4.4 HISTOGRAMS WITH RESPONSE OVERLAY
A histogram is a graphical representation of a frequency distribution for a numerical variable. Figure 4.6 shows a histogram of the age variable with an overlay of response. Most customers range from, say, mid‐20s to about 60 years of age. So (non‐normalized) histograms are useful for seeing the distribution of the values of a numeric variable.
![Image described by caption and surrounding text.](https://static.packt-cdn.com/products/9781119526810/graphics/images/c04f006.gif)
Figure 4.6 Histogram from R of age with response overlay.
Again, however, it is somewhat difficult to ascertain any pattern in the response proportions. To better clarify these response proportions, we turn to a normalized histogram with response overlay, shown in Figure 4.7. Suddenly the response pattern becomes crystal clear. Customer response starts off high for 20‐year olds, gradually decreases, flattening out low for 30–60‐year olds, and rising sharply again for those over 60. So, the normalized histogram allows us to better distinguish these response patterns...