Skip to content
CalcGospel 國際數學圖譜
返回

S2 2023 Oct Q5

A Level / Edexcel / S2

IAL 2023 Oct Paper · Question 5

题目

Problem

A supermarket receives complaints at a mean rate of 6 per week.

(a) State one assumption necessary, in order for a Poisson distribution to be used to model the number of complaints received by the supermarket.

(1)

(b) Find the probability that, in a given week, there are

(i) fewer than 3 complaints received by the supermarket,

(ii) at least 6 complaints received by the supermarket.

(3)

In a randomly selected week, the supermarket received 12 complaints.

(c) Test, at the 5% level of significance, whether or not there is evidence that the mean number of complaints is greater than 6 per week.

State your hypotheses clearly.

(5)

Following changes made by the supermarket, it received 26 complaints over a 6-week period.

(d) Use a suitable approximation to test whether or not there is evidence that, following the changes, the mean number of complaints received is less than 6 per week.

You should state your hypotheses clearly and use a 5% significance level.

(7)

(Total for Question 5 is 16 marks)

题目中文翻译

一家超市以每周平均6 次的速率收到投诉。

(a) 说明一个必要的假设,以便可以使用泊松分布来建模该超市收到的投诉数。

(1)

(b) 求在给定一周内以下情况的概率

(i) 该超市收到少于3 次投诉,

(ii) 该超市收到至少6 次投诉。

(3)

在随机选择的一周内,该超市收到12 次投诉。

(c) 在5% 的显著性水平下,检验是否有证据表明平均每周投诉数大于6 次。

清楚地陈述你的假设。

(5)

在超市做出改变后,它在6 周内收到26 次投诉。

(d) 使用合适的近似方法检验是否有证据表明改变后平均每周收到的投诉数少于6 次。

你应清楚地陈述你的假设,并使用5% 的显著性水平。

(7)

(第5 题共16 分)

解答

(a)

解法一

思路

展开

泊松模型要求事件独立发生、平均发生率保持不变,并且事件通常逐个发生。题目只要求一个必要假设,结合投诉语境说明其中一点即可。

答题过程

展开

The complaints received by the supermarket occur independently of one another.

(b)(i)

解法一

思路

展开

一周投诉数服从均值为 6 的泊松分布。“少于 3 次”即投诉数为 0、1 或 2,求对应累积概率。

答题过程

展开

Let XX be the number of complaints received in one week. Then

X∼Po⁡(6).X\sim\operatorname{Po}(6).

Therefore,

 P(X<3) =P(X⩽2) =e−6(1+6+622!) =0.061968… ≈0.0620.\begin{align*} &\,P(X<3)\\[2mm] &\,=P(X\leqslant2)\\ &\,=\mathrm{e}^{-6}\Big(1+6+\frac{6^2}{2!}\Big)\\ &\,=0.061968\ldots\\ &\,\approx\boxed{0.0620}. \end{align*}

(b)(ii)

解法一

思路

展开

“至少 6 次”即 X⩾6X\geqslant6。用补事件,从 1 减去 X⩽5X\leqslant5 的累积概率。

答题过程

展开  P(X⩾6) =1−P(X⩽5) =0.55432… ≈0.554.\begin{align*} &\,P(X\geqslant6)\\[2mm] &\,=1-P(X\leqslant5)\\ &\,=0.55432\ldots\\ &\,\approx\boxed{0.554}. \end{align*}

(c)

解法一

思路

展开

要检验平均投诉数是否大于 6,应作右尾泊松检验。在原假设下以 Po⁡(6)\operatorname{Po}(6) 计算观察到 12 次或更多投诉的概率,并与 5%5\% 比较,最后给出有语境的结论。

答题过程

展开

Let λ\lambda be the mean number of complaints received per week.

