OCR MEI S2 2012 January — Question 2 19 marks

Exam BoardOCR MEI
ModuleS2 (Statistics 2)
Year2012
SessionJanuary
Marks19
PaperDownload PDF ↗
Mark schemeDownload PDF ↗
TopicPoisson distribution
TypeSingle period normal approximation - scaled period (normal approximation only)
DifficultyModerate -0.3 This is a straightforward application of standard Poisson distribution techniques with some routine extensions. Part (i) is pure recall of conditions, parts (ii)-(iv) involve direct formula application and calculator work, and part (v) requires knowing the normal approximation—all standard S2 content with no novel problem-solving required. Slightly easier than average due to the mechanical nature of most parts.
Spec5.02i Poisson distribution: random events model5.02j Poisson formula: P(X=x) = e^(-lambda)*lambda^x/x!5.02k Calculate Poisson probabilities5.04b Linear combinations: of normal distributions

2 The number of printing errors per page in a book is modelled by a Poisson distribution with a mean of 0.85 .
  1. State conditions for a Poisson distribution to be a suitable model for the number of printing errors per page.
  2. A page is chosen at random. Find the probability of
    (A) exactly 1 error on this page,
    (B) at least 2 errors on this page. 10 pages are chosen at random.
  3. Find the probability of exactly 10 errors in these 10 pages.
  4. Find the least integer \(k\) such that the probability of there being \(k\) or more errors in these 10 pages is less than \(1 \%\). 30 pages are chosen at random.
  5. Use a suitable approximating distribution to find the probability of no more than 30 errors in these 30 pages.

(i)
Answer: Errors have a uniform average rate of occurrence
and occur randomly and independently
AnswerMarks
Marks: E1E1
Guidance:
- E1 must refer to 'errors' not 'events', 'data' or 'conditions'.
- Condone 'constant/fixed average/mean rate/per page' but not 'constant average', 'constant rate' or 'uniform rate', etc.
- Allow large n and small p if both defined
- E1 for randomly and independently
- If 'errors' not referred to then SC1 if otherwise correct. Condone 'the number of errors'
(ii)(A)
Answer:
\[P(X = 1) = e^{-0.85} \frac{0.85^1}{1!} = 0.3633\]
AnswerMarks
Marks: M1A1
Guidance:
- M1 for attempt to find P(X = 1) either by Poisson p.d.f. or use of tables.
- A1 CAO 3s.f. for answers which round to 0.363
- www
- NOTE If P(X ≤ 1) used for final answer, award M0A0.
- Interpolation gives 0.79065 − 0.42795 = 0.3627
- www
(ii)(B)
Answer:
\[P(X \geq 2) = 1 - P(X \leq 1) = 1 - e^{-0.85}\frac{0.85^0}{0!} - e^{-0.85}\frac{0.85^1}{1!}\]
\[= 1 - 0.4274 - 0.3633 = 0.2093\]
AnswerMarks Guidance
Marks: M1M1 A1
Guidance:
- M1 for method for P(X = 0)
- M1 for correct structure used
- A1 CAO 3s.f for answers which round to 0.209. Allow 0.2094 if interpolation used. [Interpolation gives 0.42795 for P(X = 0) and 0.20935 for P(X ≤ 1)]
(iii)
Answer:
New \(\lambda = 10 \times 0.85 = 8.5\)
P(Exactly 10 in 10 pages) = \(0.7634 - 0.6530 = 0.1104\)
Or \(= e^{-8.5}\frac{8.5^{10}}{10!} = 0.1104\)
AnswerMarks Guidance
Marks: B1M1 A1
Guidance:
- B1 for 8.5
- M1 for P(X = 10) calculation using \(\lambda = 8.5\)
- CAO Allow 0.110 and 0.11
- www
- Award M1 only if \(\lambda = 8.5\) used
Question 2(iv)
Answer: So P(k − 1 or less in 10 pages) > 99%
From tables
\[P(X \leq 15) = 0.9862, \quad P(X \leq 16) = 0.9934\]
\[P(X \geq 16) = 1 - P(X \leq 15) = 0.0138 > 1\%\]
\[P(X \geq 17) = 1 - P(X \leq 16) = 0.0066 < 1\%\]
P(k or more in 10 pages) < 1% means
k − 1 = 16, k = 17
AnswerMarks Guidance
Marks: M1A1 A1
Guidance:
- M1 for P(X ≤ k − 1) > 0.99 seen, or evidence of a search for values > 0.99 from cumulative Poisson tables seen.
- A1 for finding either one of 0.9862 and 0.9934 (or either one of 0.0138 and 0.0066)
- A1 for both (3s.f.)
- A1 CAO for k = 17
- SC1 for evidence of a search for values > 0.9 from cumulative Poisson tables seen. Or for k = 17with no supporting evidence seen.
Question 2(v)
Answer:
Mean number in 30 pages = \(30 \times 0.85 = 25.5\)
Using Normal approx. to the Poisson,
\[X \sim N(25.5, 25.5)\]
\[P(X \leq 30) = P\left(Z \leq \frac{30.5-25.5}{\sqrt{25.5}}\right)\]
\[= P(Z < 0.9901) = \Phi(0.9901) = 0.8389\]
AnswerMarks Guidance
Marks: B1B1 B1
Guidance:
- B1 for Normal approx attempted.(SOI)
- B1 for correct parameters (SOI)
- B1 for correct continuity correction.
- M1 for correct structure with their parameters
- CAO (Do not FT wrong or omitted CC) Allow 0.839
- NOTE Using B(100, 0.4873) gives 0.0494. which gets 0/5
- SC If p small enough to justify a Poisson approximation, e.g. 0.05, then B1 for Poisson used, B1 ft for parameter, M1 for structure,M1 attempt at summation, A0
## (i)

