| Exam Board | OCR |
|---|---|
| Module | S1 (Statistics 1) |
| Year | 2006 |
| Session | June |
| Marks | 10 |
| Paper | Download PDF ↗ |
| Mark scheme | Download PDF ↗ |
| Topic | Hypothesis test of Spearman’s rank correlation coefficien |
| Type | Effect of data changes |
| Difficulty | Standard +0.3 This is a straightforward Spearman's rank correlation question requiring ranking data, applying the formula, and interpreting results. Part (i)(c) tests conceptual understanding of how rank changes affect rs, while part (ii) requires basic interpretation of scatter diagrams. All components are standard S1 material with no novel problem-solving required, making it slightly easier than average. |
| Spec | 5.08e Spearman rank correlation5.08f Hypothesis test: Spearman rank5.08g Compare: Pearson vs Spearman |
| Agent | \(A\) | \(B\) | \(C\) | \(D\) | \(E\) | \(F\) | \(G\) |
| Distance travelled | 18 | 15 | 12 | 14 | 16 | 24 | 13 |
| Commission earned | 18 | 45 | 19 | 24 | 27 | 22 | 23 |
| Answer | Marks | Guidance |
|---|---|---|
| 6(i)(a) | Ranks: \(2 \, 4 \, 7 \, 5 \, 3 \, 1 \, 6\) \(6 \, 4 \, 1 \, 3 \, 5 \, 7 \, 2\) \(7 \, 1 \, 6 \, 3 \, 2 \, 5 \, 4\) \(1 \, 7 \, 2 \, 5 \, 6 \, 3 \, 4\) \(\sum d^2\) \((= 60)\) | M1: —; A1: —; M1: — |
| \(r_s = 1 - \frac{6 \times 60}{7 \times 48}\) | M1: — | calc \(r\) for ranks: \(S_x = S_y = 140 - 28^2/7\) \((= 28)\) \(S_xy = 110-28^2/7\) \((= -2)\) corr subst in one corr \(S\) (any version):M1 corr subst in \(r = S_xy / \sqrt{(S_x S_y)}\) :M1 |
| \(= -^1/_{14}\) or \(-0.071\) (3 dps) | A1: 5 | \(-0.07\) without wking: M1A1M2A0 |
| 6(i)(b) | Little (or no) connection (agreement, rel'nship) between dist and commission Allow disagreement | B1f f: 1 |
| 6(i)(c) | Unchanged. No change in rank | B1B1: 2 |
| 6(ii)(a) | \(= -1\) | B1: 1 |
| 6(ii)(b) | Close to \(-1\) or, eg \(\approx -0.9\) | B1: — |
6(i)(a) | Ranks: $2 \, 4 \, 7 \, 5 \, 3 \, 1 \, 6$ $6 \, 4 \, 1 \, 3 \, 5 \, 7 \, 2$ $7 \, 1 \, 6 \, 3 \, 2 \, 5 \, 4$ $1 \, 7 \, 2 \, 5 \, 6 \, 3 \, 4$ $\sum d^2$ $(= 60)$ | M1: —; A1: —; M1: — | $> 5 \text{ ranks correct in each set}$ all correct dep ranks attempted even if opp orders, allow arith errors Correct formula with $n = 7$, dep 2nd M1 |
| $r_s = 1 - \frac{6 \times 60}{7 \times 48}$ | M1: — | calc $r$ for ranks: $S_x = S_y = 140 - 28^2/7$ $(= 28)$ $S_xy = 110-28^2/7$ $(= -2)$ corr subst in one corr $S$ (any version):M1 corr subst in $r = S_xy / \sqrt{(S_x S_y)}$ :M1 |
| $= -^1/_{14}$ or $-0.071$ (3 dps) | A1: 5 | $-0.07$ without wking: M1A1M2A0 |
6(i)(b) | Little (or no) connection (agreement, rel'nship) between dist and commission Allow disagreement | B1f f: 1 | No mks unless $\|r_s\| \leq 1$ ft their $r_s$. Must refer to context. Not "little corr'n between dist and com" not "strong disagreement" Ignore other comment |
6(i)(c) | Unchanged. No change in rank | B1B1: 2 | |
6(ii)(a) | $= -1$ | B1: 1 | indep |
6(ii)(b) | Close to $-1$ or, eg $\approx -0.9$ | B1: — | cao not referring to "corr'n" rather than $r$ allow "neg", not neg corr'n or neg skew |
**Total: 10**
---
6 The table shows the total distance travelled, in thousands of miles, and the amount of commission earned, in thousands of pounds, by each of seven sales agents in 2005.
\begin{center}
\begin{tabular}{ | l | c | c | c | c | c | c | c | }
\hline
Agent & $A$ & $B$ & $C$ & $D$ & $E$ & $F$ & $G$ \\
\hline
Distance travelled & 18 & 15 & 12 & 14 & 16 & 24 & 13 \\
\hline
Commission earned & 18 & 45 & 19 & 24 & 27 & 22 & 23 \\
\hline
\end{tabular}
\end{center}
(i) (a) Calculate Spearman's rank correlation coefficient, $r _ { s }$, for these data.\\
(b) Comment briefly on your value of $r _ { s }$ with reference to this context.\\
(c) After these data were collected, agent $A$ found that he had made a mistake. He had actually travelled 19000 miles in 2005. State, with a reason, but without further calculation, whether the value of Spearman's rank correlation coefficient will increase, decrease or stay the same.
The agents were asked to indicate their level of job satisfaction during 2005. A score of 0 represented no job satisfaction, and a score of 10 represented high job satisfaction. Their scores, $y$, together with the data for distance travelled, $x$, are illustrated in the scatter diagram below.\\
\begin{tikzpicture}[
>=Stealth,
x=0.6cm,
y=0.55cm
]
% Grid lines (light) - drawn first so they remain behind the axes
\foreach \x in {10,11,...,26} {
\draw[gray!20, very thin] (\x, 0) -- (\x, 10.5);
}
\foreach \y in {0,1,...,10} {
\draw[gray!20, very thin] (10, \y) -- (26.5, \y);
}
% Axes
\draw[->, thick] (10, 0) -- (27.5, 0) node[right] {$x$};
\draw[->, thick] (10, 0) -- (10, 11.5) node[above] {$y$};
% X-axis tick marks and labels
\foreach \x in {10, 15, 20, 25} {
\draw[thick] (\x, 0) -- (\x, -0.2) node[below=2pt] {$\x$};
}
% Y-axis tick marks and labels
\foreach \y in {0, 5, 10} {
\draw[thick] (10, \y) -- (9.8, \y) node[left=2pt] {$\y$};
}
% Data points as crosses using standard TikZ plot marks
\draw plot[only marks, mark=x, mark options={thick}, mark size=4pt]
coordinates {
(12,10)
(13,7.5)
(14,6.5)
(15,5)
(16,4)
(21,2.5)
(24,1)
};
\end{tikzpicture}\\
(ii) For this scatter diagram, what can you say about the value of\\
(a) Spearman's rank correlation coefficient,\\
(b) the product moment correlation coefficient?
\hfill \mbox{\textit{OCR S1 2006 Q6 [10]}}