7 topics. Each one opens with its key idea, then the detail you need to use it.
1. Breaking the claim into jobs
Key idea: List the jobs the claim requires before you look at a single source.
Start every synthesis question with the claim, not the sources. Read the writer's goal and underline the words that limit it. Then turn the claim into a short list of jobs the evidence must do.
Say a student argues that city bike lanes "reduce traffic deaths but only when they are physically separated from cars." That is two jobs. Job one is showing that bike lanes can reduce deaths. Job two is showing that separation is the condition that matters. A source proving that bike lanes cut deaths by 30% does job one. A source comparing painted lanes with curb-protected lanes does job two. A source showing that cycling has doubled since 2015 does neither job, even though it is about bikes.
Watch for small qualifying words. "Primarily," "more than," "only," "both," and "regardless" each add a job. "Necessary and feasible" needs one source for necessity and another for feasibility. "Tailored to regional conditions" needs evidence that the same policy works differently in different places. A single national average cannot show that, however large it is.
This step takes about fifteen seconds, and it does most of the work. Once you know the jobs, most wrong answers fall away. They tend to pick sources that are on topic but fill no job, or that fill the same job twice.
2. Evidence selection: matching sources to jobs
Key idea: Label what each source shows, then pick the set whose labels cover every part of the claim.
Evidence selection questions ask which source or pair "best supports" a specific argument. Treat it as a matching task. For each source, jot one or two words on your scratch paper or with the annotation tool naming what it shows, such as "real effect," "small effect," "public opinion," "cost" or "demographics." Then compare those labels to your job list.
Imagine a student arguing that a four-day school week saves money but harms learning. Source A says a district saved 5% of its budget. Source B says math scores fell in four-day districts. Source C says 80% of teachers like the schedule. Source D says the schedule spread to 900 districts. The right pair is A and B, because one covers "saves money" and the other covers "harms learning." The pair C and D is popular and on topic, but it fills neither job. A pair of A and D proves only the savings half.
The size of a number can mislead you. In a "real but limited" argument, a small effect is the right evidence, because it shows the effect exists and is modest. A dramatic finding would actually work against "limited." Choose the source whose meaning fits the claim, not the one that sounds most impressive.
Also check the explanation inside each answer choice. Some choices name the correct sources but describe them wrongly, for example saying a study "proves" something it only suggests. The reasoning has to be right too.
3. Weighing what kind of evidence a source is
Key idea: Match the type of evidence to the type of claim: beliefs prove beliefs, measurements prove effects.
Sources prove different things depending on how the information was gathered. Learn to sort them quickly.
A controlled study or a large data analysis measures what actually happened. An opinion poll measures what people believe. A self-report survey records what people say about their own behavior or experience. That is a real measurement, but a weaker one, because people misremember or shade their answers. Testimony from someone with direct experience explains why something happens, which is useful but personal. An expert's argument is a reasoned opinion, not data.
The exam tests this by putting a perception source next to an outcome claim. Suppose a writer argues that energy drinks do not really improve test scores. A poll where 65% of students say energy drinks help them focus does not show they improve scores. It shows students believe they help, which the writer might use as the belief to correct. Likewise, officers saying cameras made them more professional is a feeling, not a measured change in incidents.
Two more traps appear often. First, one study in one place cannot prove something "always" works everywhere. Watch for answer choices with absolute words like "always," "proves" or "primary cause" attached to modest evidence. Second, a bigger sample makes a study more precise, but it does not make a different study wrong. When two studies disagree, sample size alone does not settle why.
Ask of every source: what exactly did this measure, and does my claim need that measurement?
4. Handling counterevidence: concede, then refute
Key idea: Name the opposing source fairly, concede what is true, and answer it with stronger evidence.
Many questions give you a source that argues against the writer's thesis and ask how to use it. The correct answer is nearly always to present it as a counterargument, grant what is reasonable in it, and answer it with the other sources.
Here is the logic. A reader who already knows the opposing view will distrust an essay that pretends it does not exist. Naming the objection shows you have thought it through, which builds your credibility. Answering it with evidence then turns the objection into extra support for your side.
