Analyzing and Presenting Quantitative Data

A number without a unit is ethereal and abstract. With a unit, it acquires a meaning—but at the same time, it loses its purity. A number with a unit can no longer inhabit the Platonic realm of absolute truth; it becomes tainted with the uncertainties and imperfections of the real world. To mathematicians, numbers represent indisputable truths; to the rest of us, they come from inherently impure, imperfect measurements.

Seife (2010, pp. 9–10)

Lies, damned lies, and statistics.

Mark Twain

They say numbers don't lie. Perhaps numbers don't lie, but the conclusions drawn from numbers and the methods used to draw those conclusions may often be subject to criticism. Communications professionals must be able to present data in a clear and concise manner that makes the conclusions drawn obvious and logical to the reader.

Validating Your Data Sources

It is important to recognize the power of quantitative data, and the inherent risks this power poses for its user. Because quantitative data looks scientific, it is often believed to be true. By presenting data in numerical or statistical form, communicators can improve the chances of having their message be accepted, and thus improve the chances of influencing the target publics, or audiences. If using data and statistics is so effective, it might seem like a no-brainer. But, as with many things that seem too good to be true, there is a downside.

If you are of a certain age, you may remember an advertising campaign for chewing gum that relied on a very specific presentation of quantitative data. The data presentation in the advertisements went something like this: Four out of five dentists surveyed recommend sugarless gum for their patients who chew gum (Newman, 2009). This sounds impressive and probably caused people to stop and consider the gum. (Note: Another, if less persuasive, way to express the same data would be: Eighty percent of dentists surveyed recommend sugarless gum for their patients who chew gum.)

For a professional communicator acting ethically, the problem with the above data is that there's no indication where it came from, how many dentists it represents, who these dentists are, and when the data was gathered. Theoretically, we could be looking at the results of a survey involving four dentists who work for the gum company, and one dentist in private practice located in a major sugar growing area. There's no way to know.

In order to use data in your communications, you must make every effort to ensure the validity of the data and the manner in which it was obtained. This involves looking for the source of a particularly appealing data point or statistic, then ensuring that that source is credible and has gathered the data in an ethical manner. Sometimes there are valid reasons why it may be difficult to validate a piece of quantitative data: for example, to protect the privacy of the subjects, or because the study was commissioned by an organization and is thus their proprietary intellectual property (IP). In such cases, use the sources with caution.

Some data sources that we can assume are valid include major research organizations like Pew and Gallup, large public universities, and government agencies and departments. Never use a statistic or other quantitative data from social media unless you have validated it with a reputable third party.

Analyzing Your Data

Whether you obtain data on your own, or as is often the case, use data obtained from third party sources, you need to analyze it in order to pull out the facts most relevant to your goals and objectives. Sometimes this is as easy as identifying a specific number that supports a contention, e.g., our organization ranked first in a reputable industry survey. Other times the process can be much more involved and require the application of statistical processes and data modeling.

Payne (2011) reminds us to "measure what matters." Metrics that communicators often want to pay particular attention to include the following:

  • percent change in perceptions,
  • percent change in awareness,
  • percent increase in preference, and
  • percent of visitors likely to return, purchase, or recommend.

There are many potential sources of data, some of which may already be clearly tabulated and presented, while others require you to pull the numbers that you need and tabulate them yourself. Once you have validated the data source, you can use a tool like the Pell Institute's basic data analysis toolkit (see the References list for more information) to evaluate the numbers you have and help you determine how much confidence you should have in making claims based on them. The Pell Institute suggests that we should first decide what kind of data we have:

  • nominal (conceptual with no special order),
  • ordinal (ordered in some fashion),
  • interval (standardized distances between values but no zero value), or
  • ratio (continuous, ordered, has standardized differences between values, and a natural zero; to a scale).

The Institute then suggests you tabulate your data. In a frequency tabulation you can verify that you entered the information correctly, identify frequency of each category, high and low scores, and the spread of the scores.

In a percentage tabulation, you can identify the proportions of your results.

Once tabulated, you can describe your results. Standard descriptions include:

  • mean,
  • median,
  • mode, and
  • minimum and maximum values.

You can also dive deeper into the data by using cross tabs to separate the data across different variables and subcategories. With the data separated or disaggregated, you can conduct additional analyses to discover any relationships or correlations that might exist.

WIth properly validated data, the basic data points described above can be very helpful in supporting your communications and proving your contentions with mathematical precision. That said, you may come across more advanced statistical data that is presented using arcane expressions like Pearson's r, the coefficient of variation (CV) and chi-squared. This is not meant to confuse you, it is just scientific shorthand used by statisticians. Still, if you are not versed in this language, approach this type of data with caution.

