题目
A supermarket sells apples in trays. Each tray contains 10 apples.
Trays are inspected for rotten apples daily. The following table shows the numbers of rotten apples in a random sample of 150 trays on one day.
| Number of rotten apples | 0 | 1 | 2 | 3 | 4 | 5 | 6 |
|---|---|---|---|---|---|---|---|
| Frequency | 25 | 47 | 37 | 24 | 8 | 6 | 3 |
The manager believes that these data can be modelled by a binomial distribution B(10, p).
(a) Use these data to estimate the value of the parameter p for this model. Give your answer to 2 decimal places.
The manager uses the estimated value of p to 2 decimal places and calculates expected frequencies to 2 decimal places as follows.
| Number of rotten apples | 0 | 1 | 2 | 3 | 4 | 5 | |
|---|---|---|---|---|---|---|---|
| Expected frequency | 20.62 | 45.26 | 44.71 | 26.17 | 10.05 | r | s |
(b) Show that the value of r is 2.65 to 2 decimal places.
(c) Find the value of s to 2 decimal places.
The manager says that the cells for 4, 5 and need to be combined into a single cell.
(d) Explain why it is necessary for the manager to do this.
The value of χ^2 for the first 4 frequencies given in each table is 2.51.
(e) Using a 10% significance level, test whether or not a binomial distribution is a suitable model. You must state your hypotheses, test statistic and the critical value used.
题目中文翻译
一家超市出售装有 10 个苹果的苹果筐。
每天都会检查这些苹果筐中的腐烂苹果数。 下表给出了某一天随机抽取的 150 个苹果筐中腐烂苹果的数量。
| 腐烂苹果数 | 0 | 1 | 2 | 3 | 4 | 5 | |
|---|---|---|---|---|---|---|---|
| 频数 | 25 | 47 | 37 | 24 | 8 | 6 | 3 |
经理认为这些数据可以用二项分布 B(10, p) 来建模。
(a) 用这些数据估计参数 p 的值,答案保留 2 位小数。
经理将 p 估计到 2 位小数后,计算出如下期望频数。
| 腐烂苹果数 | 0 | 1 | 2 | 3 | 4 | 5 | 6 |
|---|---|---|---|---|---|---|---|
| 期望频数 | 20.62 | 45.26 | 44.71 | 26.17 | 10.05 | r | s |
(b) 证明 r 的值为 2.65,保留到 2 位小数。
(c) 求 s 的值,保留到 2 位小数。
经理说必须把 4、5 和至少 6 的格子合并成一个格子。
(d) 解释为什么有必要这样做。
对两张表中前 4 个频数计算得到的 χ^2 值为 2.51。
(e) 在 10% 的显著性水平下,检验二项分布是否适合作为模型。你必须写出原假设、检验统计量和所用临界值。
解答
(a)
解法一
思路
展开
先由频数表求样本中腐烂苹果的总数,再除以苹果总数 ,所得样本比例就是 的估计值。
答题过程
展开
The total number of rotten apples is
There are apples in total. Therefore,
(b)
解法一
思路
展开
使用题目指定的两位小数估计值 。每筐恰有 5 个腐烂苹果的概率由二项分布给出,再乘以 150 筐便得到期望频数 。
答题过程
展开
Let be the number of rotten apples in a tray. Using the estimated value of ,
Hence
(c)
解法一
思路
展开
各格期望频数之和应等于总频数 150。按照题目表格已经保留两位小数的数值,从 150 中减去其余各格即可得到 ,这也与官方评分方案采用的数值一致。
答题过程
展开
The expected frequencies must sum to 150. Therefore, using the expected frequencies in the table,
(d)
解法一
思路
展开
卡方拟合优度检验要求每一格的期望频数至少为 5,而腐烂苹果数为 5 和至少为 6 的两格都不满足此条件,所以必须与相邻格合并。
答题过程
展开
The expected frequencies for 5 and at least 6 rotten apples are both less than 5. The cells must therefore be combined with the cell for 4 rotten apples so that the expected frequency in the combined cell is at least 5.
(e)
解法一
思路
展开
合并后三格的观察频数为 17、期望频数为 13.24。把这一格的卡方贡献加到题目已给出的 2.51,再根据合并后的 5 格及已估计的一个参数确定自由度为 3,最后与 临界值比较。
答题过程
展开
Let
For the combined cell,
Therefore, the test statistic is
There are 5 cells after combining, and one parameter has been estimated. Hence
At the significance level, the critical value is . Since
is not rejected. There is insufficient evidence that the binomial distribution is not a suitable model, so a binomial distribution is suitable for these data.