Ten lessons on how AP Research is scored, finding a gap and a question, the literature review, method, sampling and ethics, data collection, results, discussion and limits, writing on a schedule, and the presentation and oral defense, with sliders and worked examples.
A study guide to AP Research, not a replacement for your teacher, your school's research review or the College Board materials. The format and weights are from AP Central as read on October 8, 2026. Practice values in the labs are invented.
['AP Seminar or strong research writing']
Course outline
The shape of AP Research and how it is scored
Know what you produce and how the two parts are weighted.
Finding a gap and asking an inquiry question
Choose a question that a study of your own can actually answer.
The literature review
Show what is known and what is missing, and place your study in that picture.
Choosing a method
Match qualitative, quantitative or mixed methods to the question.
Sampling and research ethics
Choose participants responsibly and protect them.
Collecting data and keeping it valid
Gather data carefully and notice threats to validity.
Analyzing and reporting results
Turn data into clear findings and describe them honestly.
Discussion, limits and conclusion
Say what the study means and where it stops.
Writing the paper on a schedule
Plan word counts and revision so the paper is finished on time.
Read every lesson below. The interactive reader above contains the same explanations, with visual tools and quizzes.
1. The shape of AP Research and how it is scored
Learning goal: Know what you produce and how the two parts are weighted.
AP Research has one through-course performance task and no exam. You write an academic paper of 4,000 to 5,000 words, which is 75 percent of the score and is scored by the College Board. You then give a presentation and an oral defense of 15 to 20 minutes total, including 3 to 4 questions from a panel of three evaluators. That part is teacher scored and is 25 percent.
The paper is almost the whole grade, so most of your time belongs in the research and the writing. The presentation and defense still matter: you must explain your choices, defend your method and admit limits in real time.
Submission goes through the AP Digital Portfolio, with final deadlines set by the College Board each year (April 30, 2027 for the final performance tasks, as listed on AP Central on October 8, 2026). The College Board also publishes guidance on the use of AI tools; read it before you start.
Worked example
Paper 70 percent, presentation and defense 90 percent. Weighted overall?
0.75 x 70 = 52.5
0.25 x 90 = 22.5
Total = 75 percent
Practice problem and solution
Paper 80, defense 60 (percent). Weighted overall percent? Enter a number.
0.75 x 80 + 0.25 x 60 = 60 + 15 = 75.
Mental model: One task: a 75 percent paper and a 25 percent presentation and defense.
Common trap: Spending the whole year on the paper and arriving at the defense unprepared.
2. Finding a gap and asking an inquiry question
Learning goal: Choose a question that a study of your own can actually answer.
Research starts with a gap: something published work has not answered, has answered for a different group, or has left disputed. You find gaps by reading reviews, noting limitations that authors admit and asking what changes when you move the setting, the group or the time.
An inquiry question must be answerable with data you can collect or analyze in a school year. How does sleep affect students? is too broad. How does self-reported sleep duration relate to first-period quiz scores among 11th graders at one school? names a group, two measures and a relationship.
Check feasibility early: time, access to participants or data, cost, and ethics review. A brilliant question you cannot study is not a good project. Write a one-paragraph proposal that states the gap, the question and the planned method, and revise it after talking to your teacher.
Worked example
Narrow: Does social media hurt grades?
Group: 11th graders at one school
Measure: self-reported hours of short video per day
Outcome: last semester's GPA
Question: How is daily short-video time related to GPA among 11th graders at one school?
Practice problem and solution
With 3 months of data collection, about how many weeks is that? Use 4.3 weeks per month. Enter a number to one decimal place.
3 x 4.3 = 12.9.
Mental model: Find a gap and ask a specific question you can study in time.
Common trap: Picking a question you cannot collect data for.
3. The literature review
Learning goal: Show what is known and what is missing, and place your study in that picture.
A literature review does not list sources one by one. It groups them by idea, shows where they agree and disagree and ends at the gap your study fills. Each paragraph makes a point about the research, supported by several sources.
Use mostly scholarly sources: peer-reviewed articles, books from academic presses and reports from research bodies. Read abstracts and conclusions first, then the method and results of the sources you rely on. Note the sample, method and limits of each study.
Keep a table with columns for source, question, method, sample, main finding and limit. Sorting by theme in that table gives you the review's outline. Cite every source in a consistent style, and paraphrase rather than quote long passages.