It's always a good idea to seek help from your more experienced colleagues or from your manager if you run into data that you do not understand. Don't try to figure it out by yourself: If you don't understand the data, you can't communicate it to your audiences or engage your publics about it with any degree of veracity. Your publics will notice if you aren't comfortable with and knowledgable about the data you use.

Presenting Your Data

Now that you have good meaningful data from a valid source, you need to communicate it to your target publics. This task can be easy or hard depending on your publics, and their ability and willingness to accept and understand quantitative data. If you have a lot of important data and a numbers-literate public, you may be able to simply present a data table with some supporting text and be done. But, much of the time, this approach will not be effective.

Presenting in Graphic Form

For most publics, you will need to develop easily understood representations of the data you are communicating. Putting your data into graphic form is a great way to create interest, and perhaps engagement in your data and what it represents. Fortunately, there are many data visualization tools available to help you create the most impactful display of your primary data points. Of course, before you start trying to use an industrial strength graphics generator, you should probably see if Excel can meet your needs. Pie charts and line charts (both two and three dimensional), help array complex data into easily discernible relationships. Tables are less exciting but can often make relationships clear as well.

Infographics are a wonderful tool for explaining complex organizations, processes, and concepts. Venn diagrams are useful in showing conceptual relationships. You will need the assistance of a good graphics artist to create strong infographics,

Other visual media, such as video and animation of still images help tell stories about the data you have collected.

Below are some examples using the four out of five dentists example discussed above.

pie graph showing that four out of five dentists recommend sugar-free gum. 80 percent of the pie graph is blue, representing the dentists who recommend the gum, and 20 percent is orange, representing the dentists who do not recommend the gum.

Pie Graph Example

bar graph showing that four out of five dentists recommend sugar-free gum. The y-axis has tick marks going from 0 to 5. A bar reaching up to the 4 tick mark on the y-axis is blue, representing the dentists who recommend the gum. A bar reaching up to 1 on the y-axis is orange, representing the dentists who do not recommend the gum.

Bar Graph Example

illustration of four happy dentists who recommend sugar-free gum separated with a toothbrush from the one outlier who does not recommend the gum

Illustrated Example

Presenting in Text

There are some situations where you may have a wealth of data, but only want to pull out one or two highly relevant insights for your publics, or audiences. In this situation, the graphical and/or tabular presentation methods will not work—you will have too few data points. In situations like this, it is best to highlight your key finding or data point in the body of your text. If possible, you can incorporate the finding in a subtitle, or a callout box, or use bold and/or italics to draw the reader's eye to it. The narrative discussion of the finding should be clear and straightforward, using simple language (no arcane scientific or statistical references).

For examples of this type of material, you can look at the media releases about a company's financial performance—especially when it is struggling. The organization's communicators will work hard to identify some positive data points in the financials, and then highlight those in order to balance the otherwise negative results.     

When Not to Use Data

While quantitative data and statistics can be powerful tools to leverage in our communications programs, they are not appropriate for all occasions. Sometimes emotional appeals will be more effective than rational arguments, and in these situations quantitative indicators should not be used, or should be used sparingly.

Other situations when quantitative data and/or statistics should be used with caution are when a situation is fast moving, such as a crisis. If you cannot be certain of the data, do not provide it. Provide qualitative information and indicate when you expect to be able to provide hard data—guessing is worse than saying you don't know.

Another situation when quantitative information may be inappropriate is in the aftermath of a tragedy. If a passenger plane crashes, the airline's message should be one of sympathy for the victims and a commitment to determining the cause of the crash. It should not provide data indicating that air travel remains the safest form of transportation, the current unfortunate incident notwithstanding.

Quantitative data and statistics are important tools that communicators can use to make their messaging more credible and impactful. As the availability of data continues to increase due to the digitization of our economy we can expect ever greater amounts of quantitative information to work with, and so this is an area where you will want to hone your skills. That said, it is just as important to know when not to employ this type of information, as to know how to employ it correctly.

References

Newman, A. A. (2009, July 17). Selling gum with health claims. New York Times. Retrieved from https://www.nytimes.com/2009/07/28/business/media/28adco.html

Payne, K. D. (2011). Measure what matters. Hoboken, NJ: John Wiley & Sons, Inc.

Seife, C. (2010). Proofiness: The dark arts of mathematical deception. New York: Penguin Group

The Pell Institute for the Study of Opportunity in Higher Education, the Institute for Higher Education Policy, and Pathways to College Network. (n.d.). Evaluation toolkit: Analyze quantitative data. Retrieved from  http://toolkit.pellinstitute.org/evaluation-guide/analyze/analyze-quantitative-data/