Teaching the Teachers

The Economist, as only that magazine can, covered the vast world of teacher training and good classroom practice a couple of weeks ago with a cover story titled “How to Make a teacher.” Succinctly put — and anything else would be a mis-step considering the publication in question — the article is topnotch. A few pages of perfect summation.

I would add — with some pride and enthusiasm — that the story features our very own Sposato GSE (Scott and Orin pontificate and profess at various points in the article, and one of our Match Teacher Residents is quoted), our good friends at Relay GSE (Doug Lemov, notably) from whom we continue to learn and with whom we collaborate all the time, and several of the leading researchers and writers in our space who have pushed us over the years (Tom Kane, Roland Fryer, Elizabeth Green, Erik Hanushek).  

In short, the little corner of the teacher training and instructional design world to which we belong just flew out to the masses on the cover of The Economist. Feels pretty good.

You can see the whole article here or here (on our website).  It’s a great read, truly. Below are some of the best bits:

  • In a study updated last year, John Hattie of the University of Melbourne crunched the results of more than 65,000 research papers on the effects of hundreds of interventions on the learning of 250m pupils. He found that aspects of schools that parents care about a lot, such as class sizes, uniforms and streaming by ability, make little or no difference to whether children learn (see chart). What matters is “teacher expertise”.
  • Thomas Kane of Harvard University estimates that if African-American children were all taught by the top 25% of teachers, the gap between blacks and whites would close within eight years.
  • The new teaching schools [Sposato GSE, Relay GSE] believe that those skills which teachers now pick up haphazardly can be systematically imparted in advance. 
  • After finishing an undergraduate degree in education “I didn’t feel I was anywhere near ready,” says Jazmine Wheeler, now a first-year student at the Sposato Graduate School of Education, which grew out of the Match charter schools in Boston.
  • Trainees at Sposato undertake residencies at Match schools. They spend 20 hours per week studying and practising, and 40-50 tutoring or assisting teachers. Mr Gutlerner says that the most powerful predictor of residents’ success is how well they respond to the feedback they get after classes.  “We have thought a lot about how to teach 22-year-olds,” says Scott McCue, who runs Sposato. He and his colleagues have crunched good teaching into a “taxonomy” of things to do and say.

Advanced Placement Exams: The Achievement Gap in Boston (Part III of III)

In this third and final blog on the AP, I’ll cover the AP exam participation and performance gap in Boston and our ongoing work at Match to close that gap.  I am struck – and have always been struck – by how little debate is devoted to the glaring race- and income-based achievement gaps that characterize the AP in our city. It’s a topic worth talking about. 

The Exam High School versus Non-exam High School Gap on the AP in Boston

The table below shows the number of upperclassmen in Boston’s high schools who took at least one AP exam in 2015, and it shows the number of AP exams that resulted in a passing score of 3 or higher.

I split out Boston’s high schools between exam high schools and open-enrollment high schools. Boston has three exam high schools that admit students based in large part on an academic entrance exam. 

Notice that Boston’s open-enrollment (non-exam) high schools account for about three-quarters of Boston’s high school students, but account for only 11% of exams (282 exams total) with a passing grade.

AP Exam Participation and Performance in Boston High Schools (2014-15)
 Total # of 11th & 12th GradersTotal # of AP Exam Test TakersTotal # of AP Exams Passed
BPS Exam High Schools182118592277
BPS Open-Enrollment High Schools59901013282
All BPS High Schools781128722559

The table below tells the same story with different data. It lists the participation and pass rates on AP exams for all Boston’s exam high schools and Boston’s open-enrollment high schools.

AP Exam Participation and Performance in Boston High Schools (2014-15)
 Participation Rate (# of Unique AP Exam Takers/# of 11th and 12th Grade Students)Passing Rate (% of AP Exams Taken that Resulted in a Passing Grade of 3+)
BPS Exam High Schools102%*66%
BPS Open Enrollment High Schools18%18%
All BPS High Schools 37%51%
*The participation rate at BPS exam high schools exceeds 100% because of the inclusion in the data set of 10th grade test takers at those schools.

In short, AP success in Boston is concentrated heavily in our city’s three exam high schools. A relatively small number of students in our city’s open-enrollment high schools take or pass AP exams.

The Race and Income Gap on the AP in Boston

Household income strongly predicts performance on AP exams in Boston. Forty-three percent of BPS high school students are economically disadvantaged, meaning they qualify for food stamps, Medicaid, or certain other public subsidies. These students take AP exams at less than half the rate of their wealthier peers. And when they do take AP exams, they pass at a substantially lower rate. Here is the data: 

AP Exam Participation and Performance in Boston Public High Schools by Income (2014-2015)
 % of BPS High School Students% of All BPS AP Exams Taken By the SubgroupPass Rate (AP Exams Passed/AP Exams Taken)
Economically Disadvantaged Students43%30%38%
All Other Students57%70%51%

Now consider the statistics when sorted by race. Per the table below, African-American and Latino students account for 38% and 34% of BPS high school students, respectively, but they account for a smaller percentage of tests taken or passed.

