Andrew Smith

The 2026 AP Psychology Exam scores: 5: 15%; 4: 35%; 3: 24%; 2: 18%; 1: 8% This AP Psychology Exam was taken by over 379,000 students, 2% of the U.S. high school population, making it one of the most widely taken AP Exams. AP Psychology Multiple-Choice Questions (MCQ) • Students performed exceptionally well on questions assessing Unit 1: Biological Bases of Behavior and Unit 2, Cognition, with 45-50% of students earning most or all available points for these sections. • Students also performed well on MCQs assessing Skill 2: Evaluate Research Design, with 45% earning most or all available points — an indicator that many students are internalizing AP Psychology’s emphasis on scientific literacy and critical evaluation of studies. • The most challenging MCQ content area was Unit 4, Social Psychology and Personality, suggesting students may benefit from more focus on these topics. AP Psychology Free-Response Questions (FRQ) apcentral.collegeboard.org/m… I’ve been so impressed by the way the exam Development Committee of professors and teachers has designed these two FRQs to provide students with real-world, relevant opportunities to demonstrate scientific reasoning. Both questions place students in the role of a practicing psychological scientist: reading real peer-reviewed research, evaluating methodology, interpreting data, and constructing evidence-based arguments about human behavior. FRQ#1, the Article Analysis Question on cell phone absence and anxiety, presented students with a study on whether not answering a ringing phone triggers anxiety. To earn all points, students needed to: identify the research method, operationally define anxiety as measured, interpret a statistically significant result, identify an ethical guideline, assess generalizability, and argue whether the evidence supports or refutes a FOMO explanation. Part C, in which students needed to describe the statistical significance, was the key differentiator for AP 5s, the most challenging part of this FRQ for students. Otherwise, to earn an AP 3, students needed to answer most parts of this question correctly, and generally scored very well on Part E, in which they explained the generalizability of the study, and Part A, in which they identified the research method. FRQ#2, the Evidence-Based Question on retrieval and long-term memory retention, asked students to read and analyze three peer-reviewed studies examining how different learning schedules affect retrieval and retention of recently learned information. Students selected evidence from the sources and explained how each piece of evidence connected to a psychological concept, applying a different psychological concept to their second piece of evidence. Students did a terrific job forming a defensible claim (94% of AP Psychology students earned this point) and identifying supporting evidence. The challenge for some was connecting evidence to a psychological concept in Part C, which was a key differentiator between AP 4s and AP 5s. Students earning AP 5s are typically meeting all of the standards represented in this question, not missing any points. All subjects’ AP score distributions for 2026 will be posted here when available: apstudents.org/score-distrib…

The 2026 AP Macroeconomics Exam scores: 5:19%; 4: 22%; 3: 25%; 2: 20%; 1: 14% AP 2026 Macroeconomics exam was taken by 189,000 students, ~1% of the U.S. high school population. AP Macroeconomics Multiple-Choice Questions (MCQ): • Once again, students demonstrated impressive performance on Unit 1, Basic Economic Concepts questions, with over 55% of students earning all or nearly all the available points in this section. Students also performed well on Unit 2, Economic Indicators and the Business Cycle questions, with 46% of students earning all or nearly all such available points. • The Financial Sector (Unit 4) was the toughest climb within the MCQs: ~15% of students earned all or nearly all of these points. • Across the course skills, students found manipulation tasks and numerical analysis most demanding, skills that reflect college-level expectations that students not only recognize economic relationships, but calculate and apply them with precision. AP Macroeconomics Free-Response Questions (FRQ): apcentral.collegeboard.org/m… The committee of professors and teachers designed this year’s FRQs to require students to demonstrate analytic fluency across the full scope of the AP Macroeconomics course: graphical construction and interpretation, quantitative calculation, verbal explanation, and multi-step causal reasoning. Since AP scores are reported on a 5-point scale, the free-response questions deliberately include some especially difficult points, designed to differentiate AP 5s from AP 4s, points of varying difficulty levels to differentiate AP 4s, 3s, and 2s, and more foundational points to separate AP 2s from AP 1s. • FRQ #1, a question about the AD-AS Model and Monetary Policy, was the exam’s 10-point long free-response question, inviting students to construct a correctly labeled AD-AS graph showing an inflationary gap, explain the economy’s long-run self-adjustment, and then analyze the effects of a specific open-market operation on the money market, interest rates, international capital flows, bond prices, investment spending, and unemployment. Performance on this task differentiated especially well between AP 5s and AP 4s. Long, multi-part FRQs like this one provide great measurement value, enabling educators to evaluate student abilities across the full range of the AP score scale, with several points focused on foundational skills expected of AP 2s, and other points requiring greater depth and mastery, as expected of AP 3s and AP 4s, and finally several very challenging points (international capital flows, unemployment changes, and the long-run self-adjustment process) designed to differentiate AP 5s from AP 4s. • FRQ #2, a question about the Labor Market and the Phillips Curve Model, required students to work from a data table for an economy to calculate the unemployment rate, draw a correctly labeled short-run and long-run Phillips curve, plot numerical equilibrium values, and reason about how shifts in retirement affect the unemployment rate. This question’s combination of data-based calculations, graphing, and verbal explanations is representative of tasks that students earning an AP 4 or higher must consistently be able to do. • FRQ #3, a question about Policy Effects and the Foreign Exchange Market, required students to work across multiple interconnected tasks: 1) identifying a central bank action in an ample-reserves system, 2) calculating the government spending needed to close a recessionary gap, 3) explaining the price-level effect of the change in government spending, and 4) graphing the resulting exchange-rate effect. This was the most challenging question on the exam, focused on differentiating AP 5s and AP 4s, from AP 3s, as students receiving AP 1s and AP 2s are generally unable to earn these points. All subjects' AP score distributions for 2026 will be posted here when available: apstudents.org/score-distrib…

