We got our report card last month.

It didn’t come home in an intimidating, official white envelope carefully placed in our backpacks. No one had to sign it and turn it back the next day and no one hid it from their parents. No one got in trouble for poor marks, no one got a new car for getting all A’s. There were no emergency teacher-parent conferences frantically arranged to discuss the results.

None of that happened because it wasn’t that kind of report card.

Every year, the National Center for Education Statistics (NCES) measures the performance of our nation’s 4^th and 8^th graders across a wide range of academic subjects at the state and local levels. These tests are known collectively as the National Assessment of Educational Progress (NAEP). When the NAEP releases the results, they’re called The Nation’s Report Card. This April, we received our scores in math and reading for 2017.

No one got grounded and no one got any gold stars, but every time a large-scale assessment like this comes out, a large kerfuffle ensues. Not at home, and for the most part, not in classrooms or principal’s offices. The mad scramble comes one, two, or even three steps removed from where the learning happens. It happens in the offices of local and state superintendents, in local and national departments of education, in local and national editorial pages, and in reports published by education watchdog groups and public policy lobbyists.

What All the Noise is About

For better or for worse, large-scale standardized testing is how we, in the United States, determine the broad strokes of our national education policies. Critics and advocates debate the merits of standardized tests ad nauseam. Critics fault schools for teaching to the test and ignoring the non-quantifiable elements of student development and education such as social and emotional learning. Advocates insist big-picture education policy must derive directly from objective data, and consistently convince decision-makers that standardized testing is the best way to get reliable feedback on student performance.

Neither camp is entirely right, and neither camp is entirely wrong. However, what’s clear to most observers is that we’re stuck. A recent article called published on the website Getting Smart called Reflections on the NAEP: Breaking the Reactionary Cycle elucidates our national dilemma:

“This cycle of designing and administering assessments, and then responding to the results, is the dominant input to decision making related to improving educational outcomes. Decisions on maintaining or adjusting educational approaches, on setting resource allocation priorities, and on what feelings leaders convey about the public educational system emerge from reacting to the test scores. It is how we gauge whether or not specific school systems are doing right by their kids in preparing them for the future.”

That’s where we are. We test, we react, we retool policies, we implement them in classrooms across the nation, then we re-test. And do it all over again. When we strip away the competing agenda, there’s really only one question that matters:

Does it work?

Let’s take a look at the numbers and find out.

NAEP 2017: Reading and Math Results

A little background and the NAEP: conceived in 1963, developed under a grant from the Carnegie Corporation between 1964 and 1968, they conducted their first nationwide assessments in 1969. Their current model includes a random selection of around one hundred schools from each state. Schools opt-in on a voluntary basis. Past participation in the assessment has no bearing on whether a school will be included in future assessments. According to their internal reporting, the NAEP tests approximately 6,000 – 20,000 students per grade, per subject, per year. The large sample size enables them to formulate broad, generalized conclusions about the current state of learning in the United States. The current data set compares scores from 2017 to scores in 2015, and refers to scores recorded in 1990/1992 as long-range benchmarks to determine our progress across decades.

Enough preamble: to the scores.

NAEP 2017 Mathematics Results

General Info:

• Tests administered between January and March, 2017 to:
  • 149,400 fourth-graders from 7,840 schools
  • 144, 900 eighth-graders from 6,500 schools.
• Scores reported on a scale of 0-500.
  • For 4th graders, scores above 214 indicate basic knowledge, scores above 249 indicate proficient knowledge, and scores over 282 indicate advanced knowledge.
  • For 8th graders, scores above 262 indicate basic knowledge, scores above 299 indicate proficient knowledge, and scores over 333 indicate advanced knowledge.

National Assessment:

• FLAT OVERALL compared to 2015.
• Average math scores for 4^th graders:
  • 2017: 240
  • 2015: 240
  • 1990: 213
• Average math scores for 8^th graders:
  • 2017: 283
  • 2015: 282
  • 1990: 263

The good news from these numbers is that over the past thirty years, our math scores show significant improvement. Many of you may remember the panic that occurred when our students started to fall behind other developed nations in technical subjects back in the 1990s. That’s basically what caused our recent obsession with Science, Technology, Engineering, and Mathematics (STEM) curricula: most of our focus – meaning money – has been directed at playing catch up with the rest of the world in math- and science-related subjects.

Where We Need to Improve

The less inspiring news hides deeper within the data:

• Compared to 2015, scores for students in the middle and higher percentiles either improved slightly or stayed the same.
  • High performing 4^th graders – 75^th and 90^th percentile students – did not improve.
  • Middle performing 4^th graders – 50^th percentile students – did not improve.
  • High performing 8^th graders improved slightly.
  • Middle performing 8^th graders did not improve.
• Compared to 2015, scores for students in the lower percentiles declined.
  • Low performing 4^th graders – 25^th and 10^th percentile students – saw statistically significant drops of 2 points and 4 points, respectively.
  • Scores for 25^th percentile 8^th graders also saw a statistically significant drop of 2 points.
  • Scores for 10^th percentile 8^th graders did not improve.
• The nationwide percentage of 4^th grade students at or above Proficient was 40%.
  • Two states/provinces improved in comparison to 2015
  • Ten states declined in comparison to 2015.
• The nationwide percentage of 8^th grade students at or above Proficient was 33%.
  • Two states improved in comparison to 2015
  • Three states declined in comparison to 2015.

We’ll discuss these trends in a moment. For now, we’ll move on to the reading scores.

