1. Signals beyond the report card
In the fall of 2026, a mother in Queens, New York, sat down to review math with her 14-year-old son and found a situation very different from the impression his report cards had given her.
According to the account she later gave in the New York Post, her son had just entered a highly regarded high school that had been his first choice. His math grades had always been good: he often received the highest rating in elementary school, finished sixth grade with a 92, had no final grade below 89 in seventh grade, and received a 93 in eighth grade. Nothing on those report cards seemed to warrant concern.
Then he scored 78 on his first high school math test. His mother decided to help him prepare for the second, only to discover that he could answer just the first few practice questions. He could not complete some basic algebra and geometry problems and even struggled with the fraction calculations taught in elementary school. She said tears welled up in his eyes as he faced those questions. His teacher subsequently called to say that, of the second test's 10 multipart questions, he had answered only 2 correctly, earning 10 points out of 100.
The mother is Adriana Aviles. A former New York City police officer, she had long been involved in school affairs and had served in leadership roles in a parent-teacher association and a community education council. She had specifically moved so her child could attend District 26, a district she trusted. She was hardly a parent who paid no attention to school. Yet she found that her assessment of her son's learning might have been far removed from reality.
What shocked her even more was what she noticed only when she went back through his records: his seventh-grade New York State standardized math test had already shown that he had not met the relevant proficiency standard.[1]
Some might ask: if he had gaps in basic math, how had he been admitted to a good high school? New York City's high schools do not all use the same admission process. Some screened programs group applicants by their final seventh-grade marks in four core subjects; when a group has more applicants than places, admission is decided by random number.[2] But the original report did not disclose the boy's high school or his specific admission route. General rules can help explain the context, but they cannot replace verification of an individual's experience.
The questions that need asking are these: why did different assessments send different signals? Were those signals explained at the time? Did the child receive help promptly?
On October 26, Aviles described her experience at a public meeting attended by the chancellor of New York City Public Schools. Five days later, the New York Post published her first-person account. Using strong language such as “deception” and “faking,” she accused the school system of concealing her child's learning difficulties behind attractive grades.[1]
A boundary matters here: this is a parent's public allegation. The individual grade records have not been fully independently verified, and the material we found is insufficient to establish that the school deliberately raised his grades. This cannot be treated as a proven case of grade falsification.
She also said that some other parents at her son's high school had reported that their children were struggling with algebra. Those reports, however, do not constitute a group of independently verified cases with the same experience. Nor is it clear whether the students concerned attended the same middle school or came from the same district.[1]
Meanwhile, the New York City Department of Education's 2024–25 School Quality Snapshot reports a math proficiency measure of 91% for his former middle school, M.S. 67 Louis Pasteur. That measure includes converted results for some students who took high school math Regents examinations early; it is not simply the pass rate on the ordinary state test.[3] Strong overall performance neither explains this boy's discrepant grades nor rules out the possibility that some difficulties went insufficiently recognized. Equally, one child's experience cannot establish that the school's grading is generally inaccurate.
Yet the question she raises deserves serious discussion: how much can an excellent report card really tell us about learning?
2. Grade inflation: when A's multiply, what does an A mean?
A mismatch between one child's high grades and actual performance is not enough to establish system-wide grade inflation. In research, “grade inflation” generally refers to grades rising over time without corresponding improvement in academic achievement.
In 2022, the US college admission testing organization ACT published a study of public high school graduates who took the ACT from 2010 through 2021. It covered more than 4.39 million students at 4,783 schools. The researchers calculated high school GPA from students' self-reported grades in core courses. After adjusting for certain student and school characteristics, the model-estimated average GPA rose from 3.17 to 3.36, without a matching improvement in ACT scores.[4]
This does not mean GPA and standardized test results should match perfectly. One reflects coursework over a period of time; the other is a test administered under particular conditions. They do not measure exactly the same things. Nor can this study's sample directly represent all US high school students. It supports discussion of the broader issue of grade inflation, but it does not establish that District 26 or the school in this case has the same problem.
Still, when classroom grades keep rising while relatively independent measures of achievement do not improve alongside them, there is reason to ask: are students learning more, or has the meaning of the grades changed?
The ACT report found that the increase in grades was especially pronounced from 2018 to 2021. But the trend began before the pandemic, so the entire change cannot simply be attributed to it.[4]
The concern about grade inflation is not just that A's are becoming more common. It is that the information conveyed by an A may no longer be as reliable as it once was.