H0: λ=6,H1: λ>6.\begin{align*} H_0:&\,\lambda=6,\\ H_1:&\,\lambda>6. \end{align*}

Under H0H_0, X∼Po⁡(6)X\sim\operatorname{Po}(6). The pp-value is

 P(X⩾12) =1−P(X⩽11) =0.0200919635…\begin{align*} &\,P(X\geqslant12)\\[2mm] &\,=1-P(X\leqslant11)\\ &\,=0.0200919635\ldots \end{align*}

Since 0.0201<0.050.0201<0.05, reject H0H_0. There is sufficient evidence, at the 5%5\% significance level, that the mean number of complaints received is greater than 6 per week.

解法二

思路

展开

也可以先求右尾临界区域。寻找使右尾概率不超过 0.050.05 的最小投诉数,并检查相邻边界以确认临界值;再判断观测值 12 是否落入临界区域。

答题过程

展开

Under H0H_0, X∼Po⁡(6)X\sim\operatorname{Po}(6). Now

P(X⩾11)≈0.0426,P(X\geqslant11)\approx0.0426,

whereas

P(X⩾10)≈0.0839.P(X\geqslant10)\approx0.0839.

Therefore the 5%5\% upper-tail critical region is

X⩾11.X\geqslant11.

Since the observed value is 1212, it lies in the critical region. Reject H0H_0 and conclude that there is sufficient evidence that the mean number of complaints is greater than 6 per week.

(d)

解法一

思路

展开

这里检验改变后的每周平均投诉数是否低于 6,因此作左尾检验。原假设下 6 周总投诉数服从 Po⁡(36)\operatorname{Po}(36);按题意使用均值与方差均为 36 的正态分布近似。对观测事件“总数不超过 26”使用连续性修正 26.5,求左尾概率并与 5%5\% 比较。

答题过程

展开

Let λ\lambda be the mean number of complaints received per week after the changes.

H0: λ=6,H1: λ<6.\begin{align*} H_0:&\,\lambda=6,\\ H_1:&\,\lambda<6. \end{align*}

Under H0H_0, let TT be the total number of complaints received in 6 weeks. Then

T∼Po⁡(36).T\sim\operatorname{Po}(36).

Using a normal approximation,

 T≈Y, Y∼N⁡(36,36).\begin{align*} &\,T\approx Y,\\ &\,Y\sim\operatorname{N}(36,36). \end{align*}

Let Z∼N⁡(0,1)Z\sim\operatorname{N}(0,1). Applying a continuity correction,

 P(T⩽26) ≈P(Y<26.5) =P(Z<26.5−366) =P(Z<−1.5833…) =0.0566727546…\begin{align*} &\,P(T\leqslant26)\\[2mm] &\,\approx P(Y<26.5)\\ &\,=P\Big(Z<\frac{26.5-36}{6}\Big)\\ &\,=P(Z<-1.5833\ldots)\\ &\,=0.0566727546\ldots \end{align*}

Since 0.0567>0.050.0567>0.05, do not reject H0H_0. There is insufficient evidence, at the 5%5\% significance level, that the mean number of complaints received after the changes is less than 6 per week.

解法二

思路

展开

先求正态近似下的左尾临界区域。要使整数边界 xx 落在左尾 5%5\% 内,带连续性修正的标准化边界应小于 −1.6449-1.6449。解出不等式后取满足条件的最大整数,再判断观测值 26 是否在区域内。

答题过程

展开

Under H0H_0, use Y∼N⁡(36,36)Y\sim\operatorname{N}(36,36). Let xx be an integer in the lower-tail critical region. We require

x+0.5−366<−1.6449.\frac{x+0.5-36}{6}<-1.6449.

Hence

x+0.5< 36−1.6449(6),x< 25.6306.\begin{align*} x+0.5<&\,36-1.6449(6),\\ x<&\,25.6306. \end{align*}

The largest integer satisfying this inequality is 25, so the critical region is

T⩽25.T\leqslant25.

The observed value, 26, is not in the critical region. Therefore, do not reject H0H_0; there is insufficient evidence that the mean number of complaints after the changes is less than 6 per week.