Author Question: Why do statisticians sometimes use inferential statistics to draw conclusions about a population? In ... (Read 104 times)

lak · **on:** Jul 25, 2018

Why do statisticians sometimes use inferential statistics to draw conclusions about a population? In what situations might a statistician draw conclusions about a population using descriptive statistics only?

What will be an ideal response?

Question 2

The variables height and weight could reasonably be expected to have a positive linear correlation coefficient, since taller people tend to be heavier, on average, than shorter people.

Give an example of a pair of variables which you would expect to have a negative linear correlation coefficient and explain why. Then give an example of a pair of variables whose linear correlation coefficient is likely to be close to zero.

juicepod · **Reply #1 on:** Jul 25, 2018

Answer to Question 1

If a population is large, it may be too expensive and time-consuming to interview every member of the population. In such cases, a sample is drawn from the population, and based on the information in the sample, conclusions are drawn about the population; in other words, inferential statistics are used. If the population is relatively small, it may be realistic to interview every member of the population, in which case only descriptive statistics are needed.

Answer to Question 2

Answers will vary. One possible answer: The variables age of Corvette and price of Corvette could reasonably be expected to have a negative linear correlation, since cars tend to depreciate in value over time. The variables height and degree of color blindness could reasonably be expected to have no linear correlation (and thus a linear correlation coefficient close to zero), since there is no apparent relationship between the two variables.