3. Why can a low grade be hard to give?
When a child works hard but still has not mastered essential knowledge, a teacher may face several competing demands: describe the gaps honestly, recognize effort and progress, prevent the child from losing confidence, and explain disappointing grades to the parents.
These goals do not necessarily conflict. But when one grade must encourage, communicate, evaluate, and help determine future school admission, it is no longer merely a record of what a student knows.
A 2024 Education Week article cited a nationally representative survey of about 1,500 teachers conducted in October 2023. In responses about pressure to change grades during the previous year, roughly four in ten teachers reported pressure from parents and a similar proportion reported pressure from students. More than a quarter also mentioned pressure from principals.[5]
Experiencing pressure does not mean teachers actually changed grades, still less that the pressure can be shown to have caused grade inflation. But it does show that grading is not independent of relationships between schools and families, students' future admission opportunities, or teachers' working circumstances.
Alongside outside pressure, the way a grade is constructed deserves attention.
In her account, Aviles said her son could correct answers after a test was marked and receive a new score.[1] That description alone does not allow us to determine the specific grading rules, much less conclude that corrections caused his learning difficulties. But it raises an important distinction: correcting an old answer is not the same as demonstrating mastery anew.
Correcting mistakes can support learning, and one poor test performance should not determine an entire semester's evaluation. The question is whether new evidence of learning supports the higher grade. Can the student independently solve a new problem of the same kind? Can the student explain why the solution works?
Likewise, effort, participation, and progress all deserve recognition, but they are different from mastery of knowledge. If all this information is compressed into one letter, parents may struggle to know what that A actually means.
Encouraging children and honestly explaining their difficulties need not be an either-or choice. A more helpful assessment would make both possible.
4. Parents think their children have learned it. Have they?
In the fall of 2023, Learning Heroes and Gallup surveyed nearly 2,000 parents of US public school students in K–12, from kindergarten through high school.
About 89% of the parents surveyed believed their children were at or above grade level in math; for reading, the figure was about 88%. At the same time, 79% said their children mostly received B's or better. Report cards were among the sources parents relied on most often when judging their children's academic standing.[6]
Another assessment presented a very different picture. In the 2024 National Assessment of Educational Progress (NAEP), about 28% of eighth graders performed at or above its Proficient standard in mathematics.[7]
Care is essential here: NAEP Proficient is not equivalent to what parents understand as “at grade level,” nor to individual states' proficiency standards. The parent survey covered K–12, whereas the result cited here concerns eighth-grade math; the years and populations also differ. We therefore cannot subtract one percentage from the other to calculate how many parents were “misled” by report cards.
These sources are better used as a reminder that understanding a child's learning requires knowing what question each assessment is answering. Different standards should not be treated as interchangeable.
Citywide state-test results released by New York City Public Schools in August 2026 showed that 57.0% of tested students in grades three through eight met the state math proficiency standard. Among eighth graders, the figure was 47.6%, down 1.6 percentage points from 49.2% in 2025.[8] These are citywide New York State test results, not a picture of any particular district or school.
Failing to meet that standard does not mean a student cannot do basic math. Nor are standardized tests unquestionable “truth.” They have their own scope and limitations and must also be interpreted alongside specific assignments, classroom performance, and teachers' observations.
But if a student consistently earns good grades while other assessments reveal marked difficulties, further inquiry is warranted. Do the grading standards differ? Is the content being tested different? Or are there gaps in knowledge that have not yet been identified?
Aviles later acknowledged that she had not examined her son's assignments and state-test results closely because she assumed a good report card meant everything was going well.[1] That makes the problem concrete: even a parent familiar with school affairs can come to rely too heavily on the prominent grade.
5. “Set the grades aside—first tell me whether the child is happy”
Many readers educated in China will probably smile in recognition at that line. Exams, rankings, tutoring, and competition over grades are familiar parts of growing up for many people. A child who does not have to worry constantly about scores, and who has more free time and more room for the pleasures of childhood, can indeed seem enviable. That wish is real and important.
Joy in learning does not mean an absence of difficulty or pressure. The challenge of trying to understand a hard problem is a different experience from the strain of constantly worrying about rank or fearing rejection.
