What this rubric measures
The AP Seminar PT2 Individual Essay Rubric (IWA) is the official scoring guide used to evaluate student writing on AP Seminar assessments. It is an Analytic (7 rows) rubric that scores responses across 7 distinct criteria, allowing teachers to give precise, targeted feedback on each area of writing.
All 7 scoring criteria
Click any criterion to expand its score level descriptors. The language below is taken verbatim from the official College Board AP Seminar scoring guide.
1 Row 1: Understand and Analyze Context (Stimulus Integration)
The response demonstrates the relevance of at least one of the stimulus materials to the argument by integrating it as part of the response (for example, as providing relevant context for the research question, or as evidence to support relevant claims). Responses earning 5 points include a reference to the stimulus material that:
- Reflects an accurate understanding of the source and demonstrates an understanding of its context (e.g., date, region, topic).
- Performs a relevant and authentic function within the argument (i.e., it serves a purpose that enhances, forwards and/or directly supports the argument).
The response does not incorporate any of the stimulus material, or, at most, it is mentioned in only one sentence. OR the response includes a discussion of at least one of the stimulus materials; however, it does not contribute to the argument. Responses earning 0 points include a reference to the stimulus material that:
- Is tangential.
- May misrepresent what the sources are discussing/arguing or may use the source in such a way that ignores its context.
- Is only used for a definition or facts that could be obtained from other, more relevant sources.
- Is no more than a jumping-off point for the student's argument, no more than a perfunctory mention.
- Could be deleted with little to no effect on the response.
Row 1 scores integration of the College Board stimulus material. Binary scoring: 0 or 5 points. References to stimulus materials may be included multiple times in the response; only one successful integration is required to earn the 5 points.
2 Row 2: Understand and Analyze Context (Significance)
The response explains the significance or importance of the research question by situating it within a larger context. Responses earning 5 points:
- Address an area of investigation that is narrow enough to address the complexity of the problem or issue. The context, once established, remains relevant throughout the argument.
- Provide specific and relevant details (i.e., who, what, where, when) for all elements of the research question and/or argument.
- Make a specific and compelling case for the urgency or the importance of the research question and/or argument.
The response either provides no context, OR the response makes simplistic references to or general statements about the context of the research question. Responses earning 0 points:
- Identify too many aspects of the topic to address complexity.
- Provide unsubstantiated assertions without explanations (e.g., "this is important").
- May provide contextual details, but they are tangential to the research question and/or argument.
- Provide overly broad, generalized statements about context.
- Provide context for only part of the question or argument.
Row 2 scores how well the student situates the research question within a larger context. Binary scoring: 0 or 5 points. Context is usually (but not always) found in the first few paragraphs.
3 Row 3: Understand and Analyze Perspective
The response evaluates multiple perspectives (and synthesizes them) by drawing relevant connections between them, considering objections, implications, and limitations. Responses earning 9 points:
- Elaborate on the connections among different perspectives.
- Use the details from different sources' arguments to explain specific relationships or connections among perspectives (i.e., evaluate comparative strengths and weaknesses of different perspectives by placing them in dialogue).
The response describes multiple perspectives and identifies some relevant similarities or differences between them. Responses earning 6 points:
- Make general comparisons between perspectives describing only basic agreement or disagreement.
- Explain that disagreement/agreement exists, but do not develop a nuanced, detailed discussion of how they relate.
- At times present perspectives that are clearly derived from specific sources, but may lapse into opinions or stakeholder perspectives that are not clearly linked to specific sources.
The response provides only a single perspective. OR the response identifies and offers opinions or unsubstantiated statements about different perspectives that may be overly simplified. Responses earning 0 points:
- Provide only one perspective.
- May use a lens or lenses that all work to convey the same point of view.
- Convey perspectives as personal opinions or assertions without evidence.
- Provide perspectives that are isolated from each other without explicit comparison.
- Provide perspectives that are oversimplified by treating many voices, stakeholders, or stances as one.
Row 3 scores discussion of multiple perspectives. A perspective is "a point of view conveyed through an argument" (the source's argument). Facts, topics, lenses, and general stakeholder points of view are not perspectives. Scoring note: clear attribution must link perspectives to sources consistently to score 9.
4 Row 4: Establish Argument
The response is a clear and convincing argument. The response is logically organized and well-reasoned by connecting claims and evidence, leading to a plausible, well-aligned conclusion. Responses earning 12 points:
- Organize the argument in a way that is often signposted or explicit.
- Provide commentary that explains fully how evidence supports claims (i.e., the commentary will engage with the content of the evidence to draw conclusions).
