Charts and Graphs: Visualizing Data
Sticking to the Point: One Message, One Slide, One Chart
We've emphasized the "One Slide, One Message" concept in previous chapters. This concept is even more important when creating charts for your slides. A chart which tries to say or do too much will inevitably confuse and mislead the audience. When creating chart slides remember this guideline: "One Slide, One Message, One Chart!"
Of course, there will be exceptions; but it is a good starting point, and you should exhaust all your single-chart options before deciding to use multiple charts on a slide.
Creating Message-Driven Charts and Graphs
The most effective charts, like other slides, are those which are message-driven. Each chart slide must have a single, well-defined message to present to the audience. From that message, you develop a chart which contains only the data necessary to support the message. For example, a slide with the message:"Midwestern Division Sales Volume is Rising" could have a bar or line chart showing the rise in sales volume for the Midwest region. Any other information, such as sales for other regions, is unnecessary and should be left out.
Figure 10.1: Sales Volume Chart
On the other hand, if the message of the chart is: "Midwestern Sales Volume Is Growing Faster than in the East" then the extra information would be appropriate.
Always make sure that the title of any chart slide clearly describes the message of the chart. The audience may draw their own conclusions from your graphics unless you tell them what they're supposed to see. The combination of the message in the title and the graphics in the chart will make your message much more memorable.
Basic Chart and Graph Layout
All charts are visual representations of numerical data. If you can't measure something, you can't chart it. The numerical data is represented on the chart by a graphic item; a rectangle in column and bar charts, a portion of a circle in a pie chart, lines in line charts, etc. Each graphic item in a chart represents two pieces of information:
- The name of a measurable item (called a category) which is identified on the chart by a label, and
- A quantity associated with the item (called a data point) which is plotted on the chart as a value.
The graphics in a chart are created on a plotting area which defines the area of the chart and provides a framework for customizing the look of your charts. There are two basic types of plotting areas: rectangular and circular. The circular plotting area is mainly used for pie charts. All of the other chart types described in this chapter use a rectangular plotting area.
Rectangular Plotting Areas
The design and layout of any chart has more to do with the way you choose to display the details of a chart than with the graphics you use to represent your data. For example, once you've set up a basic layout for a column chart you can use it as a model for the other chart types based on a rectangular plotting area such as bar, line, area, scatter and pictograph charts. In fact, you should develop a standard rectangular plotting area to act as a guide for your rectangular charts.
The basic elements of a rectangular plotting area include:
- The Chart Frame; which defines the boundaries of the plotting area.
- The Scale; which acts as a ruler for the audience to measure the relative size or position of the plotted values.
- The Labels; which name the items being plotted.
- The Grid or Tickmarks; which act as guidelines and markers for the scale.
Rectangular Chart Frames
The basic chart frame consists of a box which defines the area in which your data will be plotted. You can define this box in several ways, even make it invisible if necessary, but the chart frame defines the size of your chart and its placement in the slide frame.
Figure 10.2: Basic elements of a rectangular plotting area.
A full frame is a filled or unfilled box which surrounds the plotting area. For presentation purposes, this is the best all-round choice, since it will support many different chart types and is easily adaptable to a wide variety of scaling and grid options. Filling a full chart frame with a dark or black background also provides the greatest amount of contrast for your data.
A half frame is unfilled and defines only the left and bottom portion of the frame. This is the traditional way most charts are framed, and consequently it has an old-fashioned look to it. This frame type is useful with most kinds of charts which need a scale, but where an accurate grid is unnecessary. This type of frame is not recommended for scatter charts, which require a full grid to be readable.
A baseline only frame shows only the bottom edge of the frame with a single line. Because only labels and categories are shown, this type of frame should only be used with column, bar, area and pictograph charts which originate at the baseline and include visible data values. In fact, if your presentation consists only of simple bar and column charts, a baseline only is ideal, since it is very unobtrusive and you won't be cluttering up your presentation with unnecessary graphics. Do not use this type of frame for scatter or line charts, since they will suffer from a lack of visual connection between the baseline and the graphics representing the data.
The guidelines for using no frame on your charts are similar to those for baseline-only. However, you must only use this style where your chart graphics originate at the baseline, since the frame's baseline must be implied by the presence of your graphics.
Figure 10.3: Chart Frames
Rectangular Chart Scales
The values in a chart are plotted against a grid or scale which allows you to measure the relative size of the items in the chart. The basic grid for most charts consists of an X axis and a Y axis. The X axis extends from left to right, the Y axis extends from bottom to top.
Like a ruler, a chart scale allows you to measure the relative size or position of your data. Your chart scale should be located along the Y axis for column, line and area charts. Bar charts use a scale along the X axis. Pictograph and combination charts can use either axis, depending on your data. Scatter charts require a scale along both axes.
Figure 10.4: Scales on X, Y and Both Axes
Most presentation graphics, charting and spreadsheet programs will automatically select the proper scale for the chart type you choose, allowing you to customize it as needed.. While most automatic scales will work perfectly in a presentation, often they are created by the software with print applications in mind and may require some adjustment to be useful in slides. Here are some guidelines for setting up a chart scale which will be readable and enhance the message of your chart.