Worked example
Turn this list into a synthesis: Study A: 200 students, sleep and grades linked. Study B: 50 students, no link.
Group: both study sleep and grades
Compare: larger sample found a link, smaller did not
Gap: neither studied one school's morning classes
Claim: evidence for a link is mixed and depends on sample
Practice problem and solution
A 1,200-word review with 24 sources. Average words per source? Enter a number.
1200 / 24 = 50.
Mental model: Synthesize sources by idea and end at the gap.
Common trap: Summarizing each source in turn with no comparison.
4. Choosing a method
Learning goal: Match qualitative, quantitative or mixed methods to the question.
Quantitative methods use numbers: surveys with scales, experiments, existing datasets. They suit questions about how much, how many or whether two things are related. Qualitative methods use words and observations: interviews, open-ended responses, document analysis. They suit questions about how people understand or experience something.
Mixed methods combine the two, for example a survey followed by interviews that explain surprising results. Choose the method from the question, not from what feels easiest. If your question asks whether two measures are related, you need measurements. If it asks how students describe stress, you need their words.
Describe the method in enough detail that another student could repeat it: who took part, how they were chosen, what you measured or asked, in what order and how you will analyze it. State the limits of the design plainly.
Worked example
Question: Why do students choose to skip breakfast? Which method?
The question asks why, about reasons
Reasons are in words
Choose interviews or open-ended responses
Qualitative
Practice problem and solution
12 interviews at 45 minutes of coding each. Hours to code? Enter a number.
12 x 45 = 540 minutes, which is 9 hours.
Mental model: Choose the method from the question, and describe it so it can be repeated.
Common trap: Picking a method for convenience when it cannot answer the question.
5. Sampling and research ethics
Learning goal: Choose participants responsibly and protect them.
A sample is the group you collect data from. A random sample gives every member of the population a chance and supports general claims. A convenience sample, such as friends in your class, is easy but may differ from the wider group, so you must be careful about what you claim from it.
Bigger samples make estimates less noisy. For a proportion, the margin of error for 95 percent confidence is roughly 1.96 times the square root of p(1 - p)/n, widest near p = 0.5. Larger n reduces random error, not bias: a biased sample stays biased.
Research with people carries duties. Get informed consent, keep data confidential, explain risks and let people stop at any time. For students, school research review (often called an IRB) applies when a study involves human subjects, and you must have approval before you collect data. Ask your teacher about your school's process early.
Worked example
n = 400, p = 0.5. 95 percent margin of error?
SE = sqrt(0.25/400) = 0.025
Margin = 1.96 x 0.025 = 0.049
About 4.9 points
Practice problem and solution
n = 100 at p = 0.5. Margin in points to one decimal place? Enter a number.
1.96 x sqrt(0.25/100) = 9.8 points.
Mental model: Sample size reduces noise, not bias; get consent and approval.
Common trap: Collecting data before research review when people are involved.
6. Collecting data and keeping it valid
Learning goal: Gather data carefully and notice threats to validity.
Validity asks whether you measured what you meant to measure. Reliability asks whether you would get similar results again. A survey with confusing questions is unreliable, and one that asks about hours of sleep when you wanted sleep quality is not valid for your question.
Pilot your instrument with a few people and revise confusing items. Record the date, conditions and any changes in a research log. Keep raw data safe and unedited, and work on copies. Track how many people you invited and how many responded, because a low response rate may mean the respondents differ from those who did not respond.
Watch for threats: leading questions, social pressure to answer a certain way, inconsistent conditions and missing data. You cannot remove every threat, but you can state them and explain their likely effect.
Worked example
150 of 300 invited students respond. Response rate?
150 / 300
0.5
50 percent
Half did not respond; note the limit
Practice problem and solution
90 responses from 150 invited. Response rate in percent? Enter a number.
90 / 150 = 0.6, which is 60 percent.
Mental model: Pilot, record conditions, track response rate and state threats to validity.
Common trap: Ignoring who did not respond.
7. Analyzing and reporting results
Learning goal: Turn data into clear findings and describe them honestly.
Start by describing the data: counts, means, medians, spread or the common themes in text. Then compare groups or look for relationships. Report what you found, not what you hoped for. A result that goes against your guess is still a finding.
