The Education Gadfly Weekly

More than a dozen states are designing, piloting, or implementing some form of through-year (or through-course) assessment as a replacement for the traditional end-of-year test that has long been used for federal accountability. Among the latest entrants are Kansas and Texas, both led by respected chiefs known for taking accountability seriously. Through-year assessment is appealing because it promises to address multiple longstanding frustrations with existing state tests. But growing enthusiasm for it rests on a shaky assumption: that a single assessment can simultaneously satisfy accountability requirements, improve instruction, provide timely feedback, reduce testing burden, preserve local flexibility, and produce valid statewide comparisons. It slices, it dices, it even juliennes!

Complicating matters further, there is no single through-year blueprint. The term itself belies how varied these systems are. Instead, states are experimenting with different ways of structuring what is tested at each point in the year and how those results are used. In some states, all specified content is sampled across a series of tests and combined into a single annual determination of proficiency. In others, only portions of standards are assessed at each administration, with earlier results either fully incorporated into the final score, selectively weighted, or used primarily for informational purposes. Still other approaches rely primarily on the summative exam, with earlier tests serving mainly to support instruction rather than accountability. Policymakers and practitioners often use the same term to describe very different testing models.

Kansas’s approach, for example, is built around quarterly assessments that sample standards across the school year. Districts have some discretion as to how certain standards are incorporated into the testing sequence, but the system is designed to generate a single accountability score by aggregating performance across all four administrations. Texas, by contrast, is pursuing a program that gives districts greater flexibility over when and how certain assessments are administered, particularly at the beginning and middle of the year. Both states are trying to make state testing more useful for instruction while preserving its role in accountability. Yet for all the optimism surrounding these models, the empirical evidence on through-year assessment remains quite limited.

The more states try to have it both ways, the more they run into the tension of serving multiple, competing purposes at once. It is the same dilemma captured in the classic Saturday Night Live commercial for “New Shimmer,” a product advertised as both a floor wax and a dessert topping. Through-year assessment has something of a New Shimmer problem: It is being asked to function simultaneously as an accountability instrument and an instructional tool, as a system for comparability and a system for flexibility—design goals that do not naturally coexist in a single product.

Assessment experts have been warning about this dynamic for years. In a 2023 report, the Center for Assessment outlined ten key considerations for states pursuing through-year systems, many of which reflect the tradeoffs inherent in assessment design. More instructional utility often comes at the expense of comparability across students and schools. Additional testing windows may yield more information but increase logistical complexity. Faster feedback requires sacrifices elsewhere in the system, most commonly in the depth and breadth of what the assessment can cover.

These tradeoffs are not simply technical problems that can be papered over through more sophisticated psychometric design. They reflect a structural tension between systems built to support comparability and those designed to support instructional responsiveness. The expectation underlying many through-year proposals is that careful design choices can manage or smooth over this tension. But no matter how much ingenuity is brought to the ingredients list, a floor wax cannot be a dessert topping.

Indeed, the demands of accountability and instructional relevance pull in opposite directions. The more an assessment system is tailored to teaching and learning, the more it must reflect the diversity of classroom pacing, curriculum, and local priorities, potentially generating a wide array of tasks and measures. Accountability, on the other hand, depends on comparability across students as well as schools and districts, which requires common items or tasks aligned to a shared set of academic content expectations. Trying to make one system do too many things is a recipe for indigestion.

In theory, a summative test could be distributed across multiple administrations so that no single sitting carries the full weight of measurement. This idea is not new. It was, after all, a central motivation behind the original PARCC design. But that approach was never fully implemented, in part because states worried about unintended effects on curricular pacing and sequencing, as well as the added logistical and time demands of multiple testing windows.