AP Exam Participation and Performance in Boston Public Schools by Race (2014-2015)
 % of BPS Students % of All BPS AP Exams Taken By the SubgroupSubgroup Pass Rate (AP Exams Passed/AP Exams Taken)

The Challenge Ahead and Match’s Work to Date

Currently in the US, approximately 10% of students from low-income backgrounds and 4% of African-American students graduate high school having passed at least one AP exam. 

In Massachusetts high schools, the statistics are similar. Approximately 13% of high school graduates from low-income backgrounds and 3% of African-American high school graduates in Massachusetts have passed at least one AP exam.

Our schools – nationwide and in Massachusetts – are simply failing to prepare students of color and students from low income-backgrounds for the AP exam.

At Match, we have been chiseling away at the AP participation and performance gap for the last 15 years, and we will stay at it, I predict, for another decade or two. 

Our ambition is to create a PreK-12 school in which 90% of our student body – a student body that is similar to the student body in Boston’s open-enrollment schools – take and pass at least one AP exam by the time they graduate. It won’t happen overnight.  Our AP job will be made easier over time because we now enroll most of our students in early elementary school (our school is now PreK-12).  When we started 15 years ago, we enrolled students in 9th grade and only operated a high school, so we had far less time with our students than we do now to get them to AP-level work.

On the road to our ultimate AP destination, here is where we stand now. 

AP Participation rate (# of Unique Exam Takers / # of 11th and 12th grade students) AP Pass Rate (# Exams Passed/# Exams Taken)
All BPS High Schools 39.50%50.00%
BPS Open-Enrollment High Schools 20.90%15.40%
BPS Exam High Schools 97.40%68.80%
Boston Commonwealth Charter High Schools 47.60%28.40%
Match Charter Public High School 67.20%42.10%


To the data in the table above, I can add a few more statistics from our high school:

  • Over the past three years, 72% of our juniors and seniors have taken at least one AP exam each year.
  • Over the last three years, 57% of our high school graduates have passed at least one AP exam during their high school career.
  • Over the last three years, 80% of our AP calculus exams resulted in a passing grade, and 50% of our high school graduates took AP calculus. 

We have a long way to go before achieving our aspiration of a high school where 90% of our graduates take and pass an AP exam. The work is hard and we are humble, as ever, about finding new ways to help our students reach the AP bar. But we have made progress, and we believe that our students can and should master the AP with very few exceptions.

As we labor on at Match High School towards our AP goals, we hope that others will stand behind the AP as a good exam, design their schools to meet its demands, and decry the low rates at which students of color and students from low-income backgrounds take and pass these exams. 

A Note on Sources: With two exceptions, the data in this blog series comes from the College Board (particularly the 2015 National Summary of AP Performance and the 10th Annual AP Report to the Nation) and statewide reports from the Massachusetts Department of Elementary and Secondary Education, particularly the report on AP participation and AP performance. Statistics on the percentage of students attending schools with an AP program came from AP at Scale, published by the American Enterprise Institute. Statistics on the performance of Match students came from an internal Match report. Thanks also to Mass Insight for the input they offered on this blog and for their work with Match and others in Massachusetts to drive AP quality in high school. 

Advanced Placement Exams: The National Achievement Gap (Part II of III)

In Part I of this blog series, I reviewed AP exams and why we like them at Match.

In Part II of this series, I will cover the hard truth that most American high school students never take or pass an AP exam, and AP exam participation and performance varies alarmingly by race and income. 

The AP Participation Gap

Few high school students take an AP course or exam.  In 2013, just 32% of graduating high school seniors had taken at least one AP exam. AP participation rates are even lower among students of color and among students from low-income backgrounds.

Per the table below, only 19% of graduating seniors from low-income backgrounds have taken an AP exam. The analogous statistic among graduating seniors from middle class or wealthy backgrounds is 45%.

Percentage of 2013 High School Graduates Who Took At Least 1 AP Exam  
Low-income graduates 19%
Non low-income graduates 45%
All graduates 32%

The rates at which students take AP exams also vary strongly by race. For example, African-American students account for 15% of high school students in the US but only 7% of AP tests taken. 

2015 AP Exam Participation by Race     
% of High school graduates58%19%15%6%
% of all AP exams taken 54%17%7%15%

The low AP participation rate among African-American students is not because they attend schools without AP courses. In fact, African-American students attend schools with normal levels of AP programs. They are simply systematically excluded from these courses. Here is the data:

% of Students Attending Schools Offering 1 or More AP Courses (2012) 
All students89%
White 87%
African-American 88%
Asian 95%


The Performance Gap

Students from low-income backgrounds and African-American students not only take AP tests at lower rates, they also fare poorly on the tests when they do take them.

