OCR S2 2013 January — Question 1 4 marks

Exam BoardOCR
ModuleS2 (Statistics 2)
Year2013
SessionJanuary
Marks4
PaperDownload PDF ↗
Mark schemeDownload PDF ↗
TopicHypothesis test of binomial distributions
TypeCompare test procedures or parameters
DifficultyStandard +0.8 This question requires students to work backwards from a desired significance level to determine sample size and critical region, consulting cumulative binomial tables for multiple values of n. It demands understanding of how significance levels relate to critical regions and involves systematic comparison across different sample sizes—more conceptually demanding than routine hypothesis test execution.
Spec2.05b Hypothesis test for binomial proportion2.05c Significance levels: one-tail and two-tail

1 A random variable has the distribution \(\mathrm { B } ( n , p )\). It is required to test \(\mathrm { H } _ { 0 } : p = \frac { 2 } { 3 }\) against \(\mathrm { H } _ { 1 } : p < \frac { 2 } { 3 }\) at a significance level as close to \(1 \%\) as possible, using a sample of size \(n = 8,9\) or 10 . Use tables to find which value of \(n\) gives such a test, stating the critical region for the test and the corresponding significance level.
[0pt] [4]

Question 1:
AnswerMarks Guidance
AnswerMarks Guidance
\(n = 9\)B1 Stated explicitly
CR is \(\leq 2\)M1A1 2 seen but not \(\leq\): M1A0. Allow "\(P(\leq 2)\)". CR must be stated explicitly for A1. SR: \(\leq 3\) with 0.0424: (B1)M1A0. SR: If 0, give B1 for at least 3 of 0.0083, 0.0113, 0.0026, 0.0197, 0.0034 seen
0.0083A1 Or more SF. "\(n=9\), \(CR \geq 3\)", 0.0083 seen: B1M1A0A1
[4 marks]
## Question 1:

| Answer | Marks | Guidance |
|--------|-------|----------|
| $n = 9$ | B1 | Stated explicitly |
| CR is $\leq 2$ | M1A1 | 2 seen but not $\leq$: M1A0. Allow "$P(\leq 2)$". CR must be stated explicitly for A1. SR: $\leq 3$ with 0.0424: (B1)M1A0. SR: If 0, give B1 for at least 3 of 0.0083, 0.0113, 0.0026, 0.0197, 0.0034 seen |
| **0.0083** | A1 | Or more SF. "$n=9$, $CR \geq 3$", 0.0083 seen: B1M1A0A1 |

**[4 marks]**

---
1 A random variable has the distribution $\mathrm { B } ( n , p )$. It is required to test $\mathrm { H } _ { 0 } : p = \frac { 2 } { 3 }$ against $\mathrm { H } _ { 1 } : p < \frac { 2 } { 3 }$ at a significance level as close to $1 \%$ as possible, using a sample of size $n = 8,9$ or 10 . Use tables to find which value of $n$ gives such a test, stating the critical region for the test and the corresponding significance level.\\[0pt]
[4]\\

\hfill \mbox{\textit{OCR S2 2013 Q1 [4]}}