Keep Scale Increments Evenly Spaced
The spacing and values of your scale should be evenly distributed visually and numerically to give a true picture of the data you are presenting. The values of your scale do not have to be in increments of fives or tens however. Any set of evenly-spaced values can be used to create a valid scale. For example, a scale of zero to sixty can be designed in six different ways, using from four to twelve even steps:
STEP INSERT Pg 198
Special scale increments can be very useful when creating charts which require two separate scales. For example, if you wanted to compare annual sales in dollars to annual shipments in tons, you may need separate scales to make a chart which fits your message. The following example shows a dual-scale chart created from two separate charts. Note that the "Tons" scale has been altered to use the same number of steps as the "$MM" scale to simplify the chart.
Figure 10.5: Dual Scale Chart
Keep Scale Figures Readable
The text size of your scale figures should be at least half the size of the average body text used in your slides. For example, if the body text in your text slides ranges from 24 to 30 points, the scale figures should be at least 12 to 16 points. Of course, larger type may cause the scale figures to be cramped when used in scales with a large number of steps.
Limit The Number of Scale Steps
If creating a readable text size causes your scale figures to be too cramped, then you should decrease the number of scale steps to create more space between the scale figures. To make room for readable scale figures, use no more than ten steps in any scale on a slide. A scale should also contain a minimum of four steps to be a useful ruler for the audience.
Figure 10.6: Scale Steps
Figure 10.7: Scale steps should range between four and ten
Grids & Tickmarks
Grids and tickmarks are used to define exact divisions of the frame into sections which correspond to the scale. A grid line is one which extends across the entire chart frame. A tickmark is a shorter line which indicates a scale division. Grids and tickmarks can also be combined.
There are dozens of variations, and most of them are used in conjunction with a full frame. Half frame, baseline and no frame support only a few variations.
Figure 10.8: Grid and tick mark variations for full-frame charts
Figure 10.9: Grid and tick mark variations for half-frame, baseline-only and no-frame charts
Figure 10.10: Baseline Only + Staff Grid
One of the most useful variations is the "staff grid" (so called because it looks like a musical staff) which can be used as a variation on the baseline only frame. You can create column charts using only a baseline and then switch to a staff grid for line and area charts which need a Y axis scale.
Axis Titles & Labels, Data Values and Callouts
There are several kinds of text other than scale values used in charts. Axis titles, axis labels, data values and callouts all contribute to the clarity of your charts and graphs. Type size for all of these should be at least as large as your scale figures.
An axis title acts to identify the units of a scale or a larger category describing axis labels. Axis titles especially clarify scales which have very large values. For example, a column chart scale which ranges from zero to 10 million dollars could be very difficult to read with so many zeros in the scale. Adding a "$ Millions" axis title makes the scale much easier to read and comprehend.
Figure 10.11: Axis titles clarify the scale
Keep your axis titles short and to the point; don't use "Thousands of Dollars" where "$000's" will do. You can use abbreviations such as "M" or "MM" for "millions", "K" or "000's" for "thousands", and "/Yr" for "per year". A good dictionary of abbreviations can save a great deal of space on your charts and graphs.
An axis does not always act as a scale with numerical values. When plotting column charts, for example, an axis label is usually placed under each graphic on the X axis to identify the category being plotted. Labels should be spaced evenly along the axis, the same as scale values. There are two ways of setting up axis labels: object-aligned and grid-aligned.
Object-aligned labels are usually associated with a graphic object such as a bar or column. This type of label should be used with bar, column and pictograph charts. An object-aligned label does not need any sort of grid or tick mark.
Grid-aligned labels should be used with line and area charts when it is necessary to line up labels at the bottom of the chart frame with data points some distance away.
Figure 10.12: Axis Labels
Data values are the text used when you wish to show the actual value of a chart item, such as column values or pie percentages. Data values should always be placed either inside or directly adjacent to the corresponding chart graphic.
Figure 10.13: Data Values
One important thing to remember about data values is to keep them short. Large numbers should be rounded to three or four significant figures. Use abbreviations. Remember: Key the accuracy of your data values to what's being said in the speech. If the speaker is saying:
"Acme profits soared to nearly 31 million dollars in 1993."
then your data value on the accompanying chart should read "$30.6 M", not "$30,609,784.23".
A callout is descriptive text placed in a chart which is not data or part of the axes. The most common use of a callout is to identify plotted lines when there is more than one in a line chart. Other callouts identify categories in stacked column charts. A callout may also act as a notation for a particular data point. Keep callouts as simple as possible. Remember, you're creating a chart to take the place of text; don't clutter it up with more words than you need.
Figure 10.14: Callouts
When there is not enough room on a chart for a callout, or if you need to identify several categories within a single chart, it may be necessary to use a legend. A legend is a visual key which allows you to identify chart segments when labels and callouts are inadequate. Most charts complicated enough to require a legend will be too complicated for a slide. Never use a legend where you can use labels or callouts.
The main use for a legend is when you have multiple sets of data points in a single chart. Legends can be hard to read for the audience, who have to switch their attention back and forth from chart segments to the legend to make sense of the graph.
The most important part of a chart is the plotted data, not the legend. Place your legends inside the plotting area. Legends placed outside the plotting area will usually cause the actual chart to be much smaller, unnecessarily giving too much weight to the legend.