**Answer:** Errors have a **uniform average rate of occurrence**

and occur **randomly and independently**

**Marks:** E1 | E1

**Guidance:**
- E1 must refer to 'errors' not 'events', 'data' or 'conditions'.
- Condone 'constant/fixed average/mean rate/per page' but not 'constant average', 'constant rate' or 'uniform rate', etc.
- Allow large n and small p if both defined
- E1 for randomly **and independently**
- If 'errors' not referred to then SC1 if otherwise correct. Condone 'the number of errors'

---

## (ii)(A)

**Answer:**
$$P(X = 1) = e^{-0.85} \frac{0.85^1}{1!} = 0.3633$$

**Marks:** M1 | A1

**Guidance:**
- M1 for attempt to find P(X = 1) either by Poisson p.d.f. or use of tables.
- A1 CAO 3s.f. for answers which round to 0.363
- www
- NOTE If P(X ≤ 1) used for final answer, award M0A0.
- Interpolation gives 0.79065 − 0.42795 = 0.3627
- www

---

## (ii)(B)

**Answer:**
$$P(X \geq 2) = 1 - P(X \leq 1) = 1 - e^{-0.85}\frac{0.85^0}{0!} - e^{-0.85}\frac{0.85^1}{1!}$$

$$= 1 - 0.4274 - 0.3633 = 0.2093$$

**Marks:** M1 | M1 | A1

**Guidance:**
- M1 for method for P(X = 0)
- M1 for correct structure used
- A1 CAO 3s.f for answers which round to 0.209. Allow 0.2094 if interpolation used. [Interpolation gives 0.42795 for P(X = 0) and 0.20935 for P(X ≤ 1)]

---

## (iii)

**Answer:**
New $\lambda = 10 \times 0.85 = 8.5$

P(Exactly 10 in 10 pages) = $0.7634 - 0.6530 = 0.1104$

Or $= e^{-8.5}\frac{8.5^{10}}{10!} = 0.1104$

**Marks:** B1 | M1 | A1

**Guidance:**
- B1 for 8.5
- M1 for P(X = 10) calculation using $\lambda = 8.5$
- CAO Allow 0.110 and 0.11
- www
- Award M1 only if $\lambda = 8.5$ used

---

# Question 2(iv)

**Answer:** So P(k − 1 or less in 10 pages) > 99%

From tables
$$P(X \leq 15) = 0.9862, \quad P(X \leq 16) = 0.9934$$

$$P(X \geq 16) = 1 - P(X \leq 15) = 0.0138 > 1\%$$
$$P(X \geq 17) = 1 - P(X \leq 16) = 0.0066 < 1\%$$

P(k or more in 10 pages) < 1% means
k − 1 = 16, k = 17

**Marks:** M1 | A1 | A1

**Guidance:**
- M1 for P(X ≤ k − 1) > 0.99 seen, or evidence of a search for values > 0.99 from cumulative Poisson tables seen.
- A1 for finding either one of 0.9862 and 0.9934 (or either one of 0.0138 and 0.0066)
- A1 for both (3s.f.)
- A1 CAO for k = 17
- SC1 for evidence of a search for values > **0.9** from cumulative Poisson tables seen. Or for k = 17with no supporting evidence seen.

---

# Question 2(v)

**Answer:** 
Mean number in 30 pages = $30 \times 0.85 = 25.5$

Using Normal approx. to the Poisson,
$$X \sim N(25.5, 25.5)$$

$$P(X \leq 30) = P\left(Z \leq \frac{30.5-25.5}{\sqrt{25.5}}\right)$$

$$= P(Z < 0.9901) = \Phi(0.9901) = 0.8389$$

**Marks:** B1 | B1 | B1 | M1 | A1

**Guidance:**
- B1 for Normal approx attempted.(SOI)
- B1 for correct parameters (SOI)
- B1 for correct continuity correction.
- M1 for correct structure with their parameters
- CAO (Do not FT wrong or omitted CC) Allow 0.839
- NOTE Using B(100, 0.4873) gives 0.0494. which gets 0/5
- SC If p small enough to justify a Poisson approximation, e.g. 0.05, then B1 for Poisson used, B1 ft for parameter, M1 for structure,M1 attempt at summation, A0

---
2 The number of printing errors per page in a book is modelled by a Poisson distribution with a mean of 0.85 .
\begin{enumerate}[label=(\roman*)]
\item State conditions for a Poisson distribution to be a suitable model for the number of printing errors per page.
\item A page is chosen at random. Find the probability of\\
(A) exactly 1 error on this page,\\
(B) at least 2 errors on this page.

10 pages are chosen at random.
\item Find the probability of exactly 10 errors in these 10 pages.
\item Find the least integer $k$ such that the probability of there being $k$ or more errors in these 10 pages is less than $1 \%$.

30 pages are chosen at random.
\item Use a suitable approximating distribution to find the probability of no more than 30 errors in these 30 pages.
\end{enumerate}

\hfill \mbox{\textit{OCR MEI S2 2012 Q2 [19]}}