In the summer of 2019, the College Board released the new AP Statistics Course and Exam Description (CED). We knew this new framework would likely cause some shifts in what questions would be asked on the AP Exam, and only time would tell. Tossing aside the 2020 Exam, we now have three AP exams as guidance for the future. So what insights can we gain from these exams? How will this help us improve our instruction?

**What Can We Learn from the Multiple Choice?**

While we will likely never see the multiple choice questions from the recent Operational Exams, we do have access to the multiple choice questions from the **2021 and 2022 International Exams** (thanks to some savvy AP Stats teachers who realized these questions were added into the question bank on AP Classroom). We categorized each of these 80 questions into their respective CED Units, as well as noting the Essential Knowledge statements associated with each question.

This table compares the targeted percent of questions by Unit as suggested by the College Board to the actual percent by Unit found in this sample of multiple choice questions.

## Important Essential Knowledge from Each Unit!

In addition to organizing the content into units, the College Board CED also provides Essential Knowledge statements that give even more specific details about the content. Within each unit, we wanted to determine if there were any Essential Knowledge statements that are being tested more frequently than others. When teaching the unit, it is helpful to keep these key points in mind. Here are the most frequent so far:

**Unit 1: Exploring One-Variable Data**

VAR-2.B: Determine proportions and percentiles from a normal distribution.

**Unit 2: Exploring One-Variable Data**

DAT-1.H: Interpret coefficients for the least squares regression line model.

**Unit 3: Collecting Data**

VAR-3.C: Compare experimental designs and methods.

VAR-3.E: Interpret the results of a well-designed experiment.

**Unit 4: Probability, Random Variables, and Probability Distributions **

UNC-3.B: Calculate probabilities for a binomial distribution.

VAR-4.E: Calculate probabilities for independent events and for the union of two events.

VAR-5.C: Calculate parameters for a discrete random variable.

VAR-5.E: Calculate parameters for linear combinations of random variables.

**Unit 5: Sampling Distributions**

UNC-3.R: Determine whether a sampling distribution of a sample mean can be described as approximately normal.

**Unit 6: Inference for Categorical Data: Proportions**

UNC-5.C: Identify factors that affect the probability of errors in significance testing.

**Unit 7: Inference for Quantitative Data: Means**

UNC-4.R: Calculate an appropriate confidence interval for a population mean, including the mean difference between values in matched pairs.

**Unit 8: Inference for Categorical Data: Chi-Square (Not Enough Data!)**

**Unit 9: Inference for Quantitative Data: Slopes (Not Enough Data!)**

## What Can We Learn from the FRQs?

While the multiple choice questions are harder to come by, the released free response questions (FRQs) are plentiful. But once again, we wanted to focus on questions that have been released *after* the release of the College Board CED. For this analysis, we looked at the **International Exams from 2021 and 2022** and the **Operational Exams from 2021, 2022, and 2023**. Because each question crosses over several CED Units, we couldn’t do a simple unit classification like we did for the multiple choice. But we did keep track of which questions were the most frequent. Here is the list of the most frequent:

## Describe and Compare Distributions and Relationships!

Every recent AP Exam has had the students describing or comparing one-variable data (SOCS) or describing a relationship for two-variable data (DUFS). Sometimes both are coming up in the same year (see 2023 Operational Exam #1 and #5)! Take some extra time to review some past rubrics to help students know exactly what to include and how this works with a variety of graphs.

## Identify and Interpret Vocabulary Terms!

Students are frequently asked to identify and/or interpret vocabulary terms, especially from CED Units 2 and 3. Here are the terms that are most frequently assessed:

**Unit 2 - ****Exploring Two-Variable Data**

y-intercept

slope

residual

r-squared

**Unit 3 - ****Collecting Data**

experimental units

treatment

response variable

Consider using the Stats Medic Ultimate Interpretations Guide to help your students prepare for these questions.

## Know the Benefits of Methods and Conclusions

Recent exams have emphasized asking why certain methods would be better than others. As you go through Unit 3, make sure students get comfortable with why samples have to be representative, why stratified and clustering may be better than a simple random sample, and why blocked designs (including matched pairs) can be better than just random assignment alone. As with any type of observational study or experiment, make sure students also know the differences between correlation and causation.

**Probability Distributions Matter!**

Maybe not directly in the probability question (especially if they use a table-style question), but somewhere in the FRQs you can expect students to be doing a normal distribution calculation. Using normal distributions is at the heart of AP Statistics, and it has come up on every recent exam. Another favorite distribution is the binomial distribution. Recent years have been asking students to "define a random variable and state how it is distributed"(hint: if they ask what type of random variable a situation is - it’s often looking for binomial). Remember when this happens, students would not only need to say it's binomial, but also define the value of n and p.

**Know Your Inference Procedures!!**

This is a given! Every year, we have a full-blown confidence interval (2022 Operational) or significance test (2021 International, 2021 Operational, 2022 International, 2023 Operational). Make sure that your students know not only the process but have a conceptual understanding of statistical significance. While sometimes these FRQs are a standalone inference procedure, they often come paired with a follow-up question, such as “Based on your conclusion from part (a), which of the two errors, Type I or Type II, could have been made? Interpret the consequence of the error in context.” (2021) or “Based on your conclusion from part (a), which of the two errors, Type I or Type II, could have been made? Interpret the consequence of the error in context. " (2021) or “Based on the confidence interval in part (a), do the sample data provide convincing statistical evidence that the proportion of all teenagers in the United States who would respond that they use a video streaming service every day is not 0.5? Justify your answer. ” (2022).

To help keep your students organized, check out our Stats Medic Ultimate Inference Guide to help your students decide and carry out the appropriate inference procedures.

## What Can We Learn from the Rubrics?

Of course, it helps to know what questions are coming, but you also need to know what responses are needed for full credit. And while we can all read the rubrics, there are often student responses that don’t fit nice and neatly into the rubric descriptions. This is why we created the new blog series “Would This Get Credit”. Here are the blog posts that cover the free response questions from 2023 Operational Exam.

2023 #1: Alaskan Streams (Describing a distribution, constructing a boxplot, making decisions with boxplots)

2023 #2: Fiber Driveways (Identifying experimental units, treatments, response variable, describing random)

2023 #3: Bath Fizzys (Probability, conditional probability, expected value)

2023 #4 OMEGA-3 Supplement (Matched pairs t test)

2023 #5: Tule Elk (Describe scatterplot, making a prediction, residual, interpreting slope, p-value and conclusion for test of slope)

2023 #6: Cleo Gold Necklaces (Normal probability calculation, sampling distribution of xbar, significance test conclusion, sampling distribution of range, conditional probability)