Figure 10.15: Legends
Some presentation graphics programs will create a legend even for a chart with a single set of data points; delete the legend and name the set in your slide title or in a subtitle.
Selecting the Correct Chart Type
Your message and the structure of the data you need to present will affect your charting choices. Different messages and data are better suited to certain types of charts and graphs. To help you determine the right type of chart for your message, here is a guide to chart types, with useful variations and enhancements.
The column chart is one of the most common and useful kinds of charts. Column charts are best suited to displaying the relative size or volume of tangible, physical things. For example, use them to compare weights, dollar values, and object quantities.
It is common to read passing time from left to right, so the column chart is the ideal method for plotting the change in size or volume of a single item over a period of time. For example, Figure 10.14 shows the volume of apples shipped over several years.
Figure 10.16: Column charts can show time or categories
This type of column chart actually has two scales. The Y axis scale (vertical) measures the data values; the X axis scale (horizontal) measures the passing of time.
You can also use a column chart to plot the relative quantities of several different items at a particular time. The main limitation for this use is that X axis labels can often be quite long compared to date or time labels. Bar charts are usually more appropriate for this type of data. If your X axis labels are short, though, things can work out fine. For example, Figure 10.15 compares the volume of various types of fruits shipped in a particular year.
Figure 10.17: Columns should be evenly spaced
There are dozens of column chart types, many used for very specialized purposes. The design and layout for all of them is similar. We'll be discussing the main types used in general business presentations.
A column chart is made of rectangles which represent an item or series of items. Labels placed under the baseline along the X axis identify each column. The Y axis consists of a scale to measure the relative height of each column.
Your columns should be evenly spaced and should use most of the width of the plotting area. The space between columns should be no less than 25% and no more than 100% of the column width. One exception to this rule is if you have a large number of closely-spaced narrow columns; in this case you are better off having the columns touch (similar to a histogram, see below).
Your tallest column should also use as much of the full height of the plotting area as possible. For example, if the largest data value in your chart is 72, make your scale 0-75 or 0-80. In any case, your columns should use at least 75% of your plotting area height.
You can use virtually any grid/tickmark combination with a column chart; the complexity of the data should guide your decision. A single set of columns actually needs little in the way of a scale or grid, especially if you are showing data values on top of the column. Other variations on the column chart may require more complicated scales and grids. Always strive to use the minimum amount of grid necessary for clear, accurate communication.
You can place data values above your columns, eliminating the need for a grid entirely. By including data values, all you really need is a baseline. Remember to keep your data values short, however; they should not exceed the width of a column. If necessary, round off data values to three or four significant figures.
The column chart is one of the most versatile chart types. It can be adapted to provide a wide variety of information, even acting as a pie chart substitute.
A stacked column is designed to show how a series of components adds up to a whole. Each column is composed of segments which represent a series of data values which add up to a total value. Limit yourself to a maximum of six segments per column. If you have more than six series of data, combine leftover data into a single data value and label it "Other". The most important data series should be placed on the baseline, since only those segments can be easily compared.
There are two ways to determine the height of each column, and you should be aware of how each system works.
Cumulative segments combine to add up to the total height of the column. For example, three cumulative segments of 20, 25 and 30 add up to a total column height of 75.
Incremental Values indicate discrete steps in the height of the column, but do not indicate the height of each segment (although segment height can be determined with a little arithmetic). For example, a column with a height of 75 may have two incremental values at 20 and 45.
Figure 10.18: Cumulative/Incremental
While the graphic on both charts is identical, the data values used to define the graphic are different, and should be treated differently. Cumulative segment values should be placed in the center of their respective column segment. Incremental values should be placed on top of their respective column segment. Category labels should align with the data value text. In both cases, the total should be placed above the entire column.
Don't use a legend as a key to your segments. Place callouts to the right of the last column aligning with and identifying the corresponding segments.
Figure 10.20: Stacked Segment Labels
A grouped column chart is designed to compare the components of several data value series in relationship to each other. Each group is composed of several columns, each of which represents a different data series.
Grouped column charts are very easy to overdo. Keep them as simple as possible. Column groups should have no more than three to four columns or data series.
Avoid putting data values at the top of the columns, since too many figures will be hard to read. Use a good, clear scale and grid instead.
Groups can be arranged either with the columns touching, or slightly overlapping. Columns should not overlap more than one third of their total width. In addition, only use overlapping columns if each set is consistently larger or smaller than the others, forming a stair-step effect. The tallest column should be in the back, with shorter columns ranked in front of it.
Figure 10.21: Grouped Columns
You will normally need a legend or callouts to identify the series being plotted. Only use callouts if there is a substantially clear area somewhere in the chart for the text, and if there are no other bars to get in the way. A legend can be set off in a box within the plotting area to help isolate it from the columns and text in the chart.
Histograms are a special category of column chart which use a variable scale for both axes. In the other examples you've seen in this section, the columns have been evenly spaced and represent specific points in time or individual items. The width of each bar was irrelevant to the data being presented.