We surveyed AP students worldwide and asked for their perspectives on their AP teachers. Unsurprisingly, AP teachers received sky-high ratings from most students. One of the survey questions asked students to rate the degree to which they felt supported by their teachers when encountering learning challenges or instructional needs. Of all AP subjects, AP Art and Design teachers received the highest ratings on this question of supporting students when encountering learning challenges and needs. Other subjects that received especially high ratings by students for the supportiveness of their teachers: African American Studies; AP Calculus; AP European History; AP Latin; AP Research; AP Spanish Literature.

The 2026 AP Statistics Exam scores: 5: 17%; 4: 23%; 3: 22%; 2: 17%; 1: 21% The 2026 AP Statistics exam was taken by 281,000 students, ~2% of the U.S. high school population. AP Statistics Multiple-Choice Questions (MCQ): 1. Students performed strongest on questions related to Units 1 and 2 (Exploring One-Variable and Two-Variable Data). The high performance here indicated that many AP Statistics students have solid foundational skills in summarizing and interpreting distributions. To receive an AP 5, students were expected to answer all 10 of these questions right, while to receive an AP 4, students were expected to miss no more than one; students receiving AP 3s were expected to answer 7/10 correctly. 2. The most challenging MCQ content area was Units 4 and 5: Probability and Sampling Distributions, a conceptually demanding area that requires students to reason about randomness, probability rules, and the behavior of sample statistics — the theoretical backbone of all of statistical inference. Accordingly, the difficulty level of these questions is such that they serve well to identify students who qualify for an AP 5. AP Statistics Free-Response Questions (FRQ): apcentral.collegeboard.org/m… Across this 90-minute section, students were asked to calculate five-number summaries and compare distributions, design a randomized experiment and interpret statistical significance, apply normal and binomial probability models, conduct a two-sample t-test from summary statistics, analyze a two-way table with conditional probabilities and a chi-square readiness check, and work through a multi-part linear regression analysis complete with confidence and prediction intervals. In other words, the faculty committee who designed these questions did a wonderful job of replicating the scope of a first college-level course in Statistics — no corners cut. (As I think about what AP teachers do so admirably, this is one of those moments where I feel such awe and gratitude for the ways teachers use the scope of the exam to ensure the students in their class have the benefit of no topics falling by the wayside. There are so many pressures on teachers to let stuff slide – from snow days to spring fever to parents demanding that their children receive high grades – but AP teachers tell me all the time that they’re able to invoke the AP Exam’s demands as a reason for maintaining high standards and motivating students to spend the extra time on task that learning at the advanced level in high school demands. The TIMSS study found that AP students were studying two more hours per week than students taking a similar curriculum without a culminating AP Exam. No wonder the AP students in the study outperformed students from all other countries – that two additional hours of study each week makes a huge, cumulative difference in student learning. But this is also an important reminder that with such a significant amount of additional work, students should not be pressured to take more than 1-2 AP courses per year in high school – the load that the AP Program’s own research shows is sufficient to optimize college completion rates.) Since AP scores are reported on a 5-point scale, the free-response questions deliberately include some very difficult points, designed to differentiate AP 5s from AP 4s, points of varying difficulty to differentiate AP 4s, 3s, and 2s, and more foundational points to separate AP 2s from AP 1s. Q1, Exploring Distributions: asked students to determine a five-number summary from raw data, use it alongside a boxplot to compare center and variability in context, and then explain via a stem-and-leaf plot the shape characteristics a boxplot inherently conceals. This last task, asking students to articulate why a summary representation loses information about gaps or clusters in a distribution, is a sophisticated statistical reasoning skill, which distinguished scores of AP 4 and AP 5 from the other scores. Otherwise, this was the easiest FRQ, generating data essential to differentiating between AP 1s, which were only expected to earn at least 1 of these points, from AP 2s, which were expected to earn at least 2 of the 4 points possible here. Q2, Experimental Design and Statistical Significance: was the most challenging short free-response question, providing opportunities to identify students who should receive a 5 versus those who should receive a 4. Students had to identify treatments, experimental units, and the response variable; describe a randomization procedure; and interpret the meaning of statistical significance in context. Q3, Probability Models: required students to apply three distinct probability models in a single question — normal, binomial, and geometric — in the context of a sports team's game-opening musical performances. Students calculated a normal probability, used the binomial distribution to find P(X ≥ 3) across ten games, and then switched to the geometric distribution to calculate the mean and standard deviation of the number of games until a long performance occurs — and interpret that standard deviation in context. This question was almost as demanding as Q2, with each part building on independent probabilistic reasoning skills, such that it served to distinguish between AP 3s, 4s, and 5s, who are expected to earn some or all of these points, while students unable to engage with any parts of this question are receive 1s and 2s. Q4, Two-Sample t-Test: presented students with a randomized experiment comparing two fertilizer brands across 116 trees. Given summary statistics for both groups, students had to conduct a full two-sample t-test: stating hypotheses, checking conditions, computing the test statistic, finding a p-value, and drawing a conclusion in context at the α = 0.05 level. The ability to execute all steps of a significance test accurately — and communicate conclusions appropriately — is a cornerstone of college-level Statistics and this question does an excellent job, across its four points, of differentiating the knowledge and skills of a wide range of performance levels: students are expected to score perfectly here to receive an AP 5, earn at least three of the points to earn an AP 4, and at least two of the points to earn an AP 3. Q5, Two-Way Tables, Conditional Probability, and Chi-Square Readiness: gave students a two-way table of 4,193 professional basketball, football, and baseball players categorized by age group. Students were asked to calculate and interpret conditional probabilities, read a mosaic plot, reason about mutual exclusivity and independence (showing work numerically), and assess the conditions for a chi-square test of independence. This question contains several especially difficult points that evaluate students’ qualifications for an AP 5, which students receiving the other AP scores are unable to earn. Q6, Linear Regression: Inference, Confidence Intervals, and Prediction Intervals: was the capstone 4-point investigative task, and it asked students to engage deeply with linear regression analysis using data from 30 professional baseball teams. Students described a scatterplot relationship, used a least-squares regression equation to make a prediction, compared a circled outlier team to others of its salary class, evaluated the relative strength of linear association across high- and low-salary subgroups, calculated both a 95% confidence interval for a mean response and a 95% prediction interval for a single team, and — in the most sophisticated part — explained why prediction intervals are wider than confidence intervals by connecting the conceptual logic of individual-vs.-mean variability to the mathematical structure of the standard error formulas. This question provides a variety of opportunities for students aiming for AP 4s and AP 5s to distinguish themselves, as several of these are quite advanced points. All subjects' AP score distributions for 2026 will be posted here when available: apstudents.org/score-distrib…

