题目
Jian owns a large group of shops. She decides to visit a random sample of the shops to check if the stocktaking system is being used incorrectly.
(a) Suggest a suitable sampling frame for Jian to use.
(b) Identify the sampling units.
(c) Give one advantage and one disadvantage of taking a sample rather than a census.
Jian believes that the stocktaking system is being used incorrectly in 40% of the shops.
To investigate her belief, a random sample of 30 of the shops is taken.
(d) Using a 5% level of significance, find the critical region for a two-tailed test of Jian’s belief.
You should state the probability in each tail, which should each be as close as possible to 2.5%
The total number of shops, in the sample of 30, where the stocktaking system is being used incorrectly is 20
(e) Using the critical region from part (d), state what this suggests about Jian’s belief. Give a reason for your answer.
Jian introduces a new, simpler, stocktaking system to all the shops.
She takes a random sample of 150 shops and finds that in 47 of these shops the new stocktaking system is being used incorrectly.
(f) Using a suitable approximation, test, at the 5% level of significance, whether or not there is evidence that the proportion of shops where the stocktaking system is being used incorrectly is now less than 0.4
You should state your hypotheses and show your working clearly.
(Total for Question 3 is 15 marks)
题目中文翻译
简拥有一大批商店。她决定随机抽取一些商店作为样本,检查库存系统是否被错误使用。
(a) 建议简使用的一个合适的抽样框架。
(b) 识别抽样单位。
(c) 给出抽样调查而非普查的一个优点和一个缺点。
简认为 40% 的商店中库存系统被错误使用。
为了调查她的假设,随机抽取了 30 家商店作为样本。
(d) 在 5% 的显著性水平下,求简的假设的双尾检验的临界区域。
你应说明每个尾部的概率,每个尾部的概率应尽可能接近 2.5%
在 30 家商店的样本中,库存系统被错误使用的商店总数为 20
(e) 使用 (d) 中的临界区域,说明这对简的假设意味着什么。给出你答案的理由。
简向所有商店引入了一个新的、更简单的库存系统。
她随机抽取了 150 家商店作为样本,发现其中 47 家商店的新库存系统被错误使用。
(f) 使用合适的近似方法,在 5% 的显著性水平下,检验是否有证据表明库存系统被错误使用的商店比例现在低于 0.4
你应清楚地陈述你的假设并展示你的计算过程。
(第 3 题共 15 分)
解答
(a)
解法一
思路
展开
抽样框必须能够列出总体中的每一个抽样单位。本题总体是 Jian 所拥有的全部商店,因此应使用完整商店名录。
答题过程
展开
A suitable sampling frame is a complete list or database of all the shops owned by Jian.
(b)
解法一
思路
展开
抽样单位是实际被抽取并接受检查的个体;这里抽取的是商店,而不是库存系统或商店数量。
答题过程
展开
The sampling units are the individual shops.
(c)
解法一
思路
展开
与普查相比,样本只调查总体的一部分,所以通常更快、更便宜;但由于没有覆盖全部商店,结果可能不够准确或不具代表性。各给出一点即可。
答题过程
展开
An advantage is that taking a sample is quicker and cheaper than carrying out a census.
A disadvantage is that the sample may not be representative of all the shops, so the result may be less accurate.
(d)
解法一
思路
展开
在 Jian 的看法成立时,错误使用系统的商店数服从 。双尾检验把约 放在每一尾;分别寻找不超过 、且最接近它的下尾与上尾累积概率,再写出对应临界区域。
答题过程
展开
Let be the number of shops in which the system is used incorrectly. Under Jian’s belief,
For the lower tail,
For the upper tail,
Both tail probabilities are below and as close as possible to . Hence the critical region is
(e)
解法一
思路
展开
把观测值 20 与 (d) 的临界区域比较。由于 20 落在上尾临界区域,应拒绝比例为 的看法。
答题过程
展开
Since , the observed value lies in the critical region. Therefore, there is evidence that Jian’s belief that the proportion is is incorrect.
(f)
解法一
思路
展开
这是总体比例是否下降的单尾检验。先写出关于总体比例 的假设;在原假设下,二项分布的均值和方差分别为 60 与 36,且成功、失败的期望次数都足够大,所以可使用正态近似。对观测事件 使用连续性修正得到 ,求下尾概率并与 比较,最后用题目语境下结论。
答题过程
展开
Let be the proportion of shops in which the new stocktaking system is used incorrectly.
Under ,
Since
a normal approximation is suitable. Thus
Using a continuity correction,
Since , reject . There is sufficient evidence, at the significance level, that the proportion of shops in which the new stocktaking system is used incorrectly is now less than .