- Provide an argument that is driven by student voice (commentary).
- Integrate alternate views, perhaps by engaging with counterclaims or using them to demonstrate a nuanced understanding.
- Provide a conclusion that is fully aligned with the research question.
- Present enough detail to assess the plausibility of the conclusion (perhaps with an assessment of limitations and implications).
The argument presents a claim with some flaws in reasoning. The response is logically organized, but the reasoning may be faulty or underdeveloped OR the response may be well-reasoned but illogical in its organization. Responses earning 8 points:
- Organize the argument well OR link evidence and claims well in discrete sections, but do not do both.
- Provide evidence that often drives the argument, rather than contributing to the response's argument.
- Present an argument that simply repeats but does not develop.
- Present claims that lack cohesion. There is no single articulated controlling argument.
- Provide a conclusion that lacks either enough detail to assess plausibility or is not fully aligned with the research question.
- At times lack clarity on what is student generated and what is derived from sources.
The response provides only unsubstantiated opinions or claims. OR the response summarizes information (no argument). The response employs inadequate reasoning due to minimal connections between claims and evidence. Responses earning 0 points:
- Base the argument on opinion(s).
- Seek to explain a topic, rather than take a position (e.g., report, summary, chronicle).
- Completely lack a conclusion or offer an overly-general, vastly-simplified conclusion.
- Provide an argument that is very difficult to discern, that contradicts itself, or is invalid.
- Provide an argument or claims that do not allow for alternate views.
Row 4 is the highest-weight row at 12 points. It scores the overall argument's organization, reasoning, and conclusion. Scoring note: to score 12, there must be clear attribution for paraphrased material consistently.
5 Row 5: Select and Use Evidence
The response includes relevant, credible and sufficient evidence to support its argument. Responses earning 9 points:
- Include research sources that are relevant to the topic and appropriate for an academic argument on this topic.
- Establish credibility of the sources of evidence (through effective citation, attribution or explanation) consistently.
- Provide purposeful analysis and evaluation of evidence used.
- Make effective use of well-chosen evidence from scholarly work.
- Provide relevant and credible evidence that fully supports claims.
The response includes mostly relevant and credible evidence. Responses earning 6 points:
- Include research sources that are mostly relevant to the topic, only some of which are appropriate for an academic argument (e.g., may be overly reliant on journalistic sources).
- Establish credibility of the sources of evidence inconsistently.
- Include many sources that are merely referenced when they require justification.
- Draw upon outdated research without providing a rationale for using that older evidence.
- May cite several scholarly works, but select excerpts that only convey general or simplistic ideas.
- Provide evidence that at times fully supports claims (e.g., there are sometimes gaps) or provides evidence that only generally supports claims.
Any evidence presented in the response is predominantly irrelevant and/or lacks credibility. Responses earning 0 points:
- Include many sources that are not credible for the context in which they are used.
- Include no well-vetted sources beyond the stimulus materials.
- May include a well-vetted source that is not used effectively.
Row 5 scores evidence selection and use. General reference guides such as encyclopedias and dictionaries do not fulfill the requirement for a well-vetted source.
6 Row 6: Apply Conventions (Citation)
The response attributes, accurately cites and integrates the sources used through the use of in-text citations or footnotes. The bibliography or works cited accurately references sources using a consistent style. Responses earning 5 points:
- Contain few flaws.
- Provide a clear organizational principle in bibliography/works cited.
- Provide consistent evidence of linking internal citations to bibliographic references.
- Include consistent and clear attributive phrasing for paraphrased material and/or in-text parenthetical citations.
The response attributes or cites sources used through the use of in-text citations or footnotes, but not always accurately. The bibliography or works cited references sources using a generally consistent style with some errors. Responses earning 3 points:
- Provide some uniformity in citation style.
- Provide an organizational principle in bibliography/works cited which may be uneven in some places.
- Include unclear references or errors in citations (e.g., citations with missing elements).
- Provide some successful linking of citations to bibliographic references.
- Provide some successful attributive phrasing for paraphrased material and/or in-text parenthetical citations.
The response is missing a bibliography/works cited OR the response is largely missing in-text citations/footnotes. Responses earning 0 points:
- Include internal citations, but no bibliography (or vice versa).
- Demonstrate no organizational principle in bibliography/works cited.
- Provide little or no evidence of successful linking of in-text citations to bibliographic references.
- Include poor or no attributive phrasing with paraphrased material (e.g., "Studies show..." with no in-text citation).
Row 6 scores citation conventions. No particular style sheet is required in AP Seminar, but the response must use a style that is consistent and complete. Cannot score 5 points if essential elements (author/organization, title, publication, date) are consistently missing.
