A recent conference hosted by the Stanford Social Innovation Review and FSG Social Impact Advisors, titled "Good Measures: New Approaches to Evaluation," showcased evolving thinking about how to evaluate social impact. As usual, a lot of it centers on service delivery (as it should; end needs, ultimately, need to be met), but there are interesting tidbits to be gleaned on measuring advocacy impact.

Take a look at "Session 4: Evaluation for Learning: Creating Cultures of Inquiry," for instance. Moderator Alana Connor's (pictured) presentation, titled "The Wrong End of Science" bemoans the fact that grantees are "drowning in data; they increasingly have to prove that their programs 'make a difference'--i.e., they have to conduct summative evaluations." These tend to judge, yea or nay, whether the program "worked," rather than providing formative feedback to improve while the effort is underway. Sound familiar?

But Connor goes on to make a further comparison that I've not heard before. She argues than instead of focusing on the "back end of science," with its prerequisite for double-blind, control group fortified, random assignment methods, those of us looking to measure social impact should pay attention to the "front end of science." That means taking into account "accumulated evidence, logical analysis, and principled experimentation for yourself."