| Exam Board | OCR MEI |
|---|---|
| Module | S2 (Statistics 2) |
| Year | 2016 |
| Session | June |
| Marks | 18 |
| Paper | Download PDF ↗ |
| Mark scheme | Download PDF ↗ |
| Topic | Normal Distribution |
| Type | Probability calculation plus find unknown boundary |
| Difficulty | Moderate -0.3 This is a straightforward application of normal distribution techniques with standard procedures: z-score calculations, percentage point lookups, and solving simultaneous equations with inverse normal values. All parts follow textbook methods with no novel problem-solving required, though part (iv) requires slightly more algebraic manipulation than typical. Slightly easier than average due to routine nature. |
| Spec | 2.04e Normal distribution: as model N(mu, sigma^2)2.04f Find normal probabilities: Z transformation |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Marks | Guidance |
| \(P(50000 < X < 55000) = P\!\left(\dfrac{50000 - 50600}{3400} < Z < \dfrac{55000 - 50600}{3400}\right)\) | M1 | For standardising both. SOI. Penalise erroneous continuity corrections and wrong sd. Condone numerator(s) reversed. |
| \(= P(-0.176 < Z < 1.294) = \Phi(1.294) - (1 - \Phi(0.176)) = 0.9022 - 1 + 0.5699\) | M1 | For correct structure \(\Phi(\text{positive } z) - \Phi(\text{negative } z)\) |
| \(= 0.4721\) | A1 | CAO including use of difference tables (Answer from calculator 0.4722 and from tables interpolated 0.4723) |
| [3] |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Marks | Guidance |
| \(P(X > 45000) = P\left(Z \geq \frac{750-751.4}{2.5}\right)\left(Z > \frac{45000-50600}{3400}\right) = P(Z > -1.647)\) | B1* | For \(-1.647\) or \(-\Phi^{-1}(0.95) = -1.645\) or \(1.647\) seen with \(P(X < 56200)\) or numerator reversed |
| \(= \Phi(1.647) = 0.9502\) | B1* | For \(0.9502\) or \(45007\) or \(0.0498\), or B1 for \(-1.645\) if B1 for \(-1.647\) already awarded |
| \(0.9502 > 95\%\) so agree with claim | depE1* | For comparison seen e.g. \(-1.647 < -1.645\) or \(0.0498 < 0.05\) or \(1.647 > 1.645\) or \(95\%\) last longer than 45007 hours, and correct conclusion. Dependent on B1, B1 awarded |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Marks | Guidance |
| From tables \(\Phi^{-1}(0.999) = 3.09\) | B1 | \(\pm 3.09\) seen |
| \(\frac{h - 50600}{3400} = -3.09\) | M1 | For equation with their negative z-value |
| \(k = 50600 - (3.09 \times 3400) = 40100\) | A1 [3] | CAO Allow 40094, 40090 |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Marks | Guidance |
| \(P(Y < 60000) = 0.6 \Rightarrow P\left(Z < \frac{60000-\mu}{\sigma}\right) = 0.6\) | B1 | For \(\pm 0.2533\) or \(\pm 1.282\) seen |
| \(\Rightarrow \frac{60000-\mu}{\sigma} = \Phi^{-1}(0.6) = 0.2533\) | M1 | For an equation in \(\mu\), \(\sigma\), \(z\) and \(y\) formed. NB using \(z = \pm 0.2533\) with \(y = 60000\) or \(\pm 1.282\) with \(y = 50000\) |
| \(\Rightarrow 60000 = \mu + 0.2533\sigma\) | ||
| \(P(Y > 50000) = 0.9 \Rightarrow P\left(Z > \frac{50000-\mu}{\sigma}\right) = 0.9\) | ||
| \(\Rightarrow \frac{50000-\mu}{\sigma} = \Phi^{-1}(0.1) = -1.282\) | A1 | For two correct equations seen |
| \(\Rightarrow 50000 = \mu - 1.282\sigma\) | ||
| \(1.5353\sigma = 10000\), \(\sigma = 6513\) | A1 | CAO Allow 6510, 6515 |
| \(\Rightarrow \mu = 50000 + (1.282 \times 6513) = 58350\) | A1 [5] | CAO Allow 58400 |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Marks | Guidance |
| Two Normal curve shapes shown | G1 | For two Normal shapes including attempt at asymptotic behaviour with horizontal axis at each of the four ends. Penalise clear asymmetry |
| Means shown explicitly or by scale on a single diagram | G1 | If shown explicitly, positions must be consistent with horizontal scale if present. FT part (iv) |
| Greater width (variance) for Different model | G1 | FT part (iv) |
| Lower max height for Different model | G1 [4] | FT part (iv). If not labelled assume the larger mean represents Different model |
# Question 3:
## Part (i)
| Answer | Marks | Guidance |