7 Row 7: Apply Conventions (Style)
The response creates variety, emphasis, and interest to the reader through the use of effective sentences and precision of word choice. The written style is consistently appropriate for an academic audience, although the response may have a few errors in grammar and style. Responses earning 3 points:
- Contain few flaws which do not impede clarity for understanding complex ideas.
- Demonstrate word choice sufficient to communicate complex ideas.
- Use clear prose that maintains an academic or scholarly tone.
The response is mostly clear but may contain some flaws in grammar or a few instances of a style inappropriate for an academic audience. Responses earning 2 points:
- Contain some lapses in sentence control (e.g., run-ons, fragments, or awkward syntax when integrating quoted material).
- Lapse into colloquial language.
- Demonstrate imprecise word choice insufficient for communicating complex ideas.
- Use overly dense prose at the expense of coherence and clarity.
The response has many grammatical flaws, is difficult to understand, or is written in a style inappropriate for an academic audience. Responses earning 0 points:
- Contain multiple grammatical errors that make reading difficult.
- Use an overall style that is colloquial or in other ways not appropriate for an academic paper.
- Provide too few sentences to evaluate or the student's own words are indistinguishable from paraphrases of sources.
Row 7 scores grammar and academic style. Readers should focus on the sentences written by the student, not those quoted or derived from sources.
How to score with the AP Seminar PT2 Individual Essay Rubric (IWA).
A practical guide for teachers and norming teams. How to apply each descriptor consistently, the pitfalls that hurt inter-rater reliability, and a workflow for calibrating with colleagues.
Seven rows, scored independently
- The IWA is scored across 7 rows for 48 points total. Row 4 (Argument) is the highest-weight row at 12 points. Rows 3 and 5 are weighted at 9 points each. Rows 1, 2, and 6 are 5 points each. Row 7 (Style) is 3 points.
- Each row is scored independently against the preponderance-of-evidence (best-fit) standard.
- Rows 1 and 2 are binary (0 or 5). All other rows have multiple intermediate score levels.
The stimulus packet integration requirement
- To score above 0 on Row 1, the response must integrate at least one stimulus source as part of the argument, not just mention it. The College Board explicitly excludes "jumping-off point" or "perfunctory mention" uses.
- If a response cites sources from a previous year's stimulus packet and does not cite stimulus material from the current year, it is considered off-topic and scored 0 across all rows.
- A response that is not in any way related to a theme connecting at least two of the stimulus materials is considered off-topic.
Common pitfalls to avoid
- Awarding Row 3 (Perspectives) 9 points for a response that describes multiple perspectives but does not place them in dialogue or evaluate them.
- Awarding Row 4 (Argument) 12 points without checking that the conclusion is fully aligned with the research question.
- Awarding Row 5 (Evidence) 9 points when the response is overly reliant on journalistic sources rather than scholarly work (this caps at 6).
- Awarding Row 6 (Citation) 5 points when essential elements (author, title, publication, date) are consistently missing.
Tips for AP norming
- Anchor your norming session with the College Board's released sample IWAs, scored and annotated by AP Readers.
- Row 4 (Argument) is the highest-variance row. Spend extra norming time distinguishing 8-point (organized but flawed reasoning) from 12-point (logically organized, well-reasoned with aligned conclusion) responses.
- Row 3 (Perspectives) is the second-most-debated row. Practice distinguishing description of perspectives (6 pts) from evaluation and synthesis of them (9 pts).
Notes for the AP Seminar PT2 IWA Rubric
Performance Task 2 is the individual research-based essay component of AP Seminar. The College Board releases a stimulus packet each January with several sources on a broad theme. Students develop their OWN research question rooted in at least two of the stimulus sources and write a 1,200-word Individual Written Argument (IWA).
The IWA is externally scored by trained AP readers, not by classroom teachers. Total possible is 48 points across 7 rows.
Row 4 (Argument) is the highest-leverage row at 12 points and the most commonly debated in calibration. The 12-point level requires both logical organization AND a conclusion fully aligned with the research question, students often have one without the other.
Row 1 (Stimulus Integration) is binary at 0 or 5 points. A response that fails to integrate a stimulus source as functional evidence (or only uses it as a definition or jumping-off point) earns 0 on Row 1 regardless of the rest of the essay.
See this rubric in action.
EnlightenAI scores student writing on this exact rubric, with per-criterion feedback that mirrors how you grade by hand. The sample response below shows how the rubric applies to a real piece of student writing, scored against every criterion.
