Skip to content

AP U.S. Government and Politics · technique

How to attack stimulus-based multiple choice

On this page · 5 sections

Half your AP U.S. Government score comes from Section I: 55 multiple-choice questions in 80 minutes, four options each, one raw point apiece, no penalty for a wrong guess. A large block of those questions hangs off a stimulus. College Board does not publish a fixed count, but every released Section I mixes three kinds of source. Quantitative sources are line graphs, bar graphs, tables, maps, and infographics. Text sources are short excerpts from the required foundational documents, from Supreme Court opinions, or from news-style writing. Visual sources are mostly political cartoons. They appear because the course requires the exam to test data analysis, source analysis, and applying a required case to facts you have never seen.

Scoring is blunt. There is no partial credit and no human reading your reasoning. Reading the chart correctly and then picking the wrong option earns exactly what leaving it blank earns, which is why nothing should ever be left blank. Most stimulus items are one source feeding one question; occasionally one source feeds two.

Here is the part that decides your score. A stimulus item is almost always two questions welded together. First: what does this source actually say? Second: which course concept does it belong to? Students who nail the first and skip the second lose points to options that describe the data accurately but attach it to the wrong idea. Treat every source as evidence you have to name, not just read.

The steps

  1. 1

    Read the stem before you read the source

    Your eye wants to start at the chart. Don't. The stem tells you which of four jobs you have: pull one value out of the source, describe a trend or difference, explain a likely cause or implication, or connect the source to a concept, institution, or process. Each job needs a different pass over the same source. "According to the table" means the answer is printed in front of you, and anything not printed is wrong no matter how true it is. "Best explains" or "most likely accounts for" means the answer is not in the source at all and you have to supply the course term. "The author would most likely support" means you need the author's position, not yours. Read the stem, name the job in three words, then look at the source with that job in mind. This costs about ten seconds and stops you from reading one graph four different ways.

  2. 2

    Strip the frame in twenty seconds

    Work the frame before the content. On a graph or table, read the title, the units, the years, every axis label, the legend, and the source line underneath. Ask what the denominator is. Percent of registered voters, percent of eligible citizens, and percent of the adult population produce three different turnout numbers for the same election, and distractors are built on that gap. On a text excerpt, name the speaker, name what they want, and name the clause or principle at issue. Brutus No. 1 argues a republic this large cannot stay accountable and that the necessary and proper and supremacy clauses will hollow out the states. Federalist No. 10 answers that a large republic dilutes faction. On a Supreme Court excerpt, check whether the label says majority, concurrence, or dissent, because a dissent tells you what the Court refused to do. On a cartoon, the labels are the argument: work out who each figure stands for and what the artist thinks has gone wrong.

  3. 3

    Predict the answer before you read the options

    Finish the sentence in your own words first. If the job is to describe a trend, say something like "both lines rise, the gap narrows, the older group stays on top." If the job is to explain, say the term out loud: divided government, incumbency advantage, retrospective voting, selective incorporation, fiscal federalism. Then go hunting for the option that matches what you already said. This one habit changes the whole item. With a prediction in hand you compare each option against the source. Without one, you compare the options against each other, and the writers of those options know exactly what a half-read chart looks like. When two options both feel right, one of them is usually a true statement about American politics that this particular source does not show.

  4. 4

    Kill options with three tests

    Nearly every wrong option on a stimulus item fails one of three tests, so run them in order. First, true but not shown. The statement is accurate about American government, but the source does not establish it. If the stem says "according to" or "supported by the data," that option is dead on arrival. Second, the causal leap. The table shows two things moving together and the option claims one caused the other. Data alone almost never licenses cause; unless the stem explicitly asks for a likely explanation, refuse causation. Third, direction and magnitude. The option flips which group is higher, turns a slower increase into a decrease, or swaps a percentage-point change for a percent change. A move from 40 percent to 50 percent is a ten-percentage-point rise and a 25 percent increase, and options live in that difference. Also distrust absolutes: proves, eliminates, all, none, always. Sources show patterns, not proof.

  5. 5

    Name the course concept

    The exam pays for vocabulary. Correct options are written in course language, so deciding which unit an item belongs to usually cuts four options to two before you read them closely. Turnout gaps by age, income, and education point to political efficacy and political socialization. A chart of party-line voting over time points to polarization and divided government. A map of contorted districts points to gerrymandering, to Baker v. Carr making apportionment claims justiciable, and to Shaw v. Reno subjecting race-based districting to strict scrutiny. An excerpt about a president acting without new legislation points to executive orders and the split between formal and informal powers. A table of federal dollars flowing to states points to categorical versus block grants and fiscal federalism. Ask "which unit is this?" the second the source appears, not after you have read all four options.

  6. 6

    Treat a two-question set as one investment

    When one source feeds two questions, read it once and read it properly, then answer both before your working memory dumps it. These pairs usually split along the seam described above: the first question asks what the source says, the second asks which concept explains it. Do not re-strip the chart for the second question. There is also a free hint available. The second stem sometimes names a concept, an institution, or a time frame that clarifies what the source was showing, so if the first question has you stuck between two options, read the second stem before you commit.