|--------|-------|----------|
| $P(50000 < X < 55000) = P\!\left(\dfrac{50000 - 50600}{3400} < Z < \dfrac{55000 - 50600}{3400}\right)$ | M1 | For standardising both. SOI. Penalise erroneous continuity corrections and wrong sd. Condone numerator(s) reversed. |
| $= P(-0.176 < Z < 1.294) = \Phi(1.294) - (1 - \Phi(0.176)) = 0.9022 - 1 + 0.5699$ | M1 | For correct structure $\Phi(\text{positive } z) - \Phi(\text{negative } z)$ |
| $= 0.4721$ | A1 | CAO including use of difference tables (Answer from calculator 0.4722 and from tables interpolated 0.4723) |
| **[3]** | | |
## Question 3(ii):
| Answer | Marks | Guidance |
|--------|-------|----------|
| $P(X > 45000) = P\left(Z \geq \frac{750-751.4}{2.5}\right)\left(Z > \frac{45000-50600}{3400}\right) = P(Z > -1.647)$ | B1* | For $-1.647$ or $-\Phi^{-1}(0.95) = -1.645$ or $1.647$ seen with $P(X < 56200)$ or numerator reversed |
| $= \Phi(1.647) = 0.9502$ | B1* | For $0.9502$ or $45007$ or $0.0498$, or B1 for $-1.645$ if B1 for $-1.647$ already awarded |
| $0.9502 > 95\%$ so agree with claim | depE1* | **For comparison seen** e.g. $-1.647 < -1.645$ or $0.0498 < 0.05$ or $1.647 > 1.645$ or $95\%$ last longer than 45007 hours, **and correct conclusion**. Dependent on B1, B1 awarded |
## Question 3(iii):
| Answer | Marks | Guidance |
|--------|-------|----------|
| From tables $\Phi^{-1}(0.999) = 3.09$ | B1 | $\pm 3.09$ seen |
| $\frac{h - 50600}{3400} = -3.09$ | M1 | For equation with their negative z-value |
| $k = 50600 - (3.09 \times 3400) = 40100$ | A1 [3] | CAO Allow 40094, 40090 |
## Question 3(iv):
| Answer | Marks | Guidance |
|--------|-------|----------|
| $P(Y < 60000) = 0.6 \Rightarrow P\left(Z < \frac{60000-\mu}{\sigma}\right) = 0.6$ | B1 | For $\pm 0.2533$ or $\pm 1.282$ seen |
| $\Rightarrow \frac{60000-\mu}{\sigma} = \Phi^{-1}(0.6) = 0.2533$ | M1 | For an equation in $\mu$, $\sigma$, $z$ and $y$ formed. NB using $z = \pm 0.2533$ with $y = 60000$ or $\pm 1.282$ with $y = 50000$ |
| $\Rightarrow 60000 = \mu + 0.2533\sigma$ | | |
| $P(Y > 50000) = 0.9 \Rightarrow P\left(Z > \frac{50000-\mu}{\sigma}\right) = 0.9$ | | |
| $\Rightarrow \frac{50000-\mu}{\sigma} = \Phi^{-1}(0.1) = -1.282$ | A1 | For two correct equations seen |
| $\Rightarrow 50000 = \mu - 1.282\sigma$ | | |
| $1.5353\sigma = 10000$, $\sigma = 6513$ | A1 | CAO Allow 6510, 6515 |
| $\Rightarrow \mu = 50000 + (1.282 \times 6513) = 58350$ | A1 [5] | CAO Allow 58400 |
## Question 3(v):
| Answer | Marks | Guidance |
|--------|-------|----------|
| Two Normal curve shapes shown | G1 | For two Normal shapes including attempt at asymptotic behaviour with horizontal axis at each of the four ends. Penalise clear asymmetry |
| Means shown explicitly or by scale **on a single diagram** | G1 | If shown explicitly, positions must be consistent with horizontal scale if present. FT part (iv) |
| Greater width (variance) for Different model | G1 | FT part (iv) |
| Lower max height for Different model | G1 [4] | FT part (iv). If not labelled assume the larger mean represents Different model |
3 Many types of computer have cooling fans. The random variable $X$ represents the lifetime in hours of a particular model of cooling fan. $X$ is Normally distributed with mean 50600 and standard deviation 3400.\\
(i) Find $\mathrm { P } ( 50000 < X < 55000 )$.\\
(ii) The manufacturers claim that at least $95 \%$ of these fans last longer than 45000 hours. Is this claim valid?\\
(iii) Find the value of $h$ for which $99.9 \%$ of these fans last $h$ hours or more.\\
(iv) The random variable $Y$ represents the lifetime in hours of a different model of cooling fan. $Y$ is Normally distributed with mean $\mu$ and standard deviation $\sigma$. It is known that $\mathrm { P } ( Y < 60000 ) = 0.6$ and $\mathrm { P } ( Y > 50000 ) = 0.9$. Find the values of $\mu$ and $\sigma$.\\
(v) Sketch the distributions of lifetimes for both types of cooling fan on a single diagram.
\hfill \mbox{\textit{OCR MEI S2 2016 Q3 [18]}}