| Exam Board | OCR MEI |
|---|---|
| Module | S1 (Statistics 1) |
| Marks | 18 |
| Paper | Download PDF ↗ |
| Mark scheme | Download PDF ↗ |
| Topic | Hypothesis test of binomial distributions |
| Type | Multiple binomial probability calculations |
| Difficulty | Standard +0.3 This is a straightforward S1 question testing standard binomial probability calculations and two-tailed hypothesis tests. Part (i) requires routine use of binomial probability formulas and expectation. Parts (ii) and (iii) involve standard hypothesis test procedures with the critical probability value given in (iii), removing computational difficulty. The question is slightly easier than average as it's methodical application of learned procedures with no novel problem-solving required. |
| Spec | 2.04b Binomial distribution: as model B(n,p)2.04c Calculate binomial probabilities2.05a Hypothesis testing language: null, alternative, p-value, significance2.05b Hypothesis test for binomial proportion2.05c Significance levels: one-tail and two-tail |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Marks | Guidance |
| \(X \sim B(10, 0.35)\) | ||
| \(P(5 \text{ accessing internet}) = \binom{10}{5} \times 0.35^5 \times 0.65^5\) | M1 | or \(0.35^5 \times 0.65^5\); with \(p + q = 1\) |
| M1 | For \(\binom{10}{5} \times p^5 \times q^5\); also for \(252 \times 0.0006094\) | |
| \(= 0.1536\) | A1 [3] | cao; allow 0.15 or better; NB 0.153 gets A0 |
| OR from tables \(= 0.9051 - 0.7515 = 0.1536\) | M2, A1 | For \(0.9051 - 0.7515\); cao |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Marks | Guidance |
| \(P(X \geq 5) = 1 - P(X \leq 4)\) | M1 | For 0.7515 |
| \(= 1 - 0.7515\) | ||
| \(= 0.2485\) | A1 [2] | cao; accept 0.25 or better – allow 0.248 or 0.249; calculation of individual probabilities gets B2 if fully correct 0.25 or better, otherwise B0 |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Marks | Guidance |
| \(E(X) = np = 10 \times 0.35\) | M1 | For \(10 \times 0.35\) |
| \(= 3.5\) | A1 [2] | cao; if any indication of rounding to 3 or 4 allow M1A0 |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Marks | Guidance |
| Let \(X \sim B(20, 0.35)\) | ||
| Let \(p\) = probability of a customer using the internet (for population) | B1 | For definition of \(p\) in context; minimum needed is \(p\) = probability of using internet; allow \(p = P(\text{using internet})\); definition must include word probability (or chance or proportion or percentage or likelihood but NOT possibility); do NOT allow '\(p\) = the probability of using internet is different' |
| \(H_0: p = 0.35\) | B1 | Allow \(p = 35\%\); allow only \(p\) or \(\theta\) or \(\pi\) or \(\rho\); allow any single symbol if defined; allow \(H_0 = p = 0.35\); allow \(p = \frac{7}{20}\) or \(p = \frac{35}{100}\); do not allow \(H_0: P(X=x) = 0.35\); do not allow \(H_0\) and \(H_1\) reversed |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Marks | Guidance |
| \(H_1: p \neq 0.35\) | B1 | For \(H_1\). Allow '\(p < 0.35\) or \(p > 0.35\)' in place of \(p \neq 0.35\). Do not allow if \(H_1\) wrong. |
| \(H_1\) has this form because the test is to investigate whether the proportion is different (rather than lower or higher). | E1 | |
| \(P(X \geq 10)\) | B1 | For notation \(P(X \geq 10)\) or \(P(X > 9)\) or \(1 - P(X \leq 9)\) (as long as no incorrect notation). This mark may be implied by 0.1218. No further marks if point probs used. \(P(X=10) = 0.0686\) (do not even give notation mark for correct notation). DO NOT FT wrong \(H_1\), but see extra notes. |
| \(= 1 - 0.8782 = 0.1218\) | B1* | For 0.1218. Allow 0.12. Or for \(1 - 0.8782\). Independent of previous mark. |
| \(> 2.5\) | M1* dep | For comparison with 2.5% |
| So not significant. Conclude that there is not enough evidence to indicate that the probability is different. (Must state 'probability', not just 'p') | A1* E1* dep on A1 | Allow 'accept \(H_0\)' or 'reject \(H_1\)'. Must include 'sufficient evidence' or something similar such as 'to suggest that' ie an element of doubt either in the A or E mark. |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Marks | Guidance |
