Which of the following statements about nonparametric tests is ...

[newyorker26]

Which of the following statements about nonparametric tests is false?
 
  A) We can safely employ nonparametric tests even when we know nothing at all about the population from which sample data are being drawn.
  B) We can conduct nonparametric tests with nominal data.
  C) We can conduct nonparametric tests with ordinal data.
  D) We can only conduct nonparametric tests with data of a higher order than nominal or ordinal.

Question 2

Why do we have so many different measures of central tendency? Are all really necessary or do they essentially provide the same information?
 
  What will be an ideal response?

[huda]

Answer to Question 1

D

Answer to Question 2

For certain types of well-behaved data, all the measures will provide very similar information. However, because much data is not well-behaved, it is often beneficial to look at a number of measures of central tendency and report the one that best describes the location and average of the data. Wherever possible, the mean is the preferred measure of location, because it uses all the data values. However, for certain types of data that contain severe outliers (like income data), the median is the preferable measure of central tendency because the outliers do not distort it.