Suppose a student argues that school gardens improve nutrition, and one source is a principal saying gardens waste class time. The student should introduce the principal's concern, concede that time is limited, then cite a study showing garden students ate more vegetables and lost no test-score ground. The concern is acknowledged and then answered.
Three wrong moves appear again and again. Ignoring the source makes the essay look one-sided. Quoting a fragment of it so it seems to agree with you misrepresents it, which is dishonest and easy for a reader to catch. Letting the opposing view become your governing thesis or frame hands your opponent the argument. Opening with an objection is fine, as long as you answer it right away and your own thesis stays in control.
A related question asks which source "most undermines" a thesis. Find the source that directly contradicts the claim's central point, especially one that already accounts for the factors the writer would raise. Sources that complicate side details are weaker answers.
5. Explaining why sources disagree
Key idea: Conflicting results are usually explained by a condition that differed, not by declaring one study wrong.
Real research often conflicts. One study finds a large effect and another finds none. The exam asks which third source best explains the difference. The answer is usually a source that names a condition, sometimes called a moderating variable, that differed between the two settings.
Picture this. One school district found that a tutoring app raised reading scores sharply. Another found no change. A third source reports that the app helped only when teachers set aside class time for it, and did nothing when students used it at home. That third source explains the conflict. The two districts probably used the app differently, so both results can be true.
Notice what this explanation does not do. It does not declare one study wrong. It does not decide that nothing can be concluded. It shows that context changes the outcome, which is often the most useful finding of all.
Wrong answers here follow patterns. One picks the study with the bigger sample and calls the contradiction solved, but that answers "which is more reliable," not "why did they differ." Another picks the dramatic study and claims it proves the effect always happens. A third picks a source about feelings or opinions, which cannot explain different measured outcomes.
When you see "explain why Sources X and Y differ," look for a source that mentions a condition, a policy or a setting. That is where the explanation lives.
6. Evidence integration: putting sources in order
Key idea: Choose the order that fits the writer's goal and lets each source set up the next, so every link in the reasoning stays visible.
Organization questions ask which arrangement of sources best supports a goal. The answer depends on what kind of argument the writer is making.
For a cause-and-effect or chain argument, follow the chain. If a writer claims that a factory closing hurt a whole town, the clearest order is usually the closing first, then lost jobs, then empty shops downtown. Each source triggers the next. Starting with the empty shops and tracing back to the closing can also work, but only if every link is still shown. Cramming all the sources into one paragraph hides the chain completely.
For a problem-and-solution argument, a reliable order runs from scope to failure to cause to fix. First show how big the problem is. Then show that the obvious solution falls short. Next explain why it falls short, often with testimony from someone close to the problem. Finally show a solution that worked. Each step answers the question the previous step raises.
For an opening, the strongest choice is usually evidence that is surprising and concrete. A finding that links two things the audience would never connect makes people want to keep listening. A general statistic that everyone already expects does less.
Some choices suggest dropping sources that seem off-topic. Check before you agree. If the writer's goal is to show effects beyond one group, the sources about those wider effects are the point and cannot be cut.
7. Evidence integration in your own synthesis essay
Key idea: Set up, cite, then explain: the explanation is where the points are.
On the free-response synthesis essay, integration is what separates a summary from an argument. Readers reward commentary, meaning your explanation of how each piece of evidence supports your reasoning.
Use a three-step pattern for each source. First, set it up by saying what kind of source it is and why it matters, such as "A study of 40 cities found..." Second, give the evidence briefly, quoted or paraphrased, with a parenthetical citation like (Source C). Third, explain it. Say what the evidence proves and connect it to your thesis. That third step should usually be the longest.
The strongest essays make sources talk to each other. Instead of one paragraph per source, build each paragraph around one of your own reasons and bring in two sources that support it. You might use a data source for the size of a problem and a testimony source for its human cost. You can also set two sources against each other and explain which matters more and why.
Avoid "source dumping," which means stacking quotations with no explanation between them. A paragraph that lists what Sources A, B and D say has used three sources but has made no argument. Also, never let a source speak for you. Your thesis states your position, and the sources serve it.
The skills from the multiple-choice questions transfer directly. Pick evidence that fits your exact claim, concede and refute an opposing source, and order your reasons so each one builds on the last.