统计问题第64问：描述数据的均数与标准差

Question

Using a prospective cohort study design, researchers investigated the relation between slow walking speed and risk of death in older people. A sample of 3208 French men and women aged over 65 years living in the community were recruited in 1999. At baseline the sample had a mean age of 73.2 years with a standard deviation of 4.6 years.

Which of the following, if any, are true?

a) The mean age of the sample is a point estimate of the mean age of the population

b) The sample standard deviation is a measure of the spread of ages of the cohort members

c) The sample standard deviation is the square root of the sample variance

d) The sample variance has the same units as the sample standard deviation

提示：这是一道多选题。

Answer

Answers a, b, and c are true; d is false.

The mean age of all French people aged over 65 years living in the community during 1999 is known as a population parameter. We do not know the exact value of this population parameter, but we hope that the mean age of the sample is a good estimate of it. We refer to the sample mean age as a point estimate of the population parameter (answer a). Similarly, the sample standard deviation of age is a point estimate of the population standard deviation parameter.

The sample standard deviation of age measures the spread of ages in the cohort (answer b). In particular it provides a measure of how much on average the ages of the cohort members deviate about the sample mean age.

To derive the sample standard deviation, we calculate for each individual the difference between their age and the sample mean. These differences are then squared and summed across all individuals. We then find the mean of these squared deviations. The resulting quantity is called the sample variance. The square root of the sample variance is known as the sample standard deviation (answer c).

For some individuals the difference between their age and the sample mean would be positive, for others it would be negative. The sum of these differences would equal zero. Therefore, the differences were squared before adding them. The sum of the squared differences was not divided by the cohort size, but by the cohort size minus one. It has been shown that this provides a better sample estimate of the population variance parameter.

For each individual, we squared the difference between their age and the sample mean. Therefore, this difference had the units of years^2, as did the sample variance. The sample standard deviation, the square root of the sample variance, therefore had units of years—the same as the original observations of age (answer d is false).

这个是说方差是标准差的平方。所以两者级别是不一样的。

Next week’s question will discuss how we quantify the spread of observations in the sample using the sample mean and standard deviation.

所以答案是选择 a b c

每天学习一点，你会更强大！‍

‍