I'm close to having score results to begin sharing for this year's AP Exams. Here's a quick overview of how this will unfold. The scoring of AP essays and free-response questions by ~30,000 professors and teachers - this year's AP Readers - runs throughout the first three weeks of June, with some make-up exam scoring continuing thereafter. Once scoring is complete, the psychometric work begins. These annual processes ensure that this year's students are neither disadvantaged by tougher questions or stricter grading, nor advantaged by easier questions or easier grading. Statistical analysis identifies the precise difficulty and scoring stringency of each question in comparison to prior years. If a given exam version proves to be easier, a higher number of points is required to earn scores of 5, 4, 3, or 2 on that version; if harder, a lower number is required. This work is led for the AP Program by the current president of the National Council on Measurement in Education. After Readers have finished scoring and psychometricians have set the specific "cut scores" (points required) for each exam version, I'll be able to start sharing key findings as we collectively wait for this work to be completed across all 40+ AP subjects so that scores can be sent to colleges, students, and educators in early July. Here's what I'll be covering as results come in: • The focus areas of the free-response questions. (This year's exam questions were crafted by educators from virtually every state and DC; each question is worked on by a minimum of 15 professors and teachers before appearing on the test. I'll use this opportunity to pay tribute to their superb work.) • Any statistically significant increases or decreases in the difficulty of this year's questions or scoring rigor. • Particular content areas within the multiple-choice section where students scored especially well, or particularly struggled. • Specific components of the free-response questions where students scored especially well, or particularly struggled. All of this is very much an attempt to share in real time the data I'm receiving, to tide us all over until scores are released in a few weeks. A few reminders: • I'll report each subject as it's completed, and that timing will depend on how long it takes to score this year’s questions, so I can’t publish a schedule for sharing these updates. But I will have this info for all subjects no later than June 30. • Because these score distributions include all students worldwide, individual classrooms will often have score distributions that are either higher or lower than this aggregation. • AP Exams aren't scored on a curve. Rather, as many students as earn the points necessary for college credit receive a score of 3 or higher. • If you're an AP student, here's information about how to make sure you're able to view your AP scores starting July 6: apstudents.collegeboard.org/… • You have until June 20 to indicate which college should receive your free score send: apstudents.collegeboard.org/… • If you're an AP educator, here's information about how to view your students' AP scores starting July 6: apcentral.collegeboard.org/e…