continuous data with maximum and

minimum values. The shaded area represents the probability we are looking
for:

8 Level 3 Statistics | Probability Concepts | © Inspiration Education Limited 2021. All rights reserved.
 P = base x height (area of a rectangle formula)
1
Height = 𝑚𝑎𝑥𝑖𝑚𝑢𝑚 − 𝑚𝑖𝑛𝑖𝑚𝑢𝑚
1
Height = 50−0

Height = 0.02

P(X > 20) = (50 - 20) x 0.02

P(X > 20)= 0.6
The probability that he will be at least 20cm off is 0.6.

● This is a normal distribution because it is continuous data with a mean and no

upper or lower limits.
We need to first find P(X > 1500), so we go to Ncd and set:
○ Lower = 1500
○ Upper = 9999999
○ µ = 1000
○ σ = 200
We get P(X > 1500) =0.0062

To find how many out of 20 booths would fall, we multiply this probability by
20:
○ 0.0062 x 20 = 0.124, which we’ll round to 0 booths.
So out of 20 booths, we wouldn’t really expect any to fall.

● This is a triangular distribution because it uses continuous data with an upper

limit, lower limit and most likely value.

9 Level 3 Statistics | Probability Concepts | © Inspiration Education Limited 2021. All rights reserved.
 The area we’re looking for to find the probability loading the webpage will take
less than 3 seconds is:

To find the area, we first need to find the area of the unshaded triangle. Our x
value is between the mode and the maximum, so we use the formula:

2(𝑏−𝑥)
Height = (𝑏−𝑎)(𝑏−𝑐)

Our values are:

○ a=1
○ b=6
○ c=2
○ x=3
2(6−3)
Height = (6−1)(6−2)

Height = 0.3

Then, we find the probability by the area of a triangle formula:

1
p= 2
x base x height

1
p= 2
x (6-3) x 0.3
p = 0.45

Finally, we take this away from 1 to find the shaded area:

1- 0.45 = 0.55
The probability the webpage loads in less than three seconds is 0.55.

10 Level 3 Statistics | Probability Concepts | © Inspiration Education Limited 2021. All rights reserved.
Inverse Distributions

Inverse Normal
STOP AND CHECK (PAGE 43)

● To solve these questions, you need to:

○ Go to InvN on your calculator.
○ Put in the correct tail side and the area (which is your probability).
○ Set µ = 0 and σ =1, for the standard normal distribution
○ Solve to find Xinv, which will be your Z value
○ Substitute all the values you know, including the Z value, into this
equation:
𝑥−µ
Z= σ

Rearrange to find the missing value!

● You’ll know a question uses the inverse normal because you’ll be given a
probability, which corresponds to an area on the graph!

Inverse Binomial
STOP AND CHECK (PAGE 45)

● To find P(X = 0), just calculate 1 – P(X ≥ 1)

Inverse Poisson
STOP AND CHECK (PAGE 47)

● You’ll know a question uses the inverse Poisson if you’re told a probability in
the question and need to find the parameter.

11 Level 3 Statistics | Probability Concepts | © Inspiration Education Limited 2021. All rights reserved.
Inverse Distributions
QUICK QUESTIONS (PAGE 47)

● We can tell that this is an inverse Poisson question because:

○ It uses discrete data.
○ It involves a time period and random events during that time period.
○ We’re told a probability and need to find the parameter.
Remember that our golden rule to solving these is to use the probability that x
= 0, which we need to find by 1-P(X ≥ 1)

P(X = 0) = 1 – 0.92
P(X = 0) = 0.08

Next, we’ll substitute that into the formula:

−λ 𝑥
𝑒 λ
P(X=x) = 𝑥!
−λ 0
𝑒 λ
0.08 = 0!

This collapses down to:

0.08 = e-λ

Logging both sides:

ln0.08 = -λ
-2.526 = -λ
λ = 2.526

But remember this is for three months, so the last thing we need to do is divide
it by 3 to find the mean for one month.

2.526
λ= 3

λ = 0.8419 seals caught per month on average.

12 Level 3 Statistics | Probability Concepts | © Inspiration Education Limited 2021. All rights reserved.
 ● This is also an inverse Poisson question, so we need to follow the same steps
as for the last question: first, we find P(X = 0) by subtracting P(X ≥ 1) from 1.

P(X = 0) = 1-0.27
P(X = 0) =0.73

Using the formula, we get:

−λ 𝑥
𝑒 λ
P(X=x) = 𝑥!
−λ 0
𝑒 λ
0.72 = 0!

0.72 = e-λ
ln0.72 = -λ
-0.3285 = -λ
λ= 0.3285

Finally, to find the mean for 50 square metres, we multiply this mean by 50:

λ= 0.3285 x 50
λ=16.425

The mean number of chocolate coins in an area of 50m2 is 16.425.

13 Level 3 Statistics | Probability Concepts | © Inspiration Education Limited 2021. All rights reserved.