Microsoft Office Tutorials and References
In Depth Information
Scatter (XY) Charts
Scatter (XY) Charts
Scatter charts (also known as XY charts) are fascinating, versatile, and often misunderstood.
A scatter chart is made up of two numeric data series, plotted in pairs on the horizontal and
vertical axes (which are also known respectively as the x-axis and y-axis, thus explaining the
origin of the XY name). You can use a scatter chart in place of a line chart when data points
on the horizontal access aren’t linear; the visual result is similar, but without the distortion
caused by irregular spacing of data points. One common use of a scatter chart is to identify
clusters of similar data in a nonlinear set. In Figure 13-13, for example, we’ve created this
chart type by plotting survey data for 16 companies, with customer satisfaction ratings on
the vertical axis and price (from high to low) on the horizontal axis.
Figure 13-13 Each dot represents a pair of survey results for a company. From this data, we can
conclude that consumers are less satisfied with the high-priced options in our data series.
You’ll notice in this scatter chart that we deliberately hid the values on both axes. The
numbers themselves can be on any scale you create. It’s the position of the data in this chart
that matters most. The data points in the top right quadrant represent the best
combination of value and customer satisfaction.
To divide the graph into quadrants, we used the Data Series Options dialog box to define
the Major Gridline interval for each axis as equal to exactly half the distance between the
lowest and highest points on that axis (50 on a scale of 0 to 100). From the Gridlines menu
on the Layout tab, we made Major Gridlines visible on both the horizontal and vertical axis,
resulting in a neatly divided rectangle.
You can add smooth lines or straight lines to a scatter chart, with or without markers for
each data point. The Scatter With Smooth Lines chart type can produce interesting results
when you’re trying to show a pattern in data that is not uniformly distributed.