Using Central Tendency To Assess Results
German philosopher Friedrich Nietzsche said that “those who were seen dancing were thought to be insane by those who could not hear the music.” Indeed, what you see is always relative to what surrounds it. At no time is this more important to keep in mind than when analyzing the performance of your programs.
Just about anybody can click a button to run a standard report, look over a spreadsheet, or recite how many donors came in as the result of the last appeal. But don’t confuse these actions alone with being analytical. Observing the data is an important first step in analyzing your efforts, but the next step is interpreting the data. This means figuring out what stories the data have to tell.
Interpreting data requires context. Rather than looking at metrics in a vacuum, you need to compare them to something. For example, if someone asks how your alumni participation efforts are going and you reply that your alumni participation rate is 12 percent, that doesn’t provide any context. Is 12 percent good? If so, how good, and what evidence do you have to support this claim? By itself, 12 percent is just a number; but comparing that number to something else gives it context and makes it relevant.
One of the simplest and most useful ways to give context to your metrics is to compare them to one of the three measures of central tendency: the mean, the median, and the mode.
The mean (also known as the average) is the most common measure of central tendency. It’s the sum of every number in a data set divided by the total number of data points. For example, if your annual fund raised $1 million last year from 10,000 donors, then the average gift per donor would be $100 ($1,000,000/10,000 = $100). However, the mean can be skewed by outlier gifts. For example, if one of those 10,000 donors made a single donation of $500,000, it would inflate the mean. Without that one donation included in the calculation, the figure would be only $50, so it’s important to remember that the mean does not necessarily give an accurate representation of the typical donor’s gift.
The median is the middle value when the numbers in a data set are arranged in ascending or descending order. If the number of values in a data set is even, then the median is the mean of the two middle numbers. For example, imagine that your phonathon received five gifts last night in the amounts of $60, $75, $100, $125 and $500. In this case, the median gift would be $100, since it is in the middle when you sort the gift amounts from lowest to highest. This is significantly lower than the mean gift of $172. Since outlier gifts can be common in annual giving (remember the Pareto principle), the median can be a good way to measure central tendency and paint a more accurate picture of a typical giving pattern.
The mode is the value that occurs most frequently in a data set. For example, if your program received five gifts in response to an email appeal in the amounts of $25, $25, $25, $500, and $1,000, then the modal gift would be $25, because it occurred more often than any other gift amount. Understanding where a distribution of gifts tends to cluster can be helpful when assessing the messages and ask amounts that go into your appeals or reside on your online giving forms.
These three measures of central tendency offer annual giving professionals a variety of tools to evaluate the performance of their appeals and programs. By providing context to your results, each of these tools can help you understand how an individual metric compares to the greater population and where it falls within the overall range of outcomes. Ultimately, this will let you know whether you can celebrate your success, or if it’s time for your team to head back to the drawing board.
Want to learn more? CLICK HERE for AGN’s Webinar on Fundamentals of Data-Driven Annual Giving.