According to his mother, she noticed some of her son's math difficulties only after he entered high school. That gap between expectation and reality can bring new pressure of its own. We do not know whether his childhood was happy, and one episode of tears cannot establish his overall well-being. What she describes, however, is a child's frustration when facing problems he cannot solve, followed by a parent's anxiety.
Self-determination theory in educational psychology offers a useful perspective. It emphasizes autonomy, competence, and relatedness: whether a person wants to learn for reasons they endorse, feels able to master knowledge and meet challenges, and receives understanding and support in the process.
A 2022 meta-analysis combining 144 studies involving more than 79,000 students found that satisfaction of these psychological needs was associated with more positive autonomous motivation to learn. This is evidence of association, not direct causal proof that a particular educational measure will necessarily work.[9]
It can nevertheless help us understand why enjoyment in learning comes from more than ease. A child who finally understands a difficult math problem after working at it, or chooses a question and finds an answer after attempts and failures, may also gain a profound sense of satisfaction.
What is needed is support that helps children develop real ability while retaining the freedom to explore, rather than attractive grades that create the impression, “I already know this.”
Lenient grading and excessive exam preparation need not be assigned to two different countries. Children need both the security of knowing that grades do not define them and help when they encounter difficulty. We hope for a happy childhood alongside growing capability, rather than treating the two as goals of which only one can be chosen.
6. When a measure becomes a target
Education is not the only field that faces this difficulty.
In medical research, investigators often assess a treatment by how much a tumor shrinks, how a biomarker changes, or how long disease progression is delayed. These results have value, yet they do not always directly answer the questions patients care most about: will I live longer, suffer less, or maintain my everyday life?
Some such measures are used as “surrogate endpoints.” Whether, and how well, they predict benefits that matter to patients must be judged in the context of the particular disease, treatment, and evidence. Improvement in one measure does not by itself mean that every important question has been answered.[10]
Educational assessment faces a similar difficulty. Test scores, GPA, rankings, and admission rates are easy to record and compare. Understanding, independent thought, problem-solving, and the ability to keep learning after encountering difficulty are much harder to capture fully in one number.
There is nothing inherently wrong with using measurable indicators to help understand complex goals. The difficulty arises when these indicators begin to influence school evaluations, teachers' work, parents' choices, and students' futures. People may adjust their behavior to prioritize better numbers over what the indicators were meant to represent.
This brings to mind Goodhart's law. In everyday terms, when a measure becomes a target that people pursue, its reliability as a measure may suffer. This is better understood as a warning about a common risk than as a formula that must hold in every situation.[11]
Consider an everyday example: to reach ten thousand steps, someone keeps pacing around the living room before bed. The number finally reaches the target, but at the expense of needed sleep. The problem is not that walking has no value. It is that “getting the steps in” has begun to overshadow the actual goal of looking after one's health.
This also connects with the question discussed in “With More Models to Follow, Why Are We Becoming More Alike?” Reference points can help us explore, but may gradually prescribe which outcomes are worth pursuing. Models in everyday life and school assessment standards are different things, but both can influence where we direct attention and effort.
Test standards were intended to help us understand how much students have learned. When high scores become the overriding goal, teaching may increasingly center on familiar question types. Even if scores rise, we still need to ask: can a child understand and solve the problem when the question is phrased differently or placed in a new setting?
Grade inflation and excessive exam preparation appear to move in different directions, yet may have similar consequences. One can conceal learning difficulties behind attractive numbers; the other can reduce learning too heavily to training for high scores.
In one case, grades do not adequately reflect learning. In the other, learning is confined too tightly within grades.
7. What kind of assessment do we need?
One intuitive response to misleading grades is more standardized testing. Common standards can indeed provide a relatively independent reference point. Yet a single test cannot show a student's full range of abilities, and certainly should not determine a child's worth. Conversely, abolishing tests does not automatically make education more comprehensive: without other reliable assessments, schools and families may also lose tools for identifying difficulties.
The discussion we need is about more than having more or fewer exams. It is about allowing different kinds of information to serve different purposes.
Classroom grades, specific assignments, independent assessments, teachers' observations, and students' explanations of their reasoning can complement one another. Teachers' observations, too, should rest on concrete examples and evidence of learning, rather than impression alone. This information need not all be squeezed into the same score. It should help answer: what has the child learned, what remains unclear, what progress has been made, and what help is needed next?
