OCR S2 2010 June — Question 7 11 marks

Exam BoardOCR
ModuleS2 (Statistics 2)
Year2010
SessionJune
Marks11
PaperDownload PDF ↗
Mark schemeDownload PDF ↗
TopicZ-tests (known variance)
TypeTwo-tail z-test
DifficultyStandard +0.3 This is a straightforward two-tail hypothesis test with all standard steps clearly signposted. Students must calculate sample mean and standard deviation from summaries, then perform a z-test - all routine S2 procedures. The 10% significance level and large sample size make it slightly easier than average, though it requires careful arithmetic and proper hypothesis test structure.
Spec5.05b Unbiased estimates: of population mean and variance5.05c Hypothesis test: normal distribution for population mean

7 A machine is designed to make paper with mean thickness 56.80 micrometres. The thicknesses, \(x\) micrometres, of a random sample of 300 sheets are summarised by $$n = 300 , \quad \Sigma x = 17085.0 , \quad \Sigma x ^ { 2 } = 973847.0 .$$ Test, at the \(10 \%\) significance level, whether the machine is producing paper of the designed thickness.

Question 7:
Main working:
AnswerMarks Guidance
AnswerMark Guidance
\(H_0: \mu = 56.8\), \(H_1: \mu \neq 56.8\)B2 Both correct; one error B1, but not \(\bar{x}\) etc
\(\bar{x} = 17085/300 = 56.95\)B1 56.95 or 57.0 seen or implied
\(\frac{300}{299}\left(\frac{973847}{300}-56.95^2\right)\)M1 Biased [2.8541]: M1M0A0
\(= 2.8637\ldots\)A1 Unbiased estimate, a.r.t. 2.86 [not 2.85]
M1Standardise with \(\sqrt{300}\), allow \(\sqrt{}\) errors, cc
\(z = \frac{56.95-56.8}{\sqrt{2.8637/300}} = 1.535\)A1 \(z \in [1.53, 1.54]\) or \(p \in [0.062, 0.063]\), not \(-1.535\)
\(1.535 < 1.645\) or \(0.0624 > 0.05\)A1 Compare explicitly \(z\) with 1.645 or \(p\) with 0.05, or \(2p > 0.1\), not from \(\mu=56.95\)
Part \((\alpha)\) — critical value method:
AnswerMarks Guidance
AnswerMark Guidance
\(\text{CV}\ 56.8 \pm 1.645\times\sqrt{\frac{2.8637}{300}}\)M1 \(56.8 + z\sigma/\sqrt{300}\), needn't have \(\pm\), allow \(\sqrt{}\) errors
A1\(z = 1.645\)
\(c = 56.96\), FT on \(z\), compare 56.95 \([c_L = 56.64]\)A1 FT
\(56.96 > 56.95\)
Conclusion:
AnswerMarks Guidance
AnswerMark Guidance
Do not reject \(H_0\)M1 Consistent first conclusion, needs 300, correct method and comparison
Insufficient evidence that mean thickness is wrongA1 FT 11 Conclusion in context, not too assertive, e.g. "evidence that" needed
# Question 7:

## Main working:
| Answer | Mark | Guidance |
|--------|------|----------|
| $H_0: \mu = 56.8$, $H_1: \mu \neq 56.8$ | B2 | Both correct; one error B1, but not $\bar{x}$ etc |
| $\bar{x} = 17085/300 = 56.95$ | B1 | 56.95 or 57.0 seen or implied |
| $\frac{300}{299}\left(\frac{973847}{300}-56.95^2\right)$ | M1 | Biased [2.8541]: M1M0A0 |
| $= 2.8637\ldots$ | A1 | Unbiased estimate, a.r.t. 2.86 [not 2.85] |
| | M1 | Standardise with $\sqrt{300}$, allow $\sqrt{}$ errors, cc |
| $z = \frac{56.95-56.8}{\sqrt{2.8637/300}} = 1.535$ | A1 | $z \in [1.53, 1.54]$ or $p \in [0.062, 0.063]$, not $-1.535$ |
| $1.535 < 1.645$ or $0.0624 > 0.05$ | A1 | Compare explicitly $z$ with 1.645 or $p$ with 0.05, or $2p > 0.1$, not from $\mu=56.95$ |

## Part $(\alpha)$ — critical value method:
| Answer | Mark | Guidance |
|--------|------|----------|
| $\text{CV}\ 56.8 \pm 1.645\times\sqrt{\frac{2.8637}{300}}$ | M1 | $56.8 + z\sigma/\sqrt{300}$, needn't have $\pm$, allow $\sqrt{}$ errors |
| | A1 | $z = 1.645$ |
| $c = 56.96$, FT on $z$, compare 56.95 $[c_L = 56.64]$ | A1 FT | |
| $56.96 > 56.95$ | | |

## Conclusion:
| Answer | Mark | Guidance |
|--------|------|----------|
| Do not reject $H_0$ | M1 | Consistent first conclusion, needs 300, correct method and comparison |
| Insufficient evidence that mean thickness is wrong | A1 FT **11** | Conclusion in context, not too assertive, e.g. "evidence that" needed |

---
7 A machine is designed to make paper with mean thickness 56.80 micrometres. The thicknesses, $x$ micrometres, of a random sample of 300 sheets are summarised by

$$n = 300 , \quad \Sigma x = 17085.0 , \quad \Sigma x ^ { 2 } = 973847.0 .$$

Test, at the $10 \%$ significance level, whether the machine is producing paper of the designed thickness.

\hfill \mbox{\textit{OCR S2 2010 Q7 [11]}}