When is a picture worth a thousand words? High-Impact Use of Charts in Grant Writing

Imagine my chagrin! I’ve been working on an article including pie charts, yet failed to remember that Wednesday was Pi(e) Day (celebrating 3.14… on 3/14, get it?). I’m a lover of geeky holidays, from International Talk Like a Pirate Day (September 19) to Festivus (December 23), so it was a shame to miss the opportunity to talk about pie charts on Pi(e) Day . To make up for this oversight, I may need to add in an extra observance, like Morse Code Day on .- .–. .-. .. .-.. /  ..— –… (April 27).

Well, apologies for missing the holiday mark; I hope—like the best fruit pies—that this article is just as good the next day. Now, back to the task at hand…. Charts can be a powerful addition to grant narratives and other proposal elements. Conversely, they can be a confusing or misleading waste of space. In this article, I’ll talk about the considerations for using charts, types of charts and the data they represent, and tips for chart formatting.

We use charts to display quantitative data—also known as “stuff you can count or measure” like 5 pies or 300 grams banana cream filling.  Charts can help prove or disprove a point or convey important information. This can also be accomplished via table or text—so when do charts make sense? In general, we use charts when they:

• Display information more concisely than a narrative;
• Show information in a more impactful way (or hammer home a key point);
• Sum up information with too many variables, data points, or dependencies to be encapsulated in text;
• Reveal relationships between variables; or
• Provide visual relief from text blocks for a long proposal.

Looking in Excel, there is a dizzying selection of chart types. While variations abound, nearly all break down into several core classes: pie, bar, line/area, and scatterplot. Choosing which chart type to use is not based on aesthetics, but rather on the type of information displayed:

Pie Chart

Use pie charts to show parts of a whole, like proportions of your budget allocated to various categories or age group breakdowns for your clients. While pie charts can be useful, they have several drawbacks. They only show a limited amount of information (one variable); they use up a good bit of real estate; they make comparisons between roughly proportional subcomponents difficult; and they make comparisons between two data series difficult (see figure 3 for an alternative to side-by-side pie charts). Still, they have a place. Here’s my favorite pie chart:

Chart 1: Pie chart (Source:  Jamie Schimley, GraphJam, 2008)

Bar Chart

Bar charts are flexible and can display a number of points including: comparisons of separate items by a common variable; changes for a single variable over time; and proportions of a whole. The bars represent numbers scaled by size (and can include negative values). Bar charts often handle two and often three variables adeptly. While bar charts fit many needs, they have limitations. For instance, you don’t really want to squash 10+ bars in a chart and they can appear heavy when something like a line chart would do. Sub-types, shown below, include vertical and horizontal standard bar charts (Excel refers to vertical bar charts as column charts), stacked bar charts, and histograms.

Chart 2: Column Chart (Source: Information Resources Inc., Bakery Production & Marketing Redbook 2003, p.26)

Chart 3: Two-Item, 100% Stacked Bar Chart (Sources: Michael Ruhlman, Ratio: The Simple Codes Behind the Craft of Everyday Cooking, 2009 and Mark Bittman, How to Cook Everything, 2008)

Chart 4: Distribution Histogram (data is fictional though U.S. average annual consumption is 46 slices, per Agriculture Council of America, 2012)

Line and Area Charts: These charts typically show time-series data (e.g. stock prices by month). They work much better than bar charts for displaying larger sets of data points.

Chart 5: Line Chart (data Harvard Business School. Mergers, Acquisitions, Restructurings, and Corporate Governance. Eskimo Pie Corporation. 1992)

Scatterplots: These charts capture data points expressed by two variables, such as hours that students studied and their test scores. Each variable is shown along either the vertical or horizontal axis of the chart. The two variables may be the same (e.g. plotting geographic coordinates, like the chart below, which simply mirrors spilled berries). Variables are more commonly two different types (e.g. elevation of various communities along the horizontal axis and their annual snowfall on the vertical axis). Once the data points are graphed, Excel can add a line or curve to show the mathematical trend (like the orange line in Chart 6).

Chart 6: Scatterplot with Trendline

1. Don’t use 3-D effects, exploding pie charts, or other whiz-bang options. This just gets in the way of your point and visually clutters the graphic.
2. Use the title of the chart to identify the variables shown in the chart. This can be general (Pie Contest Participants Eating more Each Year) or specific (Mean Participant Slice Consumption per Pie Eating Contest, 2002-2009) .
3. Use the title of the chart to reinforce the conclusion you are showing with the data; for instance, “Child Hunger Rises as Economy Dips.” The exception to this would be if you are writing to a funder that emphasizes objectivity (academic or scientific sources), in which case simply describe the data shown, “Child Hunger Rates and GDP,” with a level of precision and formality similar to the funder’s own materials.
4. Match the typefaces in charts with those in the document. Same with colors when used in text. Also, make type sizes consistent across charts.
5. Beware of using color charts unless you know reviewers will be provided with proposals in color. When in doubt use grayscale or black and white patterns.
6. Microsoft Office software (Word, Excel, PowerPoint) often defaults to position a chart legend to the right of plot area—but this is usually the most space-wasting placement. Move the legend to a position along the bottom or in a box floating over a blank part of the chart area.
7. Make sure axis titles describe what the measure is and what unit is used, for instance “Annual Mortality (per capita).” Scale the axes as needed so you don’t waste space on area that doesn’t contain data.
8. If used, gridlines should not overshadow the data (make the units are relatively large and consider using a light gray for gridline color).

All too commonly, people use charts to manipulate data or make poor quality data appear more legitimate. Grant reviewers are hip to this trick, so be sure to:

• Use quality data,
• Clearly show your sources and data parameters,
• Avoid altering scales just to exaggerate or downplay the effects shown, and
• Resist jumping to conclusions about one variable causing change to another just because they move in synch (correlation does not prove causation—otherwise we might conclude that eating soup causes the flu simply because these both rise during the cold months).

Charts can be a powerful component of proposal writing. They lend credence to your message, capture a tremendous amount of information “at a glance,” and offer reviewers a welcome break from reading text. Like the rest of your writing, charts should be clear, concise, and cogent. They should be able to stand on their own and be easily understood. Always keep your data message in mind when you are selecting what to display, what format to use, and how to design and format within that chart type.

Creating the perfect chart is much like creating a perfect pie. Both take planning and practice. We hope that this article has provided some of the essential ingredients you need to make your charts powerfully compelling…like a fresh baked pie sitting on a windowsill.

Contact: Aly Sanchez, Director of Projects, aly@thegrantplantnm.com

This post was filed under: Grant Writing