| Exam Board | OCR MEI |
|---|---|
| Module | Further Statistics A AS (Further Statistics A AS) |
| Year | 2018 |
| Session | June |
| Marks | 12 |
| Paper | Download PDF ↗ |
| Mark scheme | Download PDF ↗ |
| Topic | Bivariate data |
| Type | Calculate r from raw bivariate data |
| Difficulty | Standard +0.3 This is a straightforward application of Spearman's rank correlation coefficient with standard hypothesis testing. While it requires multiple steps (ranking data, calculating rs, hypothesis test, interpretation), each step follows a routine procedure taught in Further Statistics. The conceptual demand is low—recognizing non-linearity from a scatter diagram and applying a standard non-parametric test. This is slightly easier than average because it's a textbook application with no novel problem-solving required. |
| Spec | 5.08e Spearman rank correlation5.08f Hypothesis test: Spearman rank |
| \(x\) | 45.9 | 48.3 | 52.2 | 64.6 | 66.6 | 67.6 | 69.3 | 75.0 | 77.4 | 82.8 |
| \(y\) | 25.4 | 23.9 | 26.6 | 18.8 | 18.9 | 19.0 | 16.8 | 16.3 | 17.8 | 17.2 |
| Answer | Marks | Guidance |
|---|---|---|
| 3 | (i) | Because the scatter diagram does not suggest a |
| Answer | Marks |
|---|---|
| have two ‘islands’) | E1 |
| Answer | Marks |
|---|---|
| [2] | 3.5a |
| 3.5b | For not from bivariate Normal distn |
| For not elliptical | Do not accept ‘data is not |
| Answer | Marks |
|---|---|
| (ii) | Rank x 1 2 3 4 5 6 7 8 9 10 |
| Answer | Marks |
|---|---|
| s | M1 |
| Answer | Marks |
|---|---|
| [3] | 1.1a |
| Answer | Marks |
|---|---|
| 1.1 | For using ranks |
| For correct ranks for y | Accept both reversed |
| Answer | Marks |
|---|---|
| (iii) | H : There is no association between level of |
| Answer | Marks |
|---|---|
| radium. | B1 |
| Answer | Marks |
|---|---|
| [5] | 3.3 |
| Answer | Marks |
|---|---|
| 2.2b | B1 for H |
| Answer | Marks | Guidance |
|---|---|---|
| provided | r | < 1 |
| Answer | Marks |
|---|---|
| s | Hypotheses as shown in |
| Answer | Marks |
|---|---|
| (iv) | The significance level is the probability of rejecting |
| Answer | Marks |
|---|---|
| that there is an association between x and y. | E1 |
| Answer | Marks | Guidance |
|---|---|---|
| [2] | 2.4 | |
| 1.2 | For relating this to the context. | |
| Rank x | 1 | 2 |
| Rank y | 9 | 8 |
Question 3:
3 | (i) | Because the scatter diagram does not suggest a
bivariate Normal distribution since it is does not
appear to be roughly elliptical (seems to possibly
have two ‘islands’) | E1
E1
[2] | 3.5a
3.5b | For not from bivariate Normal distn
For not elliptical | Do not accept ‘data is not
bivariate Normal’
Do not accept ‘Normal
bivariate’
(ii) | Rank x 1 2 3 4 5 6 7 8 9 10
Rank y 9 8 10 5 6 7 2 1 4 3
r = 0.818 (= – 9/11)
s | M1
A1
A1
[3] | 1.1a
1.1
1.1 | For using ranks
For correct ranks for y | Accept both reversed
Accept – 0.82, – 0.8182 or
better
(iii) | H : There is no association between level of
0
dissolved oxygen and amount of radium.
H : There is association between level of dissolved
1
oxygen and amount of radium.
For n = 10, 1% critical value = 0.7939
0.818 > 0.7939 so significant/Reject H .
0
The evidence suggests that there is some association
between level of dissolved oxygen and amount of
radium. | B1
B1
B1
M1
A1
[5] | 3.3
1.2
1.1
1.1
2.2b | B1 for H
0
B1 for H and population soi
1
Hypotheses must be in context (allow
hypotheses in terms of x & y)
NB H H NOT ito ρ
0 1
M1 for sensible comparison with
0.7939, leading to a conclusion,
provided |r| < 1
s
For non-assertive correct conclusion
in context and in terms of H . FT their
1
r
s | Hypotheses as shown in
answer column should be
understood to imply
population
No further marks from here
if wrong cv used.
See additional notes.
(iv) | The significance level is the probability of rejecting
the null hypothesis when in fact it is true.
e.g. If there is no association between x and y only
about 1 sample in 100 would lead to the conclusion
that there is an association between x and y. | E1
E1
[2] | 2.4
1.2 | For relating this to the context.
Rank x | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10
Rank y | 9 | 8 | 10 | 5 | 6 | 7 | 2 | 1 | 4 | 3
3 Samples of water are taken from 10 randomly chosen wells in an area of a country. A researcher is investigating whether there is any relationship between the levels of dissolved oxygen, $x$, and the amounts of radium, $y$, in the water from the wells. Both quantities are measured in suitable units. The table and the scatter diagram in Fig. 3 show the values of $x$ and $y$ for the ten wells.
\begin{center}
\begin{tabular}{ | l | l | l | l | l | l | l | l | l | l | l | }
\hline
$x$ & 45.9 & 48.3 & 52.2 & 64.6 & 66.6 & 67.6 & 69.3 & 75.0 & 77.4 & 82.8 \\
\hline
$y$ & 25.4 & 23.9 & 26.6 & 18.8 & 18.9 & 19.0 & 16.8 & 16.3 & 17.8 & 17.2 \\
\hline
\end{tabular}
\end{center}
\begin{figure}[h]
\begin{center}
\includegraphics[alt={},max width=\textwidth]{e3ac0ba0-9692-4018-894e-2b04b07eaf32-3_865_786_657_635}
\captionsetup{labelformat=empty}
\caption{Fig. 3}
\end{center}
\end{figure}
(i) Explain why it may not be appropriate to carry out a hypothesis test based on the product moment correlation coefficient.\\
(ii) Calculate Spearman's rank correlation coefficient for these data.\\
(iii) Using this value of Spearman's rank correlation coefficient, carry out a hypothesis test at the 1\% significance level to investigate whether there is any association between $x$ and $y$.\\
(iv) Explain the meaning of the term 'significance level' in the context of the test carried out in part (iii).
\hfill \mbox{\textit{OCR MEI Further Statistics A AS 2018 Q3 [12]}}