| LOWER TAIL: \(P(X \leq 2) = 0.0121 < 2.5\%\); \(P(X \leq 3) = 0.0444 > 2.5\%\) | B1 | For either probability. Do not insist on correct notation as candidates have to work out two probabilities for full marks. If only upper tail of CR given (or only upper tail justified), allow max 4/5 for final 5 marks. |
| UPPER TAIL: \(P(X \geq 11) = 1 - P(X \leq 10) = 1 - 0.9468 = 0.0532 > 2.5\%\); \(P(X \geq 12) = 1 - P(X \leq 11) = 1 - 0.9804 = 0.0196 < 2.5\%\) | B1 | For either probability |
| So critical region is \(\{0,1,2,12,13,14,15,16,17,18,19,20\}\) | M1* dep | cao dep on at least one correct comparison with 2.5%. No marks if CR not justified. Condone \(\{0,1,2,12,\ldots 20\}\), \(X \leq 2\), \(X \geq 12\), oe but not \(P(X \leq 2)\) etc |
| So not significant. Conclude that there is not enough evidence to indicate that the probability is different. | A1* E1* dep on A1 | NB If CR found correctly then \(P(X=10)\) subsequently found but candidate says '10 not in CR' then allow up to all last five marks. If do not say '10 not in CR' allow none of last five marks. |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Marks | Guidance |
| \(0.0022 < 2.5\%\), so reject \(H_0\). Significant. | B1 | For either reject \(H_0\) or significant, dep on correct comparison |
| Conclude that there is enough evidence to indicate that the probability is different. | E1* dep | Dep on good attempt at correct hypotheses in part (ii). If they have \(H_1: p > 0.35\), allow SC1 if all correct including comparison with 5%. |
## Question 3:
### Part (i)(A)
| Answer | Marks | Guidance |
|--------|-------|----------|
| $X \sim B(10, 0.35)$ | | |
| $P(5 \text{ accessing internet}) = \binom{10}{5} \times 0.35^5 \times 0.65^5$ | M1 | or $0.35^5 \times 0.65^5$; with $p + q = 1$ |
| | M1 | For $\binom{10}{5} \times p^5 \times q^5$; also for $252 \times 0.0006094$ |
| $= 0.1536$ | A1 [3] | cao; allow 0.15 or better; NB 0.153 gets A0 |
| **OR** from tables $= 0.9051 - 0.7515 = 0.1536$ | M2, A1 | For $0.9051 - 0.7515$; cao |
### Part (i)(B)
| Answer | Marks | Guidance |
|--------|-------|----------|
| $P(X \geq 5) = 1 - P(X \leq 4)$ | M1 | For 0.7515 |
| $= 1 - 0.7515$ | | |
| $= 0.2485$ | A1 [2] | cao; accept 0.25 or better – allow 0.248 or 0.249; calculation of individual probabilities gets B2 if fully correct 0.25 or better, otherwise B0 |
### Part (i)(C)
| Answer | Marks | Guidance |
|--------|-------|----------|
| $E(X) = np = 10 \times 0.35$ | M1 | For $10 \times 0.35$ |
| $= 3.5$ | A1 [2] | cao; if any indication of rounding to 3 or 4 allow M1A0 |
### Part (ii)
| Answer | Marks | Guidance |
|--------|-------|----------|
| Let $X \sim B(20, 0.35)$ | | |
| Let $p$ = probability of a customer using the internet (for population) | B1 | For definition of $p$ in context; minimum needed is $p$ = probability of using internet; allow $p = P(\text{using internet})$; definition must include word probability (or chance or proportion or percentage or likelihood but NOT possibility); do NOT allow '$p$ = the probability of using internet is different' |
| $H_0: p = 0.35$ | B1 | Allow $p = 35\%$; allow only $p$ or $\theta$ or $\pi$ or $\rho$; allow any single symbol if defined; allow $H_0 = p = 0.35$; allow $p = \frac{7}{20}$ or $p = \frac{35}{100}$; do not allow $H_0: P(X=x) = 0.35$; do not allow $H_0$ and $H_1$ reversed |
## Question 3 (ii):
| Answer | Marks | Guidance |
|--------|-------|----------|
| $H_1: p \neq 0.35$ | B1 | For $H_1$. Allow '$p < 0.35$ or $p > 0.35$' in place of $p \neq 0.35$. Do not allow if $H_1$ wrong. |
| $H_1$ has this form because the test is to investigate whether the proportion is different (rather than lower or higher). | E1 | |