First consider AP performance among students from low-income backgrounds. Among high school students who took at least one AP exam in 2013, low-income students were far less likely to pass a test. Here is the data: 

Percentage of 2013 AP Exam Test Takers Who Passed At Least 1 AP Exam 
Low-income students 48%
All other students65%

Low participation rates and low pass rates among low-income students combine to create a situation where a middle- or upper-income student is three times more likely to pass an AP exam than a low-income student. In all, just 10% of low-income students who graduate high school pass at least one AP exam. 

Now consider success rates on the AP by race. Again, African-American students lag behind by an alarming degree. The chart below shows pass rates in 2015 on the AP by racial group. White and Asian-American students are not only more than twice as likely as African-American students to take an AP exam, but they are also twice as likely to pass the AP exams they take.

Low participation and pass rates among African-American students are such that African-American high school students, while 15% of the overall student population, account for only 4% of students who pass at least one AP exam in high school.

2015 AP Exam Pass Rate by Race  
White 64%

In Part III of this blog series on the AP, I will cover the scene here in Boston (hint: the AP participation and performance gaps by race and income are enormous in our city) and our ongoing work at Match to close those gaps.

A Note on Sources: With two exceptions, the data in this blog series comes from the College Board (particularly the 2015 National Summary of AP Performance and the 10th Annual AP Report to the Nation) and statewide reports from the Massachusetts Department of Elementary and Secondary Education, particularly the report on AP participation and AP performance. Statistics on the percentage of students attending schools with an AP program came from AP at Scale, published by the American Enterprise Institute. Statistics on the performance of Match students came from an internal Match report. 


Advanced Placement Exams: The Test Itself (Part I of III)

Students across the country – including students at Match High School – took this year’s Advanced Placement (AP) examinations in May. We value AP courses at Match and work hard, from kindergarten through high school, to prepare our students to do well. 

I sat down to capture our thoughts on the test and the debate that surrounds it. This is Part I of a three-part blog series about AP exams. This first blog provides an overview of AP exams and explains why we like them. My aim is to give you a concrete view of the test and of the content and skill mastery that it demands of students.

Part II covers the large differences – by income and by race – in the rate at which students take and pass the test. Regrettably, millions of high school students – often low-income, students of color – attend schools that fail to prepare them for AP exams. 

Part III covers the race- and income-based gaps in AP exam participation and performance in Boston. In this blog, I will also comment on where we stand at Match in trying to close the AP participation and performance gap.

The AP 101: A Tour of the Exam

AP exams are three-hour, subject-specific exams administered by the College Board each May. In 2015, the College Board administered approximately four million AP exams in 36 different subjects. 

AP exams cover widely the humanities (history, literature, economics, psychology, art, and foreign languages, etc.) and the sciences and math (calculus, chemistry, biology, physics, statistics, etc.)

In 2015, the five most frequently administered AP exams were English language and composition, United States history, English literature and composition, calculus AB, and government and United States politics.

High school students typically take an AP exam after enrolling in a year-long AP course in 11th or 12th grade. Some schools offer AP courses to 9th and 10th grade students. The full cost to take an AP exam is $92. But through a combination of federal, state, and school subsidies, students from low-income backgrounds can take AP exams at no cost or significantly reduced cost. 

All AP exams are scored on a scale of 1-5. According to the College Board, a student who earns a 5 on an AP exam is “extremely well qualified” in that discipline, while a student who earns a 1 on an AP exam earns “no recommendation.” 

Students can earn college credit and enroll in more advanced classes once in college, based on their AP exam performance in high school. Individual colleges and universities establish their own policies for AP-based early college credit. 

The College Board considers a score of 3 or higher as passing, and 50%-60% of AP exams in any given year earn a passing grade of 3 or higher. In 2015, 57% of the more than four million AP exams administered received a passing score.

AP exams consist of multiple choice and free response questions. 

Below, I lay out a few questions from different tests. The first two questions are multiple choice questions.  The last two questions are free response.

Please read these questions carefully since they will give you a foundational sense of the rigor of the AP exams.

Note: We are including these sample questions in isolation from the remainder of the exam in which they originally appeared. Students would be answering these questions based either on the coursework they completed in the related AP course or based on passages provided in the exam. 

Screen Shot 2016-05-24 at 3.09.37 PM.png

Why We Like AP Exams

At Match, we are devout fans of AP exams. We design the academic side of our schools, starting in early elementary school, with the AP standards in mind. We take this approach for three reasons:

1.    AP Exams Require Content Mastery

AP exams require mastery of content. They require expertise of our students. Consider the sample AP biology question above. To answer this question correctly, a student needs flat-out mastery of biology content – pancreatic hormones, target cells, and cellular transduction.