  7. 7

    Run the clock and use the tools

    Eighty minutes across 55 questions is about 87 seconds each, and that average has to absorb everything. Standalone definition items should take 20 to 30 seconds, and that is where you bank the time a dense table or a cartoon will eat. Put a hard ceiling of roughly 90 seconds on any single stimulus item. When you hit it, eliminate what you can, choose, flag it, and move. Section I runs in Bluebook, which gives you a highlighter, an annotation tool, an answer eliminator that strikes through options, and a flag for review. Use the eliminator before you flag, so a question you return to is already half solved. Check the clock at question 14, question 28, and question 42; you want to be near 20, 40, and 60 minutes. Never leave a blank, because a wrong answer costs nothing a blank does not.

Worked through

The table below is illustrative. It was written for this walkthrough, and the shape reflects the real age-turnout pattern in presidential elections, but the numbers are invented for instruction and are not official estimates. Read them; do not memorize them.

Reported voter turnout among U.S. citizens, by age group, presidential elections (illustrative figures) Age group | 2012 | 2020 18 to 29 | 45% | 51% 30 to 44 | 58% | 63% 45 to 64 | 68% | 72% 65 and older | 72% | 74%

Question 1. Which of the following is best supported by the table? (A) Turnout rose in every age group, and the gap between the youngest and oldest groups narrowed. (B) Expanded early voting caused the increase in turnout among citizens aged 18 to 29. (C) Turnout among citizens 65 and older fell from its 2012 level. (D) Younger citizens hold weaker policy preferences than older citizens.

Step 1, the job. "Best supported by the table" means describe what is printed. No outside knowledge is wanted.

Step 2, the frame. Title says reported turnout. Units are percentages. The denominator is citizens, not registered voters, which matters because turnout measured against registrants runs much higher. Two presidential years, four age bands.

Step 3, prediction. Every number rises. The youngest band rises most, 45 to 51, six points. The oldest band rises least, 72 to 74, two points. The two middle bands rise four and five points. The youngest-to-oldest gap goes from 27 points to 23. Older citizens still vote more. Say it: all rise, gap narrows a little, old still on top.

Step 4, kill. Option B fails the causal-leap test; early voting is nowhere in the table, and a table of outcomes cannot tell you why anything moved. Option C fails on direction, since 74 is higher than 72. Notice the near-miss it was built from: "grew more slowly than every other group" would have been true, two points against six, five, and four. "Fell from its 2012 level" is a different claim, and it is false. Option D fails on a different ground. The table records whether people voted. It says nothing about how strongly anyone holds a policy preference, so the option is about a quantity this source never measures. Option A survives both halves, because "rose in every group" checks against all four rows and "gap narrowed" checks 27 against 23. Answer A, about 50 seconds elapsed.

Question 2, from the same source. Which of the following best explains why citizens 65 and older turn out at higher rates than citizens aged 18 to 29? (A) Retrospective voting (B) Higher political efficacy and stronger political socialization among older citizens (C) The Twenty-sixth Amendment (D) Selective incorporation

Step 5, name the concept. "Best explains" flips the job: the source is no longer sufficient and you supply the term. Older citizens have had decades to build stable partisan attachments, they move less often so their registrations lapse less, they are more likely to believe their vote matters, and parties and interest groups target them precisely because they turn out. That is efficacy plus socialization, so B. Option A, retrospective voting, is about judging incumbents on results; it explains vote choice, not who shows up. Option C, the Twenty-sixth Amendment, lowered the voting age to 18 in 1971; it explains why the youngest band can vote at all, not why it votes less than the oldest band. Option D, selective incorporation, is the Fourteenth Amendment doctrine applying Bill of Rights protections against the states. It is a Unit 3 term dropped into a Unit 5 question, and that mismatch is one of the most common decoys on the exam. Both questions done in under two minutes, which puts time back in the bank.

Common mistakes

  • Choosing an option that is a true statement about American government but is not established by the source. If the stem says "according to" or "supported by," outside truth is irrelevant and the option is wrong.

  • Reading the source correctly and stopping there. Most stimulus items end by asking which concept, institution, or process the source illustrates, and an accurate description attached to the wrong course term earns nothing.

  • Treating co-movement in a table or graph as cause. Two lines rising together license a description, not an explanation, unless the stem asks for a likely cause.

  • Misreading the frame: ignoring the units, the years, the legend, or the denominator, and confusing a percentage-point change with a percent change. A rise from 40 percent to 50 percent is ten percentage points and a 25 percent increase, and both numbers appear as options.

  • Answering a text or cartoon item with your own politics, or with the view the author is arguing against. Federalist and Brutus excerpts both describe the opposing position before attacking it, and a Supreme Court excerpt labeled as a dissent tells you what the Court declined to do.

  • Spending three minutes on one dense stimulus and running out of clock. At roughly 87 seconds per question there is no room for a hero effort, and blanks at the end cost more than a flagged guess.

Checklist

  • Can you name the job in a stimulus stem in three words before you look at the source?
  • Can you state a data source's units, years, and denominator in under twenty seconds?
  • Can you predict the answer in your own words before reading any option?
  • Can you tell "true but not shown" apart from "supported by this source," and kill the first on sight?
  • Can you compare the size of a change across every row of a table, instead of assuming the biggest number moved the most?
  • Can you convert a rise from 40 percent to 50 percent into both a percentage-point change and a percent change without hesitating?
  • Given a text excerpt, can you name the document or case, the speaker's position, and the clause or principle at issue?
  • Can you tell a majority opinion from a dissent in an excerpt and explain why that label changes the answer?
  • Can you hold about 87 seconds per question, and use Bluebook's answer eliminator and flag instead of stalling on one item?

More AP U.S. Government and Politics technique