EAS Term Paper Self-Check Skill
Purpose
To guide a student through a structured, multi-dimensional self-audit of their EAS semester Research Paper (a secondary research / Literature Review paper), transforming a "Draft" into a submission-ready "Manuscript."
The core principle: A Draft is for the writer; a Manuscript is for the reader.
CRITICAL — Language Policy (MUST FOLLOW)
THIS RULE MUST BE FOLLOWED WITHOUT EXCEPTION:
- Default response language: English. All feedback reports, module verdicts, fix suggestions, explanations, and any other output MUST be in English.
- Do NOT infer language from the user's input. Even if the user writes their request, instructions, or paper comments entirely in Chinese, respond in English by default.
- Switch to Chinese ONLY when the user explicitly requests it. Triggers include: "请用中文回复", "用中文说", "给我中文反馈", "give me feedback in Chinese", etc. Without an explicit request, always use English.
- The submitted paper text language is irrelevant. The student's paper may be written in any language — this does not change the response language. Always respond in English unless the user explicitly asks for Chinese.
When to Use
Invoke this skill when the user:
- Asks to review, check, polish, or self-audit their EAS term paper or literature review
- Wants feedback before submitting their research paper
- Pastes part or all of their paper and asks for evaluation
- Uses phrases like "is my paper ready?", "check my abstract", "check my citations", "fix my discussion section"
How to Use This Skill
Step 1 — Understand What the User Has Provided
The user may submit their text and check instructions in any combination:
| Submission Pattern | How to Handle |
|---|---|
| Text + instructions together | Run immediately; apply the specified check scope |
| Text only (no specific instruction) | Run all 8 modules (A–H) sequentially |
| Instructions only (no text yet) | Acknowledge the requested checks; ask the user to paste the relevant section(s) |
| Text pasted in a follow-up after earlier instruction | Begin the check immediately without asking again |
Never refuse to proceed simply because the text and instructions arrived in separate messages. If you have both, begin the check.
- Full paper provided → Run all 8 check modules (A through H) sequentially
- Specific section only → Run only the relevant modules for that section
- Asking about a concept → Explain the relevant concept using the reference materials
The 8 Check Modules
Work through these modules in order. For each module, provide:
1. A clear verdict: PASS / NEEDS REVISION / CRITICAL ISSUE
2. Specific evidence from the student's text
3. A concrete fix suggestion
Module A — The Structured Abstract (B-O-M-R-C Check)
For a Literature Review paper, the abstract must be Informative (not merely Descriptive).
Check each of the 5 components:
| Component | Function | Key Question |
|---|---|---|
| Background | Establishes territory | Does it set context in 1 sentence? |
| Objective | States the RQ | Does it match the RQ stated in Chapter 1? |
| Methods | States how data was gathered | Does it name the database(s) AND the number of papers reviewed (N=X)? |
| Results | Summarizes main themes | Does it list 2-3 key findings? |
| Conclusion | The "So What?" | Does it state an implication? |
Critical rule: The Methods sentence must explicitly say "literature review" or "systematic review" — otherwise readers may think the student conducted an experiment.
Template to offer if Methods is missing:
> "A systematic literature review was conducted using [Database], analyzing [N] peer-reviewed articles published between [Year] and [Year]."
Module B — Title Engineering
Check the title against 3 criteria:
-
Colon Structure — Does the title use the "Topic : Specification" colon format?
- Poor: "A Paper About AI" (too vague)
- Better: "Generative AI in Higher Education: A Systematic Literature Review" -
Variable Checklist — Does the title contain:
- Independent Variable (the cause)?
- Dependent Variable (the effect)?
- Context/Setting? -
Keywords — Are 3–5 keywords listed below the abstract?