In a histogram the width of a column also carries information. In a typical histogram, the Y scale is the same as any other column chart measuring the relative value of the columns. The X axis, though, acts as another scale in which the width represents a span of time or other variable. In Figure 10.22, the column width shows the age range of the sample data, the height shows the percentage of rock and roll fans in that age range.
Figure 10.22: Histogram
The columns in a histogram should always touch, since they act as a continuous data measurement along the X axis. The labels are aligned with the center of the column they identify, the same as for other column charts. You can also use a graduated scale along the X axis similar to one used in a bar chart.
Column Chart Enhancements
The most common enhancement used with any chart is to add depth to your graphics, giving them a more solid and substantial feel. There are two common method of adding depth to column charts.
True Perspective depth is designed to give the maximum realism to the look of columns, making them appear as if they are true solid objects floating in the slide space. The column surface is extruded backwards into the field of the slide and scaled, making it appear farther away than the front. This type of depth is not usually supported in most presentation programs, and is difficult to get right, since the auidience is usually viewing the graphic from different angles,diminishing the effect.In addition, true perspective does not work well with any type of Y axis scale, since the depth distorts the columns' relationship with the scale.
Figure 10.23: Perspective Columns
Oblique depth is a simulation of perspective and depth which is the type commonly used in most presentation and charting software. In oblique depth, the surface of the column is extruded, but the back is not scaled. The most common method of oblique depth extrudes the column face up and to the right. Oblique depth is simpler to create, and it is compatible with Y axis scales. However, when using depth with a scale,it is very easy to misrepresent data. Let's look at an example using an oblique depth column chart common to many presentation programs from Harvard Graphics to Aldus Persuasion.
Figure 10.24: Oblique 3D Columns
In the left hand chart of Figure 10.24, the 1992 Annual Sales column could be interpreted as reading 69 or 75 because of the way the scale is set up. Because the scale figures are in the same plane as the front of the columns, the natural tendency is to measure the front of the column against the scale, for a data value of 69. Actually, the grid and scale are projected to the back plane of the columns, where the height of the 1992 column reads as 75. In this way, it is very easy to misinterpret data on a chart with depth. The right hand chart, however, moves the scale figures to the back plane, improving the readability of the chart.
Another alternative enabling more accuracy is the Pseudo-3D chart. In this chart, the chart frame acts as a window, and the depth is shown extruding down and to the right. The scale and grid refer to the front of the column, and the top of the column reflects a true data value in relation to the grid.
Figure 10.25: Pseudo-3D Columns
If you need to emphasize the change between segments of a stacked column chart, connecting lines can make your relationships clearer. You can also fill in the connecting areas, making a variation similar to an area chart. Filled connectors should be a darker shade of your segment color so they are not mistaken for actual data values.
Creating Column Chart Builds and Reveals
Builds on column charts can be created along the X or Y axis. A regular column chart build reveals each column or group of columns in sequence from left to right along the X axis. You can also highlight and tone back the columns as they are revealed.
Figure 10.26: Column Chart Build
On a stacked column chart you can also reveal each category in sequence from bottom to top along the Y axis. This variation can be very effective in clarifying complex stacked columns, allowing the speaker to address each category separately.
Figure 10.27: Stacked Column Chart Build
The bar chart's purpose is very similar to the column chart. It is used to compare the relative size of items. The bar chart is a rotated version of the column chart, with the vertical Y axis becoming the baseline and the horizontal X axis becoming the scale. This arrangement also makes it possible to have longer labels for bars than would be practical on a column chart.
Bar charts are particularly good for showing the rankings of several items. For example, Figure 10.30 plots monthly sales data for several salesmen, but also ranks them from top to bottom according to sales. Notice that the longer names would be too wide to use in a column chart, but they look very natural on a bar chart.
Figure 10.28: Bar Chart
Other arrangements of the bars in a bar chart are often awkward and should be avoided. Take a look at the same chart organized in alphabetical order and you'll see that the chart looks much more disorganized and less effective.
Figure 10.29: Disorganized Bar Chart
There are dozens of bar chart types, many used for very specialized purposes. The design and layout for all of them is similar. We'll be discussing the main types used in general business presentations.
Like a column chart, a bar chart is made of rectangles which represent an item or series of items. The axes of a bar chart are the reverse of a column chart. Labels placed to the left of a baseline along the Y axis identify each bar. The X axis consists of a scale to measure the relative length of each bar. Traditionally, the scale for a bar chart is placed at the top of the plotting area; but this position puts the scale in a very strong position on a slide, directly under the title. Such a strong position for the scale would be distracting on a slide, so it is usually better to place your scale at the bottom of the plotting area.
Figure 10.30: Bar Chart Scales
Your bars should be evenly spaced and should use most of the height of the plotting area. The space between bars should be no less than 25% and no more than 100% of the bar thickness. One exception to this rule is if you have a large number of closely-spaced narrow bars; in this case you are better off having the bars touch (similar to a column chart histogram).
Your longest bar should also use as much of the full width of the plotting area as possible. In any case, your bars should use at least 75% of your plotting area width. Place your data values either inside or outside the end of the bar. Data values inside the bar emphasize the graphic over the text; data values outside the bar put more (but not all) of the emphasis on the text.
