What Is Most Likely to Make You a High School Dropout?

I was intrigued today when I came across a report from the Annie E. Casey Foundation linking students' 3rd grade reading scores on the NAEP with their propensity to graduate high school.  In other words, your ability to read by 3rd grade seems to be a pretty good predictor of whether you're at risk of dropping out of high school.  Interesting stuff!  If it's true.

The report's author, Donald J. Hernandez, doesn't say any of what I'm about to tell you in his summary.  But since he's juggling a few different variables that are all important in their own right, I thought I'd repackage his stats to give you the final word on What Is Most Likely to Make You a High School Dropout.  IS IT: Your 3rd grade reading level, as the report's author seems to imply? Your race? Or your family's history of poverty?

My unscientific ranking--drumroll, please:

1. Your reading level!  Of those surveyed, those who ranked below "Proficient" on the NAEP reading test (somewhat shockingly, over two-thirds of all fourth graders fit that miserable category in 2009, although that's really nothing new) were about 4.0 times more likely to drop out than those who scored "Proficient" or "Advanced".  The categories below "Proficient" include "Basic" and "Below Basic".
2. Poverty!  The measure used here was "has experienced poverty for at least a year, at some point during childhood", and not the commonly-used "eligibility for free or reduced lunch", so good luck trying to compare this study with any other study ever conducted.  If your family fell below the poverty line at some point in your childhood, that makes you 3.7 times more likely to have dropped out.
3. Your reading level!  Wait, didn't I rank that at #1?  Yes, but it depends how you look at it.  If you scored "Below Basic" instead of a respectable "Basic", your chances of dropping out got 2.6 times higher.  Or if you dropped from "Proficient" to "Basic", the number is 2.3.  See what I did there?  Well, you know what they say about statistics.
4. Being black or Hispanic, instead of white!  The black and Hispanic populations both came in with a 2.3-times-greater-chance of dropping out.  If you look only at those black, Hispanic and white students who had experienced poverty, this actually reduces to just a 50% greater likelihood for the kids of color.  If you've never been poor though, as a black or Hispanic kid, you're still twice as likely to drop out as a white kid.

As much as I appreciate this line of inquiry, none of it says anything about what is causing what.  There's a great argument to be made for making sure kids can read by 3rd grade--and the Annie E. Casey Foundation knows how to make it persuasively--but who's to say that 3rd grade reading and the tendency to drop out aren't at least partially by-products of a series of other factors?

For example: in at least some instances, it seems social and emotional learning is where we're giving kids the short end of the stick.  And it's often that very kind of learning--or the lack thereof--that can make reading seem hard, high school pointless, poverty self-perpetuating, and race a positive determinant.

Standardized Testing Meets the Web

Today I ran a Google search for the school where my dad teaches 3rd grade.  The 2nd result (after the school district's official school webpage) was a school profile on the GreatSchools.org website, a nonprofit whose mission is "to inspire and guide parents to become effective champions of their children's education."  There also were results from several GreatSchools competitors, including Trulia.com, Schools-Data.com, Schooltree.org, and Education.com.

My hometown of Merced, CA is one of the poorest in California, and my dad's school serves a population that is roughly representative of its district as a whole.  So when I opened the GreatSchools webpage it was no surprise to see a score of just 4 out of 10 based on his school's state test results.  To put this in the local context, only four out of 19 public elementary schools in Merced rated above a 5 out of 10, so a 4 is normal--and probably deserves congratulation.  But something about this webpage still seemed deceptive.

One problem is the webpage's visual cues.  The rating number is featured at the top-left of the school's profile page, and it appears more significant than the Yelp-style "Community Rating" (5 out of 5 stars) which sits below it.  What's more, the GreatSchools rating is formatted to look like a seal (i.e. a seal of approval, which seems to confer official status).

Placing emphasis on state test results, of course, is nothing novel.  The government does this through No Child Left Behind's carrots-and-sticks approach (which I also find problematic, incidentally).  But that practice notwithstanding, I'm concerned that emphasizing test scores online may be an impoverished if not harmful approach to GreatSchools' stated mission.

GreatSchools is serving a population of parents--mostly Generation X parents--who are incredibly demanding.  One theory holds that as the most neglected generation in American history (during their own childhoods), they are driven to the opposite extreme when rearing their own children: demanding special treatment and unprecedented oversight over their children's education.  So they will--at the very least--do a background check on the schools where they're sending their kids.  And they will most likely go online as a first step.

When they do, they'll be confronted with a number, based on test scores, that seems to tell them at least a piece of what they need to know to make the right school choices for their kids.  They may not trust that number completely, but chances are they'll be influenced to weight it more heavily in their decision-making.  Many will go the extra mile to look into other factors, but a "4 out of 10" will still hover in the back of any parent's head if that's the first impression she receives.

The government, under No Child Left Behind, at least has the sense to base its rewards and punishments on schools' Annual Yearly Progress; but when parents vote with their feet, it will be with an eye not on rates of improvement but on comparative performance looking at other schools in the state.  And they will get that information from websites like GreatSchools.

As the only economically advanced nation that primarily judges school performance via multiple-choice tests, I think we need to look closely at what we're creating, and take seriously the range of issues associated with "teaching to the test".  One particularly disturbing problem confronts me every time I come home and talk to my parents and their friends about their work: they are being systematically deprived of their sense of autonomy.  And the education they are being forced to deliver--like the one now being delivered to most low-income students in high-pressure test-focused states--is becoming less and less meaningful, even as test scores rise.

I think there is a better way.  We can keep our commitment to accountability by putting trained judges in schools, and develop an assessment system that values everything in a school that is worth valuing, including social/emotional development, creativity, the ability to express ideas and other higher-order skills.  We can serve poor kids a quality education by valuing what teachers do best.  And I think websites like GreatSchools--in partnership with governments and foundations--should lead the way.