| $P(X \geq 10)$ | B1 | For notation $P(X \geq 10)$ or $P(X > 9)$ or $1 - P(X \leq 9)$ (as long as no incorrect notation). This mark may be implied by 0.1218. No further marks if point probs used. $P(X=10) = 0.0686$ (do not even give notation mark for correct notation). DO NOT FT wrong $H_1$, but see extra notes. |
| $= 1 - 0.8782 = 0.1218$ | B1* | For 0.1218. Allow 0.12. Or for $1 - 0.8782$. Independent of previous mark. |
| $> 2.5$ | M1* dep | For comparison with 2.5% |
| So not significant. Conclude that there is not enough evidence to indicate that the probability is different. (Must state 'probability', not just 'p') | A1* E1* dep on A1 | Allow 'accept $H_0$' or 'reject $H_1$'. Must include 'sufficient evidence' or something similar such as 'to suggest that' ie an element of doubt either in the A or E mark. |
**Alternative Method (Critical Region):**
| Answer | Marks | Guidance |
|--------|-------|----------|
| LOWER TAIL: $P(X \leq 2) = 0.0121 < 2.5\%$; $P(X \leq 3) = 0.0444 > 2.5\%$ | B1 | For either probability. Do not insist on correct notation as candidates have to work out two probabilities for full marks. If only upper tail of CR given (or only upper tail justified), allow max 4/5 for final 5 marks. |
| UPPER TAIL: $P(X \geq 11) = 1 - P(X \leq 10) = 1 - 0.9468 = 0.0532 > 2.5\%$; $P(X \geq 12) = 1 - P(X \leq 11) = 1 - 0.9804 = 0.0196 < 2.5\%$ | B1 | For either probability |
| So critical region is $\{0,1,2,12,13,14,15,16,17,18,19,20\}$ | M1* dep | cao dep on at least one correct comparison with 2.5%. No marks if CR not justified. Condone $\{0,1,2,12,\ldots 20\}$, $X \leq 2$, $X \geq 12$, oe but not $P(X \leq 2)$ etc |
| So not significant. Conclude that there is not enough evidence to indicate that the probability is different. | A1* E1* dep on A1 | NB If CR found correctly then $P(X=10)$ subsequently found but candidate says '10 not in CR' then allow up to all last five marks. If do not say '10 not in CR' allow none of last five marks. |
**[9 marks total]**
---
## Question 3 (iii):
| Answer | Marks | Guidance |
|--------|-------|----------|
| $0.0022 < 2.5\%$, so reject $H_0$. Significant. | B1 | For either reject $H_0$ or significant, dep on correct comparison |
| Conclude that there is enough evidence to indicate that the probability is different. | E1* dep | Dep on good attempt at correct hypotheses in part (ii). If they have $H_1: p > 0.35$, allow SC1 if all correct including comparison with 5%. |
**[2 marks total]**
3 A coffee shop provides free internet access for its customers. It is known that the probability that a randomly selected customer is accessing the internet is 0.35 , independently of all other customers.
\begin{enumerate}[label=(\roman*)]
\item 10 customers are selected at random.\\
(A) Find the probability that exactly 5 of them are accessing the internet.\\
(B) Find the probability that at least 5 of them are accessing the internet.\\
(C) Find the expected number of these customers who are accessing the internet.
Another coffee shop also provides free internet access. It is suspected that the probability that a randomly selected customer at this coffee shop is accessing the internet may be different from 0.35 . A random sample of 20 customers at this coffee shop is selected. Of these, 10 are accessing the internet.
\item Carry out a hypothesis test at the $5 \%$ significance level to investigate whether the probability for this coffee shop is different from 0.35. Give a reason for your choice of alternative hypothesis.
\item To get a more reliable result, a much larger random sample of 200 customers is selected over a period of time, and another hypothesis test is carried out. You are given that 90 of the 200 customers were accessing the internet. You are also given that, if $X$ has the binomial distribution with parameters $n = 200$ and $p = 0.35$, then $\mathrm { P } ( X \geqslant 90 ) = 0.0022$. Using the same hypotheses and significance level which you used in part (ii), complete this test.
\end{enumerate}
\hfill \mbox{\textit{OCR MEI S1 Q3 [18]}}