Figure 10.31: Bar Chart Data Values
Grid and tickmark choices are the same as for column charts, although a full grid along the X axis is definitely a help in judging the length of long bars. If you are showing data values, you may only need a baseline on the Y axis.
Excessive bar labels will diminish the impact of your chart, so keep them brief. If your bar labels are very long, you should break them into two or more lines.
The basic guidelines for stacked, grouped and histogram bar charts are the same as for their equivalent column charts, so we won't repeat them. However there are some variations on bar charts which are unique.
Paired Bar Charts
A paired bar chart enables you to plot two sets of bars for the same item, using entirely different scales and grids. It also allows you to make relative comparisons based on different criteria.
A paired bar consists of two separate plotting areas; the same set of labels is used for both Y axes, with the labels placed between the plotting areas. In the left plotting area the X axis scale reads from right to left. In the right plotting area the X axis scale reads from left to right. The scale for each plotting area can be different, but it is usually more attractive to use the same number of steps. In Figure 10.36, note that the left scale uses increments of three years so the number of steps matches the right scale.
Figure 10.32: Paired Bar Chart
You will need to place a title above each plotting area to identify the item being charted. Scales should be at the bottom of the plotting area with scale titles placed below the scale figures.
Another excellent use of bar charts is to demonstrate a series of inter-related processes over a period of time. Steps in a process are plotted as bars against a timeline on the X axis of the plotting area.
When applied to a manufacturing or construction process, this type of bar chart is called a Gantt Chart. The production schedule you encountered in Chapter 1 is a modified Gantt chart.
The Gantt chart is a method of planning and scheduling the many steps in a complicated project. Each step is a label on the Y axis and is represented on the chart with a bar which does not necessarily connect to the baseline. Each bar has a separate starting and ending point on the scale.
A Gantt chart scale can be any length from a few days to many years. Very fine scales, however, will not read well on slides, so keep your scales simple on the slide and use a handout for a more detailed chart.
Figure 10.33: Gantt Chart
Another use of timeline bar charts is in the advertising industry, where a Media Buying Chart is created to show when commercials and other promotional events are to be scheduled. Each label may have several bars aligned with it, starting and stopping for each period where it is active.
The timeline scale is always continuous. To accurately reflect blocks of time, your scale figures should fall between your grid or tick marks, rather than on them. This designates the space between the tick marks as blocks of time, rather than particular moments.
Like a Gantt chart, other timeline charts have a tendency to get very complicated and difficult to read on slides. Use them sparingly and edit them mercilessly to make them clear, readable and informative. As a general rule, use no more than five or six bars per slide. Use handouts when you need to provide more detailed data to your audience. You can also break up a large timeline charts into smaller sections like a build and talk about individual parts, letting your audience refer to handouts for the big picture.
Adding Depth to Bar Charts
Bar charts with depth receive the same benefits and suffer from the same problems as column charts. There is a tradeoff between dramatic graphics and accuracy. Remember that your audience isn't going to take tape measures to the screen to measure your graphics for accuracy, so don't be too worried about it.
Creating Bar Chart Builds and Reveals
Guidelines for bar chart builds and reveals are essentially the same as for column charts. You can often make complex timeline charts more readable and understandable by creating them as builds or as highlight reveals. One method of dealing with complex information on a Gantt chart is to selectively reveal information on each build, similar to the way a stepback reveal is created in text slides. In Figure 10.40, each bar is highlighted, with annotations which clarify the chart. As each bar is revealed, the previous annotation is removed and a new one is added.
Figure 10.34: Timeline Build
The line chart is the most common and useful chart for presenting information. Many of the messages possible with column and bar charts are also possible with a line chart. But line charts are at their best when showing trends, displaying increases and decreases, and showing relationships between several different data series.
Line charts in business presentations are frequently time-oriented, like column charts. But where column charts are limited in the number of time increments which can be plotted (because of the width of the bars), a line chart can give a smoother, more detailed comparison of data. The line chart in Figure 10.35 shows monthly sales figures for a salesperson. Treated as a column chart, this data is overbearing. As a line chart, it gives a clean, accurate picture of growth.
Figure 10.35: Line Chart
The line in a line chart is plotted against both the X and Y scale. A half or full frame on the plotting area is essential, although a staff grid can be used, since it acts like a full frame.
Your plot line should be fairly thick and easy to see relative to the grid. Choose a line thickness at least three to four times thicker than your grid lines.
If there is more than one plot line on a chart, choose bright, contrasting colors for your lines. Make sure each line is labelled to identify it. Try to avoid more than 4 lines on a chart; any more will be confusing and uninformative, even if you use builds or reveals.
You can use virtually any grid/tickmark combination with a line chart; but every line chart requires at least an X and Y axis scale. Scale labels for line charts should be "grid aligned"; data points should align on the grid lines, not between them.
Choose your grid based on the amount of accuracy required. If you show data values, or if extreme accuracy isn't important, a simple frame with tick marks will be adequate. But if you need accuracy in a line chart, use a full grid.
Placing data values at your line plot points does not eliminate the need for a grid. Generally, including data values on a line chart will make it seem cluttered. Line charts are best for showing trends, not numerical data. If you must show all of the data values, you may be better off with a table.
