Page 337 - Statistics for Dummies

Chapter 20: Ten Tips for the Statistically Savvy Sleuth
✓ Consider the units being represented by the height of the bars and what
the results mean in terms of those units. For example, are they show-
ing the total number of crimes, or the crime rate (also known as total
number of crimes per capita)?
✓ Evaluate the starting point of the axis where the counts (or percents) are
shown, and watch for the extremes: If the heights of the bars fluctuate from
200 to 300 but the counts axis starts at 0, the heights of the bars won’t look
much different. However, if the starting point on the counts axis is 200, you
are basically chopping off the bottoms of all the bars, and what differences
remain (ranging from 0 to 100) will look more dramatic than they should.
✓ Check out the range of values on the axis where the counts (or percents)
are shown. If the heights of the bars range from 6 to 108 but the axis
shows 0 to 500, the graph will have a great deal of white space and dif-
ferences in the bars become hard to distinguish. However, if the axis
goes from 5 to 110 with almost no breathing room, the bars will be
stretched to the limit, making differences between groups look larger
than they should. 321
Time charts
A time chart shows how a numerical variable changes over time (for exam-
ple, stock prices, car sales, or average temperature). Figure 20-1c is an exam-
ple of a time chart showing the percentage of yes voters from 2002 to 2010, in
2-year increments.
Here are some issues to watch for with time charts:
✓ Watch the scale on the vertical (quantity) axis as well as the horizontal
(timeline) axis; results can be made to look more or less dramatic than
they actually are by simply changing the scale.
✓ Take into account the units being portrayed by the chart and be sure
they are equitable for comparison over time; for example, are dollar
amounts being adjusted for inflation?
✓ Beware of people trying to explain why a trend is occurring without
additional statistics to back themselves up. A time chart generally
shows what is happening. Why it’s happening is another story!
✓ Watch for situations in which the time axis isn’t marked with equally
spaced jumps. This often happens when data are missing. For example,
the time axis may have equal spacing between 2001, 2002, 2005, 2006,
2008 when it should actually show empty spaces for the years in which
no data are available.

3/25/11 8:12 PM
29_9780470911082-ch20.indd 321 3/25/11 8:12 PM
29_9780470911082-ch20.indd 321