Three directions for practical improvement are worth considering.
First, make the meaning of grades clearer. Mastery, effort, participation, and progress all deserve attention, but need not be blended into a single letter. Reporting them separately makes it possible to recognize a child's work without concealing where help is needed.
Second, make it possible to compare different assessments. When classroom grades and independent assessments differ sharply, it is not enough to hand the results to parents. They also need help understanding what the difference might mean and what they can do next.
Third, connect higher grades with new evidence of learning. Corrections have value, but correcting an old question does not establish that a student can independently handle a new one. After further learning, new questions of the same kind and explanations of the reasoning can help show whether mastery has actually been achieved.
These improvements must themselves guard against another shift: could the school's goal move from helping children learn to continually proving to parents that “we have fulfilled our responsibilities”?
Keeping parents informed and clearly recording support are necessary responsibilities. But if assessment increasingly centers on how many notices were sent, how many forms were completed, or whether parents signed, procedure may become the new target. Parents may receive more material without better understanding their children. Teachers may have more to demonstrate without more time to help.
Imagine a child who repeatedly gets fraction problems wrong. What a teacher may need is to sit down and listen: why did you calculate it that way, and at which step did you become confused? If limited time is consumed by repeatedly filling out forms, uploading records, and checking requirements one by one, that conversation may be squeezed out. An interaction once centered on “How are you thinking about this?” can increasingly center on “Have these requirements been completed?”
Improving assessment should therefore include examining which repeated procedures can be reduced, alongside any new requirements. Every additional piece of information collected should prompt the question: how will this help teachers adjust instruction and parents support their children? Clear records should serve understanding and action, not replace them. Schools must be accountable to parents, but that responsibility should ultimately rest on the help children receive, rather than simply the amount of evidence the school has retained to show it has done its duty.
For the New York family, several crucial points still need verification. How were the state-test results communicated and explained at the time? What went into classroom grades, and how did corrected answers affect the final score? Once assessments diverged, did the school and family discuss possible reasons and provide support?
The material currently available cannot answer these questions. We can use it to discuss better assessment, but cannot describe possible mechanisms as though they were established facts about this child's experience.
8. A grade should be a window, not a wall
What is distressing in this mother's account is more than her son's math performance falling short of expectations. It is her feeling that the report cards she had trusted for so long had not adequately shown her how he was learning.
Parents need to attend to the particulars of their children's learning, and schools need to make important information clear enough. Both should work together around the child, rather than waiting until problems accumulate to start asking who is responsible.
For Chinese readers familiar with examinations and grades, this story offers another perspective. There is no need simply to take sides between “strict exams” and “lenient grading.” We can ask different questions:
When a child earns high marks, do we still care about what the child truly understands? When the marks are low, can we first see what help is needed? When a child feels exhausted, are we willing to ask how much room remains for self-directed exploration?
Assessment should be a window through which we see how far a child has come and what paths lie ahead, rather than a wall that compresses a complex experience of learning, effort, difficulty, and possibility into a number.
What we need to guard against is not grades themselves, but becoming so accustomed to trusting grades that we forget to look at the child who is learning.
Author's note: why move from education to drug development?
Why discuss education in a publication that starts from medicine and science? Because the puzzles surrounding grades can also help us understand difficulties in drug development: the outcomes that matter most may not be the easiest to measure, and by the time they fully emerge, a great deal of time may have passed.
Determining whether a treatment helps people live longer and better may require long observation. Yet clinical trials designed to confirm efficacy must select a primary endpoint in advance—the outcome measure used to answer the central question—and specify key analytical methods.[12] If a measure observable earlier is chosen, we need to understand how well it predicts real benefit. If we wait for longer-term outcomes, we must also face the time cost of waiting itself.[10]
Regulators face the same tradeoffs: they must protect patients from the risks of treatments supported by insufficient evidence while also considering the cost of delayed access to effective treatment. Assessment methods and standards must improve as knowledge develops. But updating standards does not mean looking at a trial's results and then working backward to select favorable “success criteria.”
Education touches many families and therefore readily sparks discussion. Clinical endpoints, study design, and regulatory requirements may sound specialized, but they too affect which treatments we may receive and which risks we may bear. Helping more people understand these difficulties is meant to make our questions more specific, rather than lower the standard of evidence: what has been established, what remains uncertain, and what further testing is needed? Through a familiar education story, we hope more readers will be willing to approach these questions and that discussion of drug development will stay closer to the lives patients actually care about.