Surprisingly, perhaps because of its simplicity, there are very few variations on a line chart.
Step Line Charts
Step line charts plot abrupt change rather than gradual change. Their purpose is to show "plateaus" of values related to time. For example, Figure 10.36 shows a step line chart of bus fares over a 35-year period. There is a vertical change wherever a fare increase occurs.
Figure 10.36: A step line chart
Line Gap Charts
A gap chart is a simple variation on a line chart which shows the relationship and difference between two separate lines. Two separate line series are plotted, then the area between them is shaded to show the specific relationship between the lines. In Figure 10.46, sales and costs are plotted, and the area between the lines is shaded for profit and loss.
Figure 10.37: Line Gap Chart
Standard Deviation Charts
The standard deviation chart is very common in scientific presentations, and is used in many types of statistical analysis. A small line is used to show the margin of error involved in the calculations used to determine the data point being plotted. On the line is a dot indicating the main data point; from this dot, an "I" shaped line is drawn vertically, with the top indicating the maximum margin of error, and the bottom indicating the minimum margin of error.
Figure 10.38: Standard Deviation
Adding Depth to Line Charts
When it comes to adding depth or perspective effects to line charts, the main advice you'll get here is: Don't! There is often a feeling that if your column, bar and pie charts have depth that your line charts should also. Adding depth to a line chart creates a "ribbon" which is usually confusing and adds nothing to your chart.
Creating Line Chart Builds and Reveals
A build on a single line is uncommon in line charts, except as an animation effect in a screen show. The slope and shape of a line are usually a single message, and therefore a single line doesn't require a step-by-step reveal.
However, a build is actually a good idea when dealing with several lines on a chart, allowing individual messages to come through and allowing clear comparisons between lines. Figure 10.49 shows a series of reveals. A reference category is left highlighted throughout, while three lines are successively revealed for comparison to the reference and then toned back.
Figure 10.39: Line Chart Build
An area chart is a plotted line with the area between the line and the baseline shaded. The large shaded surface presents a more graphic look to the chart and also serves as an indicator of volume, making it closer in effect to a column chart than a line chart. A single data set area chart is more of a decorative effect than anything else; it serves no better than a line chart. The real purpose of an area chart comes to the fore when it is used to plot multiple data sets, comparing component parts of a total amount, like a stacked column chart.
The basic design of area charts is the same as for line or column charts. Normal grid and tickmark options apply, although you should not overlay the grid on top of the plotted areas. Doing so will give your chart a "checkerboard" look. You should use external tickmarks for your X axis, since the internal grid will be hidden by the plotted areas.
Like other stacked charts, avoid using more than four or five areas in the chart. If you have several small areas to plot, combine them into a category called "Other" and plot them as a single segment.
The largest or most important area of your chart should sit on the baseline. This enables the audience to determine absolute data values of your most important information. Once an area has been plotted on top of another, without a straight baseline, it becomes difficult to assess actual values.
Always place labels for the plotted areas directly in the area. If a plotted area is too small for text, use a callout to identify it.
Figure10.40: Stacked area charts
There are no significant variations on an area chart, although a very dramatic effect can be achieved by placing a photograph in the plotted area. The photo should have a direct relation to the data being plotted - don't just put it there for the pretty picture.
Figure 10.41: Photo Area Chart
You can also add depth to an area chart, with the same scale and accuracy problems as column charts
Creating Area Chart Builds and Reveals
There are two methods of creating builds in area charts and they are strongly message-related. Which one you use has a great deal with the way your speech and message are structured.
The first method, an add-on build is the simplest, but it requires you to talk about the baseline item first. The largest segment is plotted first, and then each new category adds on on top of it. Since each new area is plotted on top of the original, uneven area, actual data values for the new segments are difficult to determine. This is the type of stacked area chart generated by most presentation and charting programs. The disadvantage of this type of build is that only the first segment displays a true picture of its individual data.
Figure 10.42: Add-on Area Chart Build
If you have the time, and it suits the content of your speech, the push-up build is a very dramatic way of creating an area chart build. In a push-up build, each new segment is plotted under the previous segment. In this way, the new segment can be viewed in relation to the straight baseline, giving a truer picture of its data. The older segment is pushed up and becomes distorted by the new segment. In this build your most important bottom segment appears last in the sequence, allowing you to build your speech to the most important topic.
Figure 10.43: Push-up Area Chart Build
Both of these build types can also work as reveals by toning back the color of older segments when a new segment is revealed.
The scatter chart is a tool for plotting coordinate "points" on two scales, where there is no direct trend or relation between the individual data. Each dot or other mark on the chart represents two data values along the X and Y axes. Both axes are usually continuous scales..A common use for the scatter chart is a price-performance graph.
Figure 10.44: Scatter Chart
Scatter chart frame layout is similar to line charts. Use a full grid, but don't space your grid too tightly. Labels identifying your plot points will have to compete with the grid.
Plotting single points for each category involves attaching a label to each dot as an identifier. Of course, this can get very complicated very quickly. Try to limit your plot points to less than ten on this type of scatter chart.