Why is content mastery good? Why do we like the AP exam for unapologetically requiring content mastery in its subject matter? 

First of all, content mastery matters because it actually matters to careers. Our high school students are, after all, future doctors, engineers, computer scientists, novelists, French teachers, and statisticians. They need to know their material. 

Second – and this is counter-intuitive – students (in fact, all human beings) develop higher-order skills most easily when they know material deeply. Baseline knowledge sets that stage for the exercise of analysis, inference, argument, and synthesis. For example, for a student to be a great historian – one who can make claims with evidence, one who can think originally and critically – she needs first to develop an encyclopedic command of history; to immerse herself in history itself. The AP, by demanding content mastery, correctly understands that content knowledge precedes and enables the development and application of higher order skills.

2.    AP Exams Require Higher Order Academic Skills

In addition to testing content mastery, the AP simultaneously challenges students to be skillful. The tests – across all subjects – require of students the higher order intellectual skills that matter in college and careers. At every turn, the AP asks students to read complex original texts, interpret seemingly ambiguous data, write precisely, argue with evidence, and so on.

Take the AP US History exam as an example. 

This exam includes a Document Based Question (DBQ) in which students are asked to read a prompt, synthesize evidence from historical documents and their own content knowledge, and then write a coherent essay in response to that prompt. 

A recent exam included the following prompt: 

“Explain the reasons why a new conservatism rose to prominence in the United States between 1960 and 1989.” 

To answer the prompt, students were given seven documents, including a quotation from Jerry Falwell, an excerpt from a Republican Party platform, and a statistical table tracking the unemployment rate from 1961 to 1988. 

To perform well, students needed to function as serious historians. They needed to be able to read and interpret the written texts and charts, construct a convincing argument grounded in evidence, and write a coherent essay. All in 60 minutes.

In all, the AP test – in addition to testing baseline content mastery – also requires students to be skillful scholars. 

3.    AP Exams prepare kids for – and are likely predictive of – success in college. 

Not surprisingly, given how they test both content mastery and academic skills, AP exams are closely correlated with students’ success in college. This is the third reason we admire the AP.

The rates at which high school graduates finish college are staggeringly low. Just 58% of high school graduates who enroll in a four-year college graduate within six years. Just 20% of high school graduates who enroll in a two-year college graduate within three years. College success rates among low-income high school graduates are even lower.

Enter the AP exam. When students take an AP exam – better yet, when they pass an AP exam – their trajectories in college improve. And though the evidence isn’t causal, the evidence is compelling enough to lead us to think that AP exams, when accompanied with the appropriate amount of student support, are predictive of college success. A 2013 study determined that even when controlling for prior academic achievement, demographic variables and school-level variables, students who took one or more AP exams were more likely to graduate from college in four years than students who hadn’t taken any AP exams by 9 percentage points. Students who scored higher on AP exams were even more likely to graduate in four years.

In short, we like AP exams because they make no bones about requiring content mastery, because they demand higher order intellectual and academic skills of students, and because they give students an edge in college.

In the next blog post, I will cover the alarming discrepancies by race and class in how frequently high school students participate in AP classes and pass AP exams. 

A Note on Sources: With two exceptions, the data in this blog series comes from the College Board (particularly the 2015 National Summary of AP Performance and the 10th Annual AP Report to the Nation) and statewide reports from the Massachusetts Department of Elementary and Secondary Education, particularly the report on AP participation and AP performance. Statistics on the percentage of students attending schools with an AP program came from AP at Scale, published by the American Enterprise Institute. Statistics on the performance of Match students came from an internal Match report. 


Introducing Big and Small (and other notes on our programming)

Stig Leschly here.  I’m the CEO of Match.  I have posted on this blog from time to time. This is a short post to let you know about a few changes underway here on the blog.

As long-time readers know, Mike Goldstein, our founder, launched and wrote most of entries on Puzzl_Ed for many years, until he took a new job a few years ago.  Since then, Mike has still been posting on Puzzl_Ed, as have I, but the blog has slowed down.

To revive it, I am taking it over week-to-week.  I’ve also given it a new name: “Big and Small: Thoughts from the crossroads of practice and policy in education.”

The entries on the blog will remain, as always, succinct, full of substance, and prone to humor. And Mike, who remains close to Match, will still blog from time to time.

Also, all the content from Puzzle_Ed is still here. 

In the next several days, we will post the first blogs to the new Big and Small – a three-part series on Advanced Placement exams. If you would prefer not to follow Big and Small, you can choose unsubscribe of course, but I hope you will stick around. And if you like what you’re reading, I hope you’ll share it. 

Many thanks,