That history points back to a familiar constraint rather than a new solution. The challenge, then, is not a lack of good intentions or even a lack of innovation. It is that state assessment systems are being asked to solve multiple problems at once, requiring choices that satisfy no purpose fully and inevitably sacrifice elements of each. In the process, states may be required to spend more time, money, and resources on unproven assessment models—and probably add to students’ total testing burden—at a time when there is little, if any, appetite for doing so.

Through-year assessment may yet prove valuable in some contexts. With strong leadership at the state level, Kansas and Texas are two worth watching closely. But policymakers should be wary of promises that sound too good to be true. The risk is that states mistake an unavoidable tradeoff for a design problem that can be engineered away.

Editor’s note: This discussion between members of the Fordham research team and Kari Kurto, The Reading League’s National Director of Policy and Partnerships, comes in the wake of Fordham’s recent survey, From the Teacher’s Desk: A Science of Reading Progress Report. Questions come from the Fordham research team.

1. Which of the report’s findings or results do you find most encouraging? And which is most troubling?

There is a lot to be encouraged by here. The growing understanding of phonics and word recognition among teachers is promising, as is the finding that 81 percent of respondents recognize the importance of breaking words into parts to support struggling readers. We were also especially glad to see the report include information on understanding the needs of students with dyslexia and English learners.

At the same time, several findings are deeply concerning. We are troubled by the continued finding that only 2 percent of teachers report learning evidence-aligned reading practices during their pre-service training, as opposed to the training they receive after years of teaching. This means the overwhelming majority of teachers have been entering classrooms without the knowledge they need. It is also troubling that 56 percent of teachers still associate dyslexia primarily with letter reversal. Finally, we are also concerned by the “science of reading gap” in high-poverty schools and at intermediate and secondary grade levels.

2. To what extent did the report confirm your expectations? Was there anything that surprised you about teachers’ responses?

Most notably, the report captures just how far the field has come in a relatively short period of time. One example that was encouraging to read was a classroom application scenario in which a student was having difficulty reading an unfamiliar word. The report states that “81 percent of teachers correctly indicate that the best strategy is to help the student break the word into parts to identify syllables or phonics patterns.” We celebrate this as tremendous growth in our collective understanding of how to support struggling readers.

Although there has been meaningful progress in shifting beliefs and practices, implementation remains uneven. Because of the frequent media attention on evidence-aligned phonics instruction, we expected to see progress in aligning word-reading practices with the science of reading. However, the survey confirms that we still have a long way to go in ensuring that all components of literacy are understood and implemented well. For this to happen, it is important for everyone, from legislators to researchers to the media, to focus on aligning all aspects of reading and writing instruction to the research.

It was surprising to see teachers’ comments on third-grade retention. In speaking with state leaders about this policy, it is often contentious. Yet the report suggests that implementation may be weakening the policy in practice.

We were also surprised that teachers and the report developers saw curriculum-based training as synonymous with training in the science of reading. While some of these programs are educative and teach teachers concepts, and support their understanding of the curriculum, these trainings are often not focused on building knowledge in the science of reading. Some of these curriculum-adjacent trainings may include practices that run counter to the science of reading, and several will only cover aspects of word reading, which may further reinforce the misconception that the science of reading focuses solely on word recognition.

3. After reading the report, what do you see as the biggest takeaway for the science of reading movement? What should we be doing differently?

The report makes clear that we have made real progress, but the work is still far from done. The biggest takeaway is that, while awareness of the science of reading has increased significantly, the typical elementary school teacher’s understanding of what it encompasses remains too narrow.

At The Reading League, we work hard to convey that reading science supports comprehensive, evidence-aligned literacy instruction, including explicit and systematic phonics and decoding, but also phonemic awareness, fluency, morphology, multisyllabic word reading, vocabulary, oral language development, discourse, writing, and reading comprehension. Our Curriculum Evaluation Guidelines identify 80 aligned practices across these domains. When the science of reading is framed narrowly as a phonics initiative, it can inadvertently validate the critics who argue it represents a narrow view of literacy, and it leaves teachers without a complete picture of what their students need. Going forward, the movement must do a better job of communicating this full picture.

