Which statement about the median is true?

Which statement about the median is true?

• A. Extreme scores have a large effect on the median.
• B. Extreme scores do not have a large effect on the median.
• C. It cannot be calculated easily.
• D. It is always the best measure for every data type.

Answer: (B)

The main idea is that the median is a robust measure of central tendency. It represents the middle value of a data set when the values are arranged in order, so it depends on the position of data rather than their exact magnitudes. Because of that, extreme scores at the high or low end don’t substantially shift the middle value unless there are many extreme scores enough to change which observation sits in the middle. This is why extreme scores have only a small effect on the median, unlike the mean, which uses every value and can be pulled toward outliers. The statement is true for that reason. The other options don’t fit: calculating the median is straightforward (order the data and take the middle value, or average the two middle values if the sample is even), and the median isn’t always the best measure for every data type (for nominal data you can’t compute a median, and in some situations the mean or mode may be more appropriate).

The main idea is that the median is a robust measure of central tendency. It represents the middle value of a data set when the values are arranged in order, so it depends on the position of data rather than their exact magnitudes. Because of that, extreme scores at the high or low end don’t substantially shift the middle value unless there are many extreme scores enough to change which observation sits in the middle. This is why extreme scores have only a small effect on the median, unlike the mean, which uses every value and can be pulled toward outliers. The statement is true for that reason. The other options don’t fit: calculating the median is straightforward (order the data and take the middle value, or average the two middle values if the sample is even), and the median isn’t always the best measure for every data type (for nominal data you can’t compute a median, and in some situations the mean or mode may be more appropriate).