Chapter 3 – Part 3

Chapter 3 – Part 3

3.8 Tables

Tables are the simplest form of data display, yet creating an effective table is an acquired skill.  The arrangement of numbers in rows and columns can enhance their meaning so that it can be understood at a glance.

Tips for Effective Tables

1. Keep the table simple and consistent with its purpose. In a slide presentation, the main point of the table should be clear to a reader within 10 seconds. If not, then consider breaking the table up to make it more clear.;

1. Display the data to be compared in columns rather than in rows. Research shows that people find it easier to compare across rather than down.;
2. For presentation purposes, round off to three or four significant digits.  People mentally round numbers anyway. (Exception is when accounting requires more precision.);
3. Physical table layout should guide the eye towards the comparison you wish to emphasize. Spaces or shading can be used to separate columns or rows., use lines sparingly.;
4. Row and column headings should be simple yet descriptive.
5. Within a column, use a consistent number of decimal digits. Right-justify or decimal align the data unless all field widths are the same within the column.

Pivot Tables

An Excel feature that allows easy editing, such as dragging variable names from a list and into a table.

 
3.9 Pie Charts - Though many data analysts consider the table or bar chart to be a better choice than a pie chart for a number of reasons, it is still used regularly in annual reports and in the general press (like USA Today, the Wall Street Journal, etc.)   For this reason, it's necessary to understand them and to understand how they can be misused.  
 
An Oft-Abused Chart

1. A pie chart can only convey a general idea of the data because it's hard to assess the areas precisely.  
2. Pie Charts should have only a few slices and the slices should be labeled with data values or percents.  
3. The only correct use of a pie chart is to portray data which sum to a total.  (e.g. pieces of a whole, like a pie itself.)

Pie Chart Options
Exploded (pieces set apart by gaps) and 3-D pie charts add visual interest.  (See Figures 3.31 and 3.32 for examples). But those forms make the sizes of the pie slices even harder to assess.

3.10 Deceptive Graphs -

As a consumer, you will need to be aware of deceptive uses of graphs.  Sometimes they get that way through ignorance.  At other times, they are deliberately crafted to deceive.

Error 1:   A Nonzero Origin   Beginning the chart or data at some number other than zero will often exaggerate the trend shown in the chart. Measured distances will not match the stated values or axis demarcations. (See comparison on p. 97 for illustration.) The accounting profession is particularly aggressive in enforcing this rule.  

Although zero origins are preferred, sometimes a nonzero origin is needed to show sufficient detail.
 
Error 2: Elastic Graph Proportions   Keep your chart in an aspect ratio (proportion of width to height) of less than 2.0.   (A ratio of 1.0 is a square)  Ex. In a time series graph, too short an X-axis in relation to the Y-axis will make a profit curve appear steep.  But too wide an x-axis in relation to a y-axis can be used to mask losses. (See figures on p.98 for comparison.)
 
Error 3: Dramatic Tiles and Distracting Pictures  A dramatic title can be used to distract from or influence the impact of the graph. (ex. "Criminals on a Spree" or "Deficit Swamps Economy".)  Sometimes a title attempts to draw a conclusion for the reader. (E.g. "Inflation Wipes Out Savings".)  Images can be used to emotionally influence or distract from the information in the graph. (e.g. a gasoline price chart atop a photo of Middle East warfare.)
 
Error 4: 3-D or Rotated Graphs  By making a graph 3-D or rotating it in a certain way, the artist can make a trend appear to be more drastic than it is.  The figure on p. 98 shows medical school applications between 2002 and 2006 portrayed in a way that makes them appear to have "soared" when they really only increased from 33,625 to 39,108. The graph distorting the change uses two errors to do so: a nonzero origin and a rotated graph.
 
Error 5:  Unclear Definitions or Scales  Missing or unclear units of measurement (Dollars? Percents?) can render a chart useless. If it's a percent, we must state clearly a percentage of what? Without tick marks under the axis, the reader cannot identify individual data values.
 
Error 6:  Vague Sources   Large federal agencies and corporations employ thousands of people and issue hundred of reports per year. Vague sources like "Department of Commerce" may indicate that the author lost the citation, didn't know the data course or mixed data from several sources.
 
Error 7: Complex Graphs  Complicated visual displays make the reader work harder. Keep your main objective in mind.  Apply the 10 second rule to graphs. (Reader must be able to discern the purpose in 10 seconds.)
 
Error 8:  Gratuitous Effects  Slide shows often use color and special effects to attract attention. But if slathered on too thickly, the audience will lose focus.
 
Error 9: Estimated Data    In the interests of including the "latest" figures, the last few data points are often estimated. Maybe some data were missing or incompatible with the trend the author wanted to show, so they were estimated.  In any case, estimated points should be kept to a minimum and always noted.
 
Error 10: Area Trick  One of the most deceptive tricks is to widen the bars in a graph along with height, making the growth appear to be greater than it is. (Or vice versa when displaying decreasing data.)  See the figure on the bottom of page 99 (physician salaries) to see an example of this kind of distortion.