IWA on centralized export-control review for federally funded AI research
Introduction (verbatim excerpt)
Open scientific exchange has historically driven American advantage in scientific research, from the Manhattan-era physics community to the post-war biomedical boom. In recent years, however, an increasing share of frontier research has dual-use character, that is, research that produces results with both civilian and national-security implications. Computational artificial intelligence is the most prominent contemporary example. This essay argues that US research universities should adopt centralized export-control review of federally funded computational AI research as a structural condition of receiving federal funds, while preserving open publication norms for non-classified results, because the dual-use risks of frontier AI research now meet the threshold the export-control regime was designed to address, and decentralized institutional review has demonstrably failed.
Stimulus integration paragraph (verbatim excerpt)
Source A (the 2023 National Academies report on dual-use research oversight) makes explicit that the existing US export-control framework is structurally mismatched to the speed and decentralization of contemporary AI research. The report's specific recommendation, that funding agencies adopt centralized review at the agency level rather than relying on per-institution compliance, anchors this essay's argument. Reading Source A in dialogue with Source C (the 2024 op-ed by former NSF director Subra Suresh) sharpens the case: Suresh argues from inside the system that institutional review boards, while well-intentioned, lack the technical depth to evaluate frontier AI research, and that consequently, dual-use risk assessment falls through the cracks. Together, the two sources establish that centralization is necessary, not optional.
Counterargument and limitations (verbatim excerpt)
An obvious counterargument, advanced by Source D (the 2024 ACLU report on academic freedom) and echoed by a substantial fraction of academic ethicists, is that centralized export-control review would chill open scientific exchange and disproportionately affect international students and faculty. The point is real and matters. Two considerations bound the concern. First, the proposal here is narrower than the full export-control regime Source D critiques, it covers federally funded computational AI research specifically, not all AI research and not basic mathematics. Second, the alternative is not the status quo, the alternative is an escalating patchwork of unilateral institutional restrictions that already affect international researchers in inconsistent ways. A centralized federal review with uniform criteria is, on balance, more protective of academic freedom than the current trajectory. This argument acknowledges Source D's concern without dismissing it, and explains why the proposed policy is a proportionate response to a real risk.
Conclusion (verbatim excerpt)
The case for centralized federal export-control review of computational AI research rests on three observations together. First, the dual-use character of frontier AI research now meets the threshold the export-control regime was designed to address. Second, decentralized institutional review has measurably failed to evaluate that risk consistently. Third, narrower centralized review, paired with preserved open publication for non-classified results, is more protective of academic freedom than the patchwork status quo. Implementation will require careful design to avoid chilling international collaboration, and the policy should sunset and be reviewed at five-year intervals. But the policy is the right balance between open exchange and security in the current research environment.
Stimulus integrated functionally; research question situated with specific context
Row 1 earns 5 (Source A integrated as functional evidence for centralization, not a definition or jumping-off point). Row 2 earns 5 by situating the research question in the dual-use debate with specific historical and contemporary detail. Earns 10 of 10.
Perspectives placed in dialogue; argument logically organized with aligned conclusion
Row 3 earns 9 by placing Source A in dialogue with Source C and engaging Source D as counterargument. Row 4 earns 12, argument signposted, conclusion aligns with research question, alternate views integrated rather than dismissed. Combined earns 21 of 21.
Evidence mostly credible; citations and style consistently academic
Row 5 earns 6, evidence is mostly relevant and credible but a few citations are stronger than others (one journalistic source). Row 6 earns 5 (consistent citation style, accurate attribution). Row 7 earns 3 (clear academic prose). Combined Rows 5, 6, and 7 earn 14 of 17.
Score this rubric consistently, with the feedback students actually use
EnlightenAI is trained on your standards and your exemplars, then scores at the speed of your classroom.
Trained on your rubric
Upload this rubric, or any custom one, and the AI learns your exact criteria, descriptor language, and score level boundaries.
Per-criterion feedback
Students receive specific, actionable comments tied to each criterion, exactly the way you'd grade by hand.
Built for K–12 schools
Roster sync, FERPA-aligned data handling, and per-school configuration so every campus uses the same standards.
About the AP Seminar PT2 Individual Essay Rubric (IWA)
What is the AP Seminar Individual Written Argument (IWA)?
How is the IWA scored?
How is the IWA different from the IRR?
Can the IWA use sources from a previous year's stimulus packet?
What if the IWA is off-topic?
Is this rubric the official version from College Board?
Where can I find the source document?
Can EnlightenAI score student writing using this rubric?
Use this rubric in EnlightenAI
Train EnlightenAI on the AP Seminar IWA rubric and start scoring student PT2 practice essays with consistent per-row feedback in a single class period.