Archaeologists often find only parts of ancient human remains. For example, they may find a small finger bone, called the metacarpal bone. Is it possible to predict the height of a human from the length of his or her metacarpal bone? To investigate, a researcher measures the heights and metacarpal lengths of 200 adults. In making the scatterplot, the researcher should

A. plot the height of the person on the horizontal axis.

B. plot the metacarpal length on the horizontal axis.

C. first determine if the heights of humans follow a Normal distribution.

D. use a plotting scale that makes the overall trend roughly linear.

A scatterplot is a graph that shows the points on a grid (coordinate plane), where horizontal axis contains the explanatory (independent variable) and the vertical axis contains the dependent variable (the variable to be predicted).

In this case, the variable that the researcher can "manipulate" is the metacarpal length, subject to the samples that the archaeologist find; thus, this is the explanatory variable and its length should be recorded on the horizontal axis.

The researcher wants to predict the height of a human, thus this is the dependent variable and it should be recorded on the vertical axis.

In conclusion, the researcher should:

1. Plot the height of the person on the vertical axis (opposed to the first choice)

2. Plot the metacarpal length on the horizontal axis (exactly what the second choice states).

The form of the scatter plot would tell the researcher the possible kind of the distribution, thus he/she cannot first determine if the heights of humans follow a normal distribution, neither must adjust the scale to make them follow a roughly linear trend. Thus, the choices C and D are absurd.