Use a legend when plotting multiple points from single or multiple categories. You can use quite a few points with this method, since individual points usually aren't labeled. Many programs allow you to use different shapes for scatter chart points, but at a distance, shapes become difficult to tell apart. You will achieve better legibility if you use color to differentiate your categories. Use high contrast colors such as yellow, green, and light blue so that closely spaced dots won't be mistaken for each other.
A Bubble Chart adds a third dimension to the scatter chart. In addition to the position of the dot, its size also acts as a data value. An important thing to remember when creating bubble charts is that your data should be reflected in the AREA of the circle, not its radius or diameter.
Figure 10.45: Bubble Chart
A pictograph is a chart which substitutes appropriate graphic icons for the traditional graphics of bar and column charts. All of the rules and suggestions for layout and design of bar and column charts also apply to their pictograph equivalents.
Figure 10.46: Pictographs
Here are a few additional guidelines for creating your pictograph charts:
- Keep your graphic symbols simple.
- Symbols should be appropriate to the topic of the chart.
- Data value changes should be represented by more or fewer symbols, not by the size of the symbols.
- Pictographs are often visually inaccurate. Include data values for precision.18
The circle has always been a symbol of the whole. A pie chart's main function is to show the relationship between the parts of that whole; showing the relative share or percentage of a total amount.
In business presentations, a pie chart is most often used to show the share of individual categories in relation to a whole, such as market share or the distribution of budget expenditures.
Of all the chart types, pie charts are the simplest to design. The scale of a pie chart is always 0 to 100 percent, and your data must always total 100 percent. All of your data, no matter what its original form, needs to be converted into percentages of the pie. Many charting programs will allow you to enter raw data in any type of unit and the program will automatically convert your data into percentages. Limit your pie charts to no more than six pie segments to maintain clarity.
You should avoid using a legend with pie charts; it is difficult to visually connect the legend colors to your pie segments. Use labels inside or directly adjacent to your pie sections.
Since people read clockwise, the most important pie segment should begin at the 12 o'clock point of a flat pie chart. The emphasis is different in three dimensional pie charts, however
You can highlight segments of a pie chart by pulling a segment out of the complete circle. The "exploded" segment can also be highlighted with a brighter color for further emphasis.
Figure 10.47: Exploded Pie Segment
Adding Depth to Pie Charts
Three dimensional pie charts are always very dramatic. The most common form is simply a thick disk tilted back, showing the front edge of the pie.
The main disadvantage of 3D pie charts is that they tend to visually distort the data. Pie segments at the top of the chart are diminished, since they seem further away from the viewer. Pie segments at the bottom are over emphasized since they seem closer; they also have the visual advantage of seeming larger because the edge adds to the size of the segment.. You can use this effect to your advantage by placing a segment you want to emphasize at the bottom of the chart. Even if the segment is the smaller than some others, it will appear larger and more important because of its position.
Figure 10.48: 3D Pie Charts
100% Stacked Column Chart
With some forms of data there is a temptation to use several pie charts on a single slide. The best way to avoid multiple pies is to use a 100% Stacked Column Chart. In a 100% stacked column chart, all of the columns are the same height, representing a value of 100%. The individual segments of each column represent percentages of the 100% total. A 100% stacked column chart requires the same basic design and layout guidelines as other column charts.
Figure 10.49: 100% Stacked Column Chart
Proportional Pie Charts
At some point you may need to compare two pie charts where the relative size of the pies also represents important information. Proportional pie charts provide the ability to compare size and volume as well as share. In a proportional pie chart, each pie has a data value represented by the area of the circle, and segment percentages which represent portions of the whole pie. Never use the diameter or radius of the circle to determine the proportions of your pies. You will be greatly exaggerating the difference in size.
Figure 10.50: Proportional Pie Chart
Note that in Figure 10.50, a legend is used to label the segments. Whenever you use proportional pies, you may end up with very small segments which make it impossible to label properly. In this case a legend is best solution.
Rotate Pies for Variety
You can add variety and style to a presentation by rotating and tilting pie charts (if your software enables this). A rotated and tilted pie chart can seriously distort your pie sections, so be sure to include data values to prevent your audience form misinterpreting the charts.
Figure 10.51: Rotated Pie Chart
Creating Pie Chart Builds and Reveals
There are several ways to create builds and reveals for pie charts. The simplest is to add each segment in order to the pie. You can achieve more dramatic effects by combining exploded segments, toneback colors and builds.
Figure 10.52: Pie Chart Highlight Reveal
Often a single type of chart won't convey the message you need for your slide. A combination chart merges two different chart types into a single, coherent message. Since the possible combinations of column, bar, line, area and pie charts are endless, you must first decide if a combination chart is appropriate for your information. Here are some guidelines for using combination charts:
The chart should be a comparison between two separate conflicting or corroborating messages. For example, "Sales volume is up, but profits have fallen." or "Widgets are a major contributor to our market share."
Figure 10.53: Combination Charts
Your information should be based on two separate measurement systems. For example: percentage and amount or dollars and volume. If you are using two scales, make sure it is clear which scale belongs to which chart graphic.
Above all, keep things simple. Whenever you combine two types of charts, you double your need to eliminate extraneous information and graphic clutter. A combination chart should be no more difficult to read than any other type.