4. One of Fordham’s core beliefs is that traditional reading comprehension strategies have diminishing returns—and that the single, most powerful “strategy” is a curriculum that systematically builds students’ knowledge of the world (and the associated vocabulary). Can you help me understand how The Reading League sees this question?

It is true that solid, empirical research supports the importance of building knowledge to enable comprehension. When a student develops a schema for a given topic, they can more easily comprehend more complex texts on that topic. This is a gateway to rigorous expository texts, which is an important goal.

However, when you elevate one strategy as the single most powerful, educators and leaders may latch onto it, believe their students will succeed based on that evidence, purchase a curriculum, and assume the problem is solved. With this view, we are concerned that the pendulum is swinging too far and that educators may abandon strategy instruction, which is also important when taught in an evidence-aligned manner.

In fact, even the meta-analysis cited in the report The Role of Background Knowledge in Reading Comprehension: A Critical Review supports both knowledge-building and strategy instruction.

“Findings from this review suggest that children would benefit from exposure to background knowledge in a specific, explicit and sequenced way (Connor et al., 2017; Kim et al., 2021): a so-called ‘knowledge-rich’ curriculum (Hirsch, 2019) in addition to teaching of comprehension strategies such as summarizing.”

The three words “in addition to” are critically important. What happens when a student learns deeply about birds but then faces a reading passage about mitosis? Their knowledge about birds does not transfer. So while a deep understanding of the topic and related vocabulary is always the hope, it makes sense to give students some automated strategies that can help them parse and comprehend texts on unfamiliar topics.

To be clear, the research does not support teaching strategies through short, disconnected texts, as was the case in the balanced literacy era. However, it does suggest that certain strategies can be an effective source of leverage for students when they are applied in the context of an ELA curriculum that intentionally focuses on building knowledge, or through direct application in science, history, and other content area instruction.

Strategies are particularly effective when they are not taught in isolation but build upon one another using a framework like Dr. Kay Wijekumar’s Knowledge, Acquisition, and Transformation (KAT), which has a strong evidence base (Wijekumar, K et al., 2012). This is not a single strategy instruction approach, but an approach that leverages text structure to help students more easily understand the text’s essential ideas. From there, the student uses this understanding to develop a summary, which is then used to further students’ reading comprehension and to help them make inferences more easily. When practiced consistently in across texts, students can master those routines and apply them on their own.

Strategies and scaffolds to support students’ comprehension when reading rich and challenging texts should not be omitted from our instruction.

5. It’s obviously difficult to gauge teacher knowledge via survey. Is there anything else that you wish we’d asked teachers? What’s the most important question that you think remains unanswered?

Reid Lyon often shares that one thing he would have done differently during Reading First would have been to work more closely and directly with teachers. He and his team knew what worked in a research setting, yet sometimes it was not actually able to be implemented.

Thinking through this lens, it would be helpful to know from teachers what has been the most helpful knowledge to gain, where implementation gets tricky, and what can be done to sustain this work. This guidance cannot come only from top-down sources, but must be a collaborative and reciprocal process among educators, educational leaders, and researchers to avoid initiative overload and burnout.

There are still several unanswered questions, of course. When The Reading League was named, the word “league” was chosen very intentionally. Not one of us knows everything. It takes researchers, experts, educators, leaders, media, and legislators to tell a complete story of what will work when it comes to teaching kids to read.

It is our responsibility to ensure we remain grounded in the evidence and do not listen to ego or sensationalism that might take our eyes off this important prize. As Neil deGrasse Tyson said, “The good thing about science is that it’s true whether or not you believe in it.” The question before us, then, is whether we are willing to break down silos and row in the same direction—toward the evidence and the science—so that our nation’s literacy outcomes do not continue to decline, but instead grow stronger for generations to come.