Use tables and graphs that answer the research question. Label axes with units, give the sample size and avoid cutting off the vertical axis in a way that exaggerates a difference. State both the size of an effect and its uncertainty. A change from 70 to 77 is a 10 percent increase, which is a change of 7 points.
Keep results and interpretation apart. The results section says what the data show. The discussion says what it might mean. For qualitative work, quote short, representative excerpts and explain how you grouped them into themes.
Worked example
Scores rise from 70 to 77. Give the change in points and in percent.
Points: 77 - 70 = 7
Percent: 7 / 70 = 0.10
10 percent increase
Practice problem and solution
Scores rise from 50 to 60. Percent change? Enter a number.
(60 - 50) / 50 = 0.2, so 20 percent.
Mental model: Describe, compare, report honestly, and keep results apart from interpretation.
Common trap: Mixing interpretation into results, or hiding a result you did not expect.
8. Discussion, limits and conclusion
Learning goal: Say what the study means and where it stops.
The discussion links your results back to the question and to the literature. Did your findings agree with earlier work? If not, offer reasons that you can support. Avoid claiming more than the data allow: a relationship in one school does not prove a cause everywhere.
The limitations section is a strength when it is specific. Name the sample, the method and the timing, and say how each could affect the findings. Weak limits such as I had little time say nothing. Strong ones read: because only volunteers took part, the results may overstate interest.
Close with implications and next steps. Who could use the findings, what would a stronger study do and what question remains? Keep the conclusion short and tied to the evidence. Do not introduce new results there.
Worked example
Write a specific limitation for a survey of 40 volunteers.
Name the sample: 40 volunteers
Name the effect: volunteers may care more about the topic
Say what it means: results may overstate interest
Practice problem and solution
A 4,500-word paper with a 900-word discussion. Share of the paper in percent? Enter a number.
900 / 4500 = 0.2.
Mental model: Interpret within the data, name specific limits and close with implications.
Common trap: Claiming cause or broad generality from a small convenience sample.
9. Writing the paper on a schedule
Learning goal: Plan word counts and revision so the paper is finished on time.
A paper of 4,000 to 5,000 words needs a schedule. Break it into sections with word budgets: introduction, literature review, method, results, discussion and conclusion. Set weekly targets and leave at least two weeks at the end for revision and for checking citations.
Draft the method early, while it is fresh, and the results as soon as the analysis is done. Write the introduction and abstract last, so they match what you found. Ask for feedback on a full draft, not only pieces, and revise for argument and structure before polishing sentences.
Cite as you write. Use one citation style throughout and keep a reference list that matches in-text citations. Follow the College Board's rules on academic honesty and on the use of AI tools, and be ready to explain every part of your own work.
Worked example
4,500 words in 9 weeks. Words per week?
4500 / 9 = 500
Draft in 7 weeks, revise in 2
Then about 640 words a week while drafting
Practice problem and solution
4,500 words in 15 weeks. Words per week? Enter a number.
4500 / 15 = 300.
Mental model: Budget sections, set weekly targets, revise on a full draft and cite as you go.
Common trap: Leaving the writing and the revision to the last two weeks.
10. The presentation and oral defense
Learning goal: Present your study and answer panel questions.
The presentation and defense together last 15 to 20 minutes. A panel of three evaluators asks 3 to 4 questions. The presentation should give the question, the gap, the method, the main findings and the limits, in a story the panel can follow.
Plan the time. If you speak for 10 minutes, about 5 to 10 minutes are left for questions. Practice answering without notes and out loud. Expect questions about why you chose this method, how you handled bias, what surprised you and what you would change.
Answer directly, then support with evidence from your study. If you do not know, say so and explain how you would find out. Do not defend a weakness by pretending it is not there; naming it shows understanding.
Worked example
A 10-minute talk in a 20-minute session with 4 questions. Minutes each?
20 - 10 = 10 minutes
10 / 4 = 2.5 minutes per question
Practice 2-minute answers
Practice problem and solution
A 12-minute talk in a 20-minute session with 4 questions. Minutes per question? Enter a number.
(20 - 12) / 4 = 2.
Mental model: Present a clear story, plan Q&A time and answer honestly.
Common trap: Filling the whole time with the talk and leaving no room for questions.