Skip to content
CalcGospel 國際數學圖譜
返回

IAL 2022 Jan S2 Q5

A Level / Edexcel / S2

IAL 2022 Jan Paper · Question 5

题目

Problem

Applicants for a pilot training programme with a passenger airline are screened for colour blindness. Past records show that the proportion of applicants identified as colour blind is 0.045

(a) Write down a suitable model for the distribution of the number of applicants identified as colour blind from a total of nn applicants.

(1)

(b) State one assumption necessary for this distribution to be a suitable model of this situation.

(1)

(c) Using a suitable approximation, find the probability that exactly 5 out of 120 applicants are identified as colour blind.

(3)

(d) Explain why the approximation that you used in part (c) is appropriate.

(2)

Jaymini claims that 75% of all applicants for this training programme go on to become pilots.

From a random sample of 96 applicants for this training programme 67 go on to become pilots.

(e) Using a suitable approximation, test Jaymini’s claim at the 5% level of significance. State your hypotheses clearly.

(7)

(Total for Question 5 is 14 marks)

题目中文翻译

一项面向客运航空公司的飞行员培训项目会筛查申请者的色盲情况。以往记录显示,被识别为色盲的申请者比例为 0.045。

(a) 写出一个合适的模型,用于表示在总共 nn 名申请者中,被识别为色盲的人数分布。

(b) 说明使该分布适合此情境的一个必要假设。

(c) 使用适当的近似方法,求在 120 名申请者中恰好 5 人被识别为色盲的概率。

(d) 解释为什么 (c) 中使用的近似是适当的。

Jaymini 声称该培训项目的所有申请者中有 75% 最终会成为飞行员。

从该培训项目的 96 名申请者随机抽样,其中 67 人最终成为飞行员。

(e) 使用适当的近似方法,在 5% 显著性水平下检验 Jaymini 的主张。 清楚陈述你的假设。

(第 5 题共 14 分)

解答

(a)

解法一

思路

展开

每名申请者只有“被识别为色盲”或“未被识别为色盲”两种结果,概率固定为 0.0450.045,因此人数适合用二项分布建模。

答题过程

展开

Let XX be the number of applicants identified as colour blind. A suitable model is

XB(n,0.045).\boxed{X\sim\operatorname{B}(n,0.045)}.

(b)

解法一

思路

展开

二项模型需要各次试验独立且成功概率不变。结合题目情境,可说明不同申请者的色盲识别结果相互独立,或该比例随时间保持不变。

答题过程

展开

The applicants’ colour-blindness outcomes are independent.

(c)

解法一

思路

展开

原分布为 B(120,0.045)\operatorname{B}(120,0.045)。由于 nn 大而 pp 很小,用参数 λ=np=5.4\lambda=np=5.4 的泊松分布近似,再计算恰好 5 人的单点概率。

答题过程

展开

Let XX be the number identified as colour blind. Then

XB(120,0.045).X\sim\operatorname{B}(120,0.045).

Using the Poisson approximation,

XPo(120×0.045)=Po(5.4).X\approx\operatorname{Po}(120\times0.045) =\operatorname{Po}(5.4).

Therefore,

P(X=5)=e5.4(5.4)55!=0.172821=0.173(3 s.f.).\begin{align*} P(X=5)=&\,\frac{\mathrm{e}^{-5.4}(5.4)^5}{5!}\\[2mm] =&\,0.172821\ldots\\[2mm] =&\,\boxed{0.173}\quad\text{(3 s.f.)}. \end{align*}

(d)

解法一

思路

展开

泊松分布近似二项分布的适用条件是试验次数较大、单次成功概率较小。本题 n=120n=120 较大而 p=0.045p=0.045 很小,符合条件。

答题过程

展开

The approximation is appropriate because the binomial distribution has a large value of nn and a very small value of pp:

n=120,p=0.045.n=120,\qquad p=0.045.

(e)

解法一

思路

展开

pp 为申请者最终成为飞行员的概率。主张是 p=0.75p=0.75,而检验是否“不同于”该比例,所以采用双侧检验。样本量足够大,可用正态分布近似二项分布,并对观测值 67 使用连续性修正 67.567.5。双侧 5% 检验的单尾临界概率是 0.0250.025

答题过程

展开

Let pp be the probability that an applicant goes on to become a pilot.

H0:p=0.75,H1:p0.75.H_0:p=0.75, \qquad H_1:p\ne0.75.

Under H0H_0, if XX is the number who become pilots, then

XB(96,0.75).X\sim\operatorname{B}(96,0.75).

Using a normal approximation,

XN(96(0.75),96(0.75)(0.25))=N(72,18).X\approx N\big(96(0.75),96(0.75)(0.25)\big) =N(72,18).

With a continuity correction,

P(X67)P(Y<67.5)=P(Z<67.57218)=P(Z<1.06066)=0.1445\begin{align*} P(X\leqslant67) \approx&\,P(Y<67.5)\\[2mm] =&\,P\left( Z<\frac{67.5-72}{\sqrt{18}} \right)\\[2mm] =&\,P(Z<-1.06066\ldots)\\[2mm] =&\,0.1445\ldots \end{align*}

Since this is a two-tailed test,

0.1445>0.025.0.1445>0.025.

Therefore, we do not reject H0H_0. There is insufficient evidence at the 5% significance level against Jaymini’s claim that 75% of applicants go on to become pilots.

解法二

思路

展开

官方评分资料还给出等价的“未成为飞行员”路线。此时零假设比例为 0.250.25,样本中未成为飞行员的有 9667=2996-67=29 人;使用连续性修正 28.528.5 计算右尾概率,结论应与解法一完全一致。

答题过程

展开

Let qq be the probability that an applicant does not go on to become a pilot.

H0:q=0.25,H1:q0.25.H_0:q=0.25, \qquad H_1:q\ne0.25.

There are 9667=2996-67=29 applicants in the sample who did not become pilots. Under H0H_0,

WB(96,0.25)N(24,18).W\sim\operatorname{B}(96,0.25) \approx N(24,18).

Using a continuity correction,

P(W29)P(Y>28.5)=P(Z>28.52418)=P(Z>1.06066)=0.1445\begin{align*} P(W\geqslant29) \approx&\,P(Y>28.5)\\[2mm] =&\,P\left( Z>\frac{28.5-24}{\sqrt{18}} \right)\\[2mm] =&\,P(Z>1.06066\ldots)\\[2mm] =&\,0.1445\ldots \end{align*}

Again, 0.1445>0.0250.1445>0.025, so we do not reject H0H_0. There is insufficient evidence at the 5% significance level against Jaymini’s claim.