OpenAI, Anthropic, and SpaceX are all planning to go public in coming months, which will create a new crop of billionaires and centi-millionaires—and likely Earth’s first trillionaire. Here’s how Nan Ransohoff put it in a Substack post that’s been making the rounds:

Hundreds of billions of dollars in new philanthropic capital will soon become liquid. The OpenAI Foundation holds 26 percent of OpenAI, worth about $220B at today’s valuation. Anthropic’s seven co-founders have pledged to give away 80 percent of their wealth and have instituted the most aggressive donor matching program for employees in tech history.

She does some back-of-the-envelope math and comes up with an estimate of $50 billion in new philanthropic spending every year. This is six times the annual spending of the Gates Foundation.

My question is how education reformers can convince the new philanthropists to invest in educational improvement. I worry that it may not be easy.

Note, for example, Ross Douthat’s off-hand comment in his New York Times column about this coming gusher of cash:

At its best, the infrastructure established by figures like Gates delivered effective efforts to reduce poverty and fight disease; at its worst, it threw money after fashionable political causes and education fads.

“Threw money after…education fads.” Harsh…but kinda fair?

Or consider what Nan thinks the newly minted zillionaires are going to care about, coming as they do from the world of AI:

Belief 1: AI is fundamentally reshaping civilization and we need to make sure that goes well. There are two major questions at hand: (1) How do we ensure we make it through the AI transition? And (2) if/once we do, how do we ensure humanity flourishes? The first is more downside-oriented, concerned with catastrophic or at least major risk (biosecurity, loss of control, etc.). The second—human flourishing—is more upside-oriented. Flourishing here is broader than the usual short list. It’s longevity and economic abundance, yes—but it also extends into aesthetics, civic life, moral imagination, relationships, and the cultural infrastructure that gives life meaning.

Belief 2: Speed matters. The next X years are critical. There’s tremendous concern especially around the group of risks driving question (1) described above. The importance of speed in mitigating these risks is critical if not existential.

Speed is not something that education, or education reform, is known for. After all, as Sam Altman said (defensively but also accurately) back in February, “It takes about 20 years of life—and all the food you consume during that time—before you become smart.” I’m a big believer in starting reform efforts early, especially in elementary school, but the payoff doesn’t come for a decade or more. Some of these folks will worry that we’ll all be replaced by AI overlords by then.

“Human flourishing,” on the other hand, is something that great schools can definitely contribute to. And it makes sense, especially for high schools and colleges, to worry deeply and urgently about helping graduates through the “AI transition,” both in terms of preparing for jobs that are unlikely to be replaced by bots and in developing the “moral imagination, relationships,” etc. that will allow them to flourish along the way.

But does that mean we need to transform secondary and postsecondary education entirely? Double down on the liberal arts? Lean into career and technical education in fields that will be hard to automate? Everyone has opinions (myself included), but nobody knows for sure.

And what about the meat-and-potatoes school reform agenda? Raising standards, expanding high-quality choices, supporting excellent teachers, spending resources more efficiently and effectively? How do we get the AI billionaires to care about all of that? Especially at a time when student achievement is in the toilet and the conventional wisdom is that “nothing works,” that past philanthropy has “thrown money after education fads” and has little to show for it? I don’t think that’s true. We made a huge amount of progress from the late 1990s till the early 2010s, thanks in part to philanthropy. And lots of things work—even according to stodgy scholars—including charter schools, pre-K, merit pay, and more. But in the competition for philanthropic funding, we are starting with some serious disadvantages.

Editor’s note: This was adapted from an article first published on Mike’s Substack, SCHOOLED.

Choosing a college major requires young people to balance their interests, academic ability, future job prospects, and expected financial return. Some majors pay off right away with solid employment outcomes and high earnings, while others take years to fully realize returns—if indeed they ever do. A recent descriptive study sheds light on one little-studied influence on the choice of a college major: family background.

