June 2024 Statistics Q4
4. The random variable \(X \sim \mathrm{B}(27,\ 0.35)\)
Historical records show that the proportion of defective items produced by a machine is 0.12
Following a maintenance service of the machine, a random sample of 60 items is taken and 3 defective items are found.
You should state your hypotheses clearly and use a 5% level of significance. (4)
| Scheme | Marks | AO |
|---|---|---|
| (i) 0.153588… awrt 0.154 | B1 | 3.4 |
| (ii) \(\mathrm{P}(X \leqslant 14) - \mathrm{P}(X \leqslant 11)\) with at least 1 from \(\mathrm{P}(X \leqslant 14) = 0.97707\ldots\) ; \(\mathrm{P}(X \leqslant 11) = 0.797603\ldots\) | M1 | 2.1 |
| \(= 0.17947\ldots\) awrt 0.179 | A1 | 1.1b |
| (3) |
Notes
(i) B1 for awrt 0.154
(ii) M1 for correct expression for \(\mathrm{P}(X \leqslant 14) - \mathrm{P}(X \leqslant 11)\) [o.e.]
with at least one correct probability substituted (2sf truncated or rounded)
or correct calculation 0.97707… – 0.797603… (2sf truncated or rounded)
ALT or \(\mathrm{P}(X = 12) + \mathrm{P}(X = 13) + \mathrm{P}(X = 14)\) with at least one from \({}^{27}C_{12}(0.35)^{12}(0.65)^{15} = 0.09176\ldots\) + \({}^{27}C_{13}(0.35)^{13}(0.65)^{14} = 0.05701\ldots\) + \({}^{27}C_{14}(0.35)^{14}(0.65)^{13} = 0.03069\ldots\)
with at least one correct probability calculation or value seen (2sf truncated or rounded)
A1 for awrt 0.179 allow 0.1795 correct answers scores 2 out of 2
| Scheme | Marks | AO |
|---|---|---|
| \(\mathrm{H}_0 : p = 0.12 \qquad \mathrm{H}_1 : p \lt 0.12\) | B1 | 2.5 |
| [\(D\) = no. of defective items in sample] \(D \sim \mathrm{B}(60,\ 0.12)\) | M1 | 2.1 |
| \([\mathrm{P}(D \leqslant 3)] = 0.06013\ldots\) awrt 0.060 or \([\mathrm{P}(D \leqslant 2)] =\) awrt 0.0196 with reference to CR [so CR: \(D \leqslant 2\)] [\(0.06\ldots \gt 5\%\) not significant, do not reject \(\mathrm{H}_0\)] | A1 | 1.1b |
| Insufficient evidence that proportion of defective items has decreased | A1 | 2.2b |
| (4) |
Notes
B1 for both hypotheses correct in terms of \(p\) or \(\pi\)
M1 for sight or correct use of \(\mathrm{B}(60,\ 0.12)\)
(implied by awrt 0.0601 or awrt 0.0405 or awrt 0.0196)
1st A1 for final answer awrt 0.060 (allow 0.06 if \(\mathrm{P}(D \leqslant 3)\) is seen with \(\mathrm{B}(60,\ 0.12)\))
or for critical region approach awrt 0.0196 with statement of CR or reference to CR
NB: \(\mathrm{P}(D \leqslant 2) =\) awrt 0.0196 on its own scores A0 here as it is treated as a \(p\)-value but can score B1ft in part (c)
2nd A1 (dep on M1A1 but independent of hypotheses) for a correct inference in context.
Must NOT reject \(\mathrm{H}_0\) (if stated) and mention underlined words o.e.
condone e.g. ‘ proportion of defective items is still 0.12/hasn’t changed’
allow e.g. ‘no’ for insufficient
allow proportion/probability/percentage but not number
allow e,g. ‘is less than 0.12’ for decreased
2nd A0 for contradictory statements
e.g. ’reject \(\mathrm{H}_0\) so no decrease in proportion of defective items’
SC A two-tailed test may score maximum in (b) B0M1A1A1 but must be 2×their \(p\)-value to 3 sf to score in part (c). Correct ft is 0.120 or better (do not accept 0.12 for the SC).
| Scheme | Marks | AO |
|---|---|---|
| “0.06…” | B1ft | 1.2 |
| (1) | ||
| (8 marks) |
Notes
B1ft for 0.06 or better allow as a percentage
or ft their final (\(p\)-value) answer from part (b) to 1sf [provided it is a probability]
NB: using a critical region approach in (b) scores B0ft if they state their CR probability as the \(p\)-value