91Ó°ÊÓ

The best reason for choosing a stemplot rather than a histogram to display the distribution of a quantitative variable?

a. Stemplots allow you to split stems; histograms don’t.

b. Stemplots allow you to see the values of individual observations.

c. Stemplots are better for displaying very large sets of data.

d. Stemplots never require rounding of values.

e. Stemplots make it easier to determine the shape of a distribution.

(a) The statement is false because you are capable of doing something similar to splitting stems in a histogram. For example, you can "split stems" in a histogram by dividing each class into two equal-sized classes.

(b) The statement is correct because seeing the values of individual observations in a graph would be useful. A stemplot, for example, could be used to display the actual data set as well as to determine the actual values of outliers or to compute numerical values such as the median, first quartile, and third quartile.

(c) The statement is incorrect because stemplots are ineffective at displaying large data sets. For example, if a data set contains 1000 observations, the stemplot will be extremely large, making it difficult to observe the data set's actual distribution.

(d) The statement is incorrect because stemplots require us to round values on occasion. For example, if all data values are integers with the exception of one decimal value, it is best to round the one decimal value before creating a stemplot (as most integer data values would otherwise have their own stems, which will make the stemplot extremely large).

(e) The statement is incorrect because the distribution in a stemplot is not always easy to determine (especially when the number of observations is very large).

As a result:

(b) We can see the values of individual observations using Stemplot.