NAEP 2017 Reading Results

General Info:

• Tests administered between January and March, 2017 to:
  • 148,400 fourth-graders from 7,830 schools
  • 141,800 eighth-graders from 6,500 schools.
• Scores reported on a scale of 0-500.
  • For 4th graders, scores above 208 indicate basic knowledge, scores above 238 indicate proficient knowledge, and scores over 268 indicate advanced knowledge.
  • For 8th graders, scores above 243 indicate basic knowledge, scores above 281 indicate proficient knowledge and scores over 323 indicate advanced knowledge.

National Assessment:

• Increase in average reading scores for 8^th graders, overall, but FLAT or DECLINING for 4^th
• Average reading scores for 4^th graders:
  • 2017: 222
  • 2015: 223
  • 1992: 217
• Average reading scores for 8^th graders:
  • 2017: 267
  • 2015: 265
  • 1992: 260

There’s not much great news in this data. The NAEP foregrounds improvements in reading scores for 8^th graders, but while the numbers they highlight are statistically significant, they’re not significant at face-value for non-statisticians: an improvement of 2 points is nothing to jump up and down about. And as a nation, it’s nothing to hang our Uncle Sam hat on. Furthermore, when we compare all three data points – 1992, 2015, and 2017 – our improvements in reading over the past thirty years are mediocre at best.

More Room for Improvement

Again, there’s some less-than-inspiring information deeper in the data:

• Compared to 2015, scores for students in the middle and higher percentiles either improved slightly or stayed the same.
  • High performing 4^th graders – 75^th and 90^th percentile students – did not improve.
  • Middle performing 4^th graders – 50^th percentile students – did not improve.
  • High performing 8^th graders improved slightly.
  • Middle performing 8^th graders improved slightly.
• Compared to 2015, scores for students in the lower percentiles fell or did not improve.
  • Low performing 4^th graders – 25^th and 10^th percentile students – saw statistically significant drops of 2 points and 3 points, respectively.
  • Scores for 25^th percentile 8^th graders did not improve.
  • Scores for 10^th percentile 8^th graders did not improve.
• The nationwide percentage of 4^th grade students at or above Proficient was 35%.
  • No states improved in comparison to 2015
  • Nine states declined in comparison to 2015.
• The nationwide percentage of 8^th grade students at or above Proficient was also 35%.
  • Ten states improved in comparison to 2015
  • One state declined in comparison to 2015.

The 2017 National Report Card: Takeaways

We’ll give our honest answer to the question posed at the beginning of this article:

Does it work?

If we’re straight with ourselves about the data we see, we get a lukewarm, anti-climactic answer:

Sort of.

Okay, that’s a terrible, depressing answer. We test, we retool, we revise, we implement new strategies, and all we get is “Meh.”

We sincerely apologize.

We’re not going to make up any numbers, though, and we’re not going to sugarcoat things. We will, however, take advice from the old Broadway hit and Accentuate the Positive: when we look at our math scores, we see a glimmer of hope. They’ve followed a general upward trend since 1990. Unfortunately, that’s about all the hope there is to find. A close look at the data shows both math and reading scores for 25^th and 10^th percentile 4^th graders dropped between 2015 and 2017. What that means is that our most vulnerable students are going in the wrong direction: backward. In contrast, math and reading scores for 75^th and 90^th percentile 8^th graders improved. Meaning the students going in the right direction – forward – are students who already do well and have already made it past their foundational elementary school years.

How to Interpret the Data

It’s tempting to lump the math and reading improvements in with the broad economic trend in the United States that’s clearly visible over the past forty years: wealth for the top earners has increased, while income for middle and lower earners has stayed flat. That’s a social, political, and cultural minefield beyond the scope of this article, so we won’t go there. But we do think it’s good food for thought.

What we will do is point out the fact that our current system seems to be stagnant. Our proficiency averages for reading and math appear woefully low: 40% or below for 4^th graders and 8^th graders in science and math. We’re sure you know how to read statistics, but we have to ask the rhetorical question(s):

Seriously? Less than half our students are proficient at reading and math?

Maybe you’re asking yourself the same thing. It’s a disturbing question with a disturbing answer: the numbers say yes, less than half our students are proficient in the most basic subjects. That makes us question the validity of the entire standardized testing system. We’ve used standardized tests for close to fifty years now, and we have NAEP data for the last thirty. What we have to show for it is 5-7 points of improvement in reading scores and 20-27 points of improvement in math scores.

We Can Do Better

Perhaps it’s time to start moving away from the traditional standardized testing model and prioritize a new set of objectives, such as those suggested in the article cited above, Reflections on the NAEP –  Breaking the Reactionary Cycle:

• Social and Emotional Learning teaches students to manage emotions, communicate effectively, and maintain productive relationships.
• Deeper Learning stresses six competencies: mastering core academic content, thinking critically, working collaboratively, communicating effectively, learning to learn, and developing a proactive academic mindset.
• Family Engagement improves student performance in all subjects across all social, racial, and ethnic groups.
• 21^st Century Skills focus on developing student collaboration, critical thinking, communication, and creativity.

Politicians, policy-makers, and administrators at the federal, state, and local level can have their cake and eat it, too: they need objective data before they can redistribute millions of dollars toward these new objectives. Fortunately, the data are out there. These objectives are one-hundred percent data-driven and evidence-based. Decades of research in education shows a myopic focus on standardized testing results in modest improvements across subject areas, while an integrated, holistic focus on the social, emotional, and family lives of students leads to significant improvements across all subject areas. The final hurdle to clear lies in convincing policy-makers to allocate resources toward these new approaches. The irony of the situation is rich: once public policy shifts toward educating the entire student – including social, emotional, and practical skills – we’re willing to bet standardized test scores will skyrocket.