Economists Margaret Leighton and Jamin Speer of the University of St. Andrews in the UK and the University of Memphis, respectively, use data from the American Community Survey (ACS) and Higher Ed Research Institute’s College Senior Survey to examine two proxies for family background—parental education and parental income. How do they influence a student’s choice of college major and how does each correlate to the earnings growth experienced by graduates with that major over time?

Leighton and Speer construct a dataset of college majors’ earnings profiles over the course of a career using a sample of about 140,000 students who were college seniors between 1994 to 2008. Data from ACS include detailed family characteristics. Unfortunately, while the higher ed survey data include hundreds of colleges and universities, they’re by no means representative of the U.S. college population and tend to oversample from private and selective universities. To attempt to correct for this, the researchers replicate their main analysis within two other, smaller datasets that show similar patterns to the larger analysis. They include lots of student-level controls in their model as well as control for differences across institutions. One limitation is they can’t follow the same individuals past their senior year of college, so they only know if students planned to go onto graduate school, not if they actually did. They look at career growth in five-year bands between ages 20 to 64. Their primary outcome is growth in earnings, which is the percentage change in average earnings from early in a career—roughly age 28—and later in a career, at around age 45.

First, the researchers identify the low-earnings-growth (a.k.a. “early return”) majors—such as nursing, education, and social work—and the high-earnings-growth (or “late return”) majors. The latter include medicine, chemistry, biology, and physics.

They find that parental education and parental income, despite being strongly correlated, have different relationships with the choice of major. Parental education level has a strong and consistent relationship with this choice: students whose parents have advanced degrees are more likely to choose majors that pay off later in life. Said another way, these young people appear willing to make less money immediately after college alongside a promise to make more later in life. Parental income level, however, is less related to the choice of major. In fact, while students from higher-income families tend to choose higher-growth majors, that link is insignificant once parents’ education is included as a control.

Importantly, the overall results appear to be driven by students with graduate-educated parents going on to jobs with high growth, and particularly by those who don’t go to graduate school right after college.

Leighton and Speer describe a role model effect even if they don’t call it that. They say that “the years between college and graduate school, in which these same students earn less than students with less-educated parents highlight that students with different backgrounds may have different time preferences.” That’s likely because they have seen their own parents make the same trade off and they have confidence that it will work out for them, too. The takeaway is straightforward: Family background may shape upward mobility not only by opening doors, but also by influencing which paths students believe are worth the wait.

SOURCE: Margaret Leighton and Jamin D. Speer, “Family Background and College Major Choice: Evidence on Major Earnings Growth,” Education Finance and Policy (January 2026).

Patrick McAlister, principal of PM Strategies and former director of the Indianapolis Mayor’s Office of Education Innovation, and Shaina Cavazos, the office’s current director, join The Education Gadfly Show to discuss charter growth after the replication era. Drawing on their experience with closures and mergers in Indianapolis, they explain why authorizers and charter boards may need new approaches as enrollment declines and the sector matures.

Then, on the Research Minute, Amber Northern examines new research on content rich reading instruction and finds that while teachers are using strong foundational skills curricula, students often get too few chances to build fluency and vocabulary.

Recommended content: 

• Beyond Replication: What Responsible Charter Growth Looks Like Now —Jed Wallace, CharterFolk
• The 10-year test for durable schools —Robert Pondiscio, Thomas B. Fordham Institute
• Do Authorizer Evaluations Predict the Success of New Charter Schools? — Adam Kho, Ph.D., Shelby Leigh Smith, and Douglas Lee Lauen, Ph.D., Thomas B. Fordham Institute
• Bridging the Divide: Connecting Word Recognition and Language Comprehension in Early Literacy —Anna Jennerjohn, Sara Rutherford-Quach, Lauren J. Cassidy, Katrina Woodworth, Sarah Dec, and Dan Reynolds, SRI (2026)

Feedback Welcome: Have ideas for improving our show? We would love to hear them. Send them to [email protected]