Flow and process charts are intended to show relationships between people and processes rather than numbers. The most common flow charts in business presentations are the organization chart, the PERT (Process Evaluation Review Technique) chart, and decision trees.
No matter which type of flow chart you use, it's important to organize your chart to fit the available space on the slide.
- Keep your type as large as possible, but don't crowd the edges of your boxes.
- Abbreviate wherever possible. Use only first initials for names in organization charts.
- Keep connecting lines short use the space for boxes and text.
- Always supply printed handouts when using complex charts.
Organization charts show the relationships between people in a business or other group. The normal layout for an organization chart is to place the highest-ranked person at the top, with subordinates layered underneath. A person's name, sometimes with their title or job description is placed in a box. Lines connecting the boxes define reporting and authority connections. The level of the person's box indicates rank and authority.
Figure 10.54: Organization Chart
When creating an organization chart, try to maintain the look of a grid, with same size boxes (especially at each level) and even spacing. It is very easy to get carried away with organization charts. The temptation is to attempt to show every person in a single chart. You will be more effective, though, if you limit the number of boxes to four rows down by six boxes across. If you need more detail, break up your chart into separate slides. Remember, keep the text readable.
A PERT (Process Evaluation Review Technique) chart diagrams the process by which a product is made, or a task completed. Each step in the process is represented by a box, with connecting arrows showing the process paths.
Figure 10.55: PERT Chart
Don't be afraid to change direction to conserve space. In Figure 10.71, the straight line of the flow chart has been "bent" (arrows flowing left and right when they should be straight up and down) to make the chart more horizontal.
A decision tree is a visual tool for making choices between various options. Questions are posed, and the answers determine the direction to follow in the chart. Eventually, the path leads to a final decision.
Figure 10.56: Decision Tree
The diamonds used for decision forks don't hold a lot of text, so be sure to keep thing short. Decision trees are seldom simple enough for slide use. Always supply printed handouts and just use your slide as a reference while speaking.
Using Color In Charts and Graphs
Color is an essential part of communicating with charts and graphs. The complex data presented in charts is made more understandable by the effective use of color.
The basic color guidelines explained in Chapter 7 are also appropriate when applied to the use of color in charts and graphs. Remember that the colors you choose for highlight colors in your palette should also be appropriate as chart colors.
- Use warm, cool and neutral colors to guide audience reactions. Use warm, bright colors to accentuate chart elements, cool and neutral colors to de-emphasize them.
- Use familiar color associations to reinforce your message: green for profit, red for loss, etc.
- Color contrast is essential to communication. Choose chart colors for maximum contrast with the background and other chart elements.
- Link related chart elements by assigning "signature" colors. to frequently used chart categories or topics.
- Use moving highlights to clarify complex charts and graphs and to pace the audience's attention.
Using Signature Colors in Charts
It is very common that during the course of a presentation you will deal with the same categories or topics in several charts and graphs. For the sake of consistency, and to aid audience understanding, you should choose "signature colors" for each different category and use that same color throughout your presentation. Figure 10.73 shows three different categories represented in four different charts. The same color is used for each category in all of the charts, making all of them easier to understand.
Figure 10.57: Color Coding
The Real World:
Lies, Damned Lies, and Statistics
Wednesday, October 10, 11:00 AM: Jim Gonzalez starts his rounds of the executive offices, looking for answers and making suggestions on the presentations. His first stop is Victoria's office.
"Morning, Victoria. I've got a few questions about this chart you added last night. I think it needs some work." He placed her storyboard sketch on the desk.
Art Insert S10.1 Victoria's Rough
"Four pie charts is a little too busy for one slide," he said. "Why don't we do this as a 100% column chart?"
"Do you have an example?" asked Victoria.
"Here, take a look at this" He sketched on his pad:
Art Insert S10.2 Jim's Rough
"I don't know, Jim. Somehow that doesn't look quite as impressive as the four charts."
"True, but it will be easier to read."
Victoria looked at Jim's rough again. "The main problem I have with this is that our share looks too small in column form."
"Well, we could just do this as a straight column chart, treating the percents as raw figures. It's not exactly Kosher because we should be showing the whole market if we're going to be showing our share."
"What do you mean?" she asked.
Jim sketched on his pad again.
Art Insert S10.3 Victoria's Final Rough
"This way makes the columns twice as high and more impressive, but doesn't misrepresent the figures."
"OK, Jim. I think that will work. Thanks."
Jim headed down the hall to Alan's office. Alan had given him a table to be turned into a slide, but Jim wanted to turn it into a chart.
"Good morning, Jim. How are you doing on the slides?"
"Fine, Alan. Almost done. I'd like to talk to you about this table."
"What about it?"
"I'd like to try doing this as an area chart instead of as a table. I think it will look a lot better."
"What did you have in mind?"
"Here's your original table.
Art Insert S10.4 SB Alan's Rough
"The three rows of figures for widgets, flotsam and thingies add up to the total on the fourth line, so I thought it might make a good area chart like this" Jim showed Alan the sketch he had made earlier.
Art Insert S10.5 SB Jim's Alan Rough
"That's perfect, Jim. The chart shows the fluctuations in our unit sales much more dramatically. Go with it."
"Thanks Alan. I'll have final proofs for your show this afternoon."