Why statistics still matter

I still believe that stories and real people are what change minds and motivate hearts. I’ve seen it so many times…in an argument, observers’ eyes glaze over when the combatants start to hurl data at each other, and if anyone ever mentions what a “multivariate regression analysis” has supposedly proven beyond (almost) (because you know that researchers never say never) any reasonable doubt, you’re sunk. They’ll never listen to another word again.

But when it comes to seeking guideposts for our own work, deciding on policy priorities, allocating scarce resources, and determining the extent to which we’re reaching our goals, good statistics can be much more helpful than anecdotes (which, as powerful as they are, are really, really bad bases for policymaking–I found that out the hard way when I gave legislative testimony that included the story of a legally-present immigrant student who was still ineligible for instate tuition, and then the legislature wanted to amend the bill to only cover kids like him, which would have excluded the VAST majority of students who were targets of the measure).

So what that means is that, while we must personalize our claims to social justice, in ways that compel right action, we also must be comfortable and skilled enough with statistics to keep those very compelling stories from leading us astray.

An example from Super Freakonomics (I’m kind of getting hooked on these guys), and one from my consulting work, illustrate this, probably better than I have so far.

The example from the freak-economists is truly crazy: apparently, an intoxicated walker is eight times more likely to die than an intoxicated driver. The authors certainly don’t suggest that the work of anti-drunk driving advocates has been in vain, but they do use data to convincingly redirect our attention to the dangers caused by excessive consumption of alcohol, period, not just the preoccupation with drunk driving. They point to other, less surprising, clashes between data and story–the high-profile death of a young boy from a shark attack, for example, when only 4 people are killed, worldwide, on average, by sharks each year.

Here’s how they state what I’ve been trying to, about why statistics still matter, even in this digital storytelling age:

“While there are exceptions to every rule, it’s also good to know the rule. In a complex word where people can be atypical in an infinite number of ways, there is great value in discovering the baseline. And knowing what happens on average is a good place to start. By so doing, we insulate ourselves from the tendency to build our thinking–our daily decisions, our laws, our governance–on exceptions and anomalies rather than on reality” (p. 14).

Which brings me to my consulting work, where I’ve been working with some anti-poverty agencies around assessing needs in their community. One of the realities that is quickly apparent is that, many times, perceived needs diverge from actual needs, such that figuring out what actual needs really are (because, when we’re talking about the willingness of powerful entities to recognize and address them, perception matters so much) becomes very difficult.

An example of this was the discussion in two communities about teenage pregnancy. In one area, teenage pregnancy featured prominently in key informants’ assertions about priority needs in the community and yet barely registered in the actual data about births to teenage mothers. Clearly, there was some reason that these different entities kept pointing to teenage pregnancy as an “alarming trend” (in the words of one respondent) when there was no trend that we could find, nor much cause for alarm. I’m not sure what all was driving that, but what is certain is that, if the community diverts resources from other efforts towards preventing teenage pregnancy, there may very well be an uptick in the incidence of those other social problems, some of which (in this community, lack of access to affordable, quality childcare, for example) may be, in any absolute terms, far more problematic.

In the other community, as you might guess, the opposite dynamic is playing out. Despite staff at the agency recognizing that teenage pregnancy is nearly epidemic, and despite statistics that show that fully 50% of young women ages 13-19 have given birth, hardly anyone talked about teenage pregnancy or related issues in the needs assessment process. Here is an example of a community that must confront its statistics in order to have half a chance at effectively solving the problem.

So, tell your stories, absolutely. Appeal to values, and connect with people’s hearts.

But when you sit down to check yourself, make sure that you can find, understand, and face what those not-irrelevant averages are telling you, too.

Advertisements

One response to “Why statistics still matter

  1. Pingback: online video marketing statistics | Search Engine Optimization

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s