Update, November 2, 2026: Official overall data for the school concerned have been added, and the evidentiary boundaries between an individual case, nationwide research, and judgments about a school district have been clarified further.
References
[1] Adriana Aviles, first-person account in the New York Post, October 31, 2026. The account of the child's experience in this article comes from this parent's statement; it is not an independent finding of grade falsification.
https://nypost.com/2026/10/31/us-news/adriana-aviles-says-her-son-was-given-phony-grades-is-behind-in-math/
[2] New York City Public Schools: Screened Admissions. Cited to explain general rules for some screened programs, not to infer the admission route of the child concerned.
https://www.schools.nyc.gov/ScreenedHS
[3] New York City Public Schools: 2024–25 School Quality Snapshot for M.S. 67 Louis Pasteur and the Educator Guide for the same school year, updated May 19, 2026. Pages 8–9 explain student attribution and conversion of certain math Regents examination results. The 91% in the text uses the school quality report's definition and is not equivalent to the proficiency rate for ordinary state math tests alone.
https://tools.nycenet.edu/snapshot/2025/26Q067/EMS/
https://infohub.nyced.org/docs/default-source/default-document-library/finalized-2024-25-educator-guide-ems--1-1.pdf
[4] ACT Research, Grade Inflation Continues to Grow in the Past Decade, May 2022. The figures 3.17 to 3.36 in the text are model-adjusted GPA estimates.
https://www.act.org/content/dam/act/unsecured/documents/2022/R2134-Grade-Inflation-Continues-to-Grow-in-the-Past-Decade-Final-Accessible.pdf
[5] Education Week, “Where Teachers Say the Pressure to Change Grades Comes From,” May 24, 2024. The methodology confirms that the survey was conducted in October 2023 and included 1,498 teachers.
https://www.edweek.org/the-state-of-teaching/2024/teaching-learning/where-teachers-say-the-pressure-to-change-grades-comes-from/
https://www.edweek.org/the-state-of-teaching/2024/teaching-learning/data-5-key-insights-into-americas-teachers/
[6] Gallup and Learning Heroes, B-Flation: How Good Grades Can Sideline Parents, 2023.
https://www.gallup.com/analytics/513881/parents-perspectives-on-grades.aspx
[7] National Assessment of Educational Progress: 2024 mathematics assessment results. NAEP Proficient is not equivalent to grade-level standards defined by states or school districts.
https://www.nationsreportcard.gov/reports/mathematics/2024/g4_8/?grade=8
[8] New York City Public Schools, “New York City Public Schools Releases Citywide Results for State Assessment,” August 7, 2026.
https://www.schools.nyc.gov/about-us/news/announcements/contentdetails/2026/08/07/new-york-city-public-schools-releases-citywide-results-for-state-assessment
[9] Bureau JS, Howard JL, Chong JXY, Guay F. “Pathways to Student Motivation: A Meta-Analysis of Antecedents of Autonomous and Controlled Motivations.” Review of Educational Research, 2022.
https://doi.org/10.3102/00346543211042426
[10] US FDA: Surrogate Endpoint Resources for Drug and Biologic Development.
https://www.fda.gov/drugs/development-resources/surrogate-endpoint-resources-drug-and-biologic-development
[11] Marilyn Strathern, “‘Improving ratings’: audit in the British University system.” European Review, 1997, 5(3): 305–321. Page 308 discusses how a measure may lose its usefulness for measurement when it becomes a target.
https://gwern.net/doc/statistics/decision/1997-strathern.pdf
[12] ICH E9, Statistical Principles for Clinical Trials, FDA guidance, 1998. See section 5.1 on prespecification of the primary analysis.
https://www.fda.gov/media/71336/download
Sources checked through November 2, 2026. This article distinguishes the parent's account, published research and official data, and the author's analysis. The allegations do not constitute proven grade falsification. The discussion of procedural burdens and classroom interactions describes possible risks; it is not a finding that these events occurred at the school concerned.
Beyond the conclusion, there are questions worth continuing to ask.
Medicine · Science · Ideas and Reading
Follow Before the Conclusion as we stay curious and think independently together.