This is part one of a series of short columns on problems in data analysis. I will use the lens of COVID-19-related data and interpretation (and misinterpretation) to shed some light on how nonprofit fundraising operations can avoid similar data-driven issues.

Amongst the annoyances of news in the time of coronavirus is the glut of charts. Charts showing that we might all die. Charts showing that no one, if you squint and average it out somehow, is going to die. Graphs adjusted for population size, or not. Charts of other charts stacked on top of each other. Models attacking the assumptions of other models.

Like everyone else, I have my own non-expert idea of what the good ones demonstrate, and why the bad ones are so bad. And my, are there some bad ones!

But notice that no study or model or fact is really breaking through the noise. Those who insist on denying or questioning the severity of COVID-19 or our response to it as a nation continue to cite one set of numbers (or a way of interpreting those numbers), and those who insist that the most extreme and costly measures are still needed to hold back the virus will cite other numbers with their own interpretations.

Despite the amount of data we now have about COVID compared to even a month or two months ago, one cannot help but notice that nearly no one seems to be changing their mind in either direction.

“The numbers don’t lie” turns out to be, well, a lie. It turns out that there simply is no such thing as “the numbers” devoid of human interpretation and judgment. Data is not self-interpreting.

Alasdair MacIntyre, perhaps the greatest living philosopher, memorably wrote that “facts, like telescopes and wigs for gentlemen, were a seventeenth century invention.” What he meant was not some sort of post-modern denial of all truths. He meant that the philosophers who conceived of our modern notion of “facts” in the 1600s were mistaken “to conceive of a realm of facts independent of judgment or of any form of linguistic expression.”  

Even the simplest questions in nonprofits admit of judgment and interpretation.

Take for example “How much money did we raise last year?” Well, do we do cash or accrual accounting with regard to multi-year pledges? What about that planned gift where we know the amount but it is still tied up in probate? What about the gifts solicited by our staff in December but received in January? Are all those t-shirts people overpaid for at our events more like raised dollars or more like program revenue?

Collecting and reporting on data is a highly rhetorical act. Charts and graphs and tables are not facts; they are arguments. And as arguments they can be rationally interrogated, questioned, and reinterpreted. 

The more we are aware of this philosophical quality of data, the better we will be at seeing nuances and inoculating ourselves against common errors in data work in fundraising and nonprofits.

It is common now to hear an organization boast of being “data-driven” in its orientation. That is always cast as a positive thing, as though eliminating any human element is a laudable goal.

But if data interpretation requires solid judgment, such a view is folly. It reminds me of the YouTube video I have probably linked more than any other: Mitchell and Webb’s provocative anti-utilitarian skit “Kill the Poor.”

The wise nonprofit leader is not a slave to data—he or she must be its master. And that requires a practical wisdom and broadness of mind that can interact successfully with, but is never provided by, the data itself.

The post Data as a philosophical field: data analysis, part one appeared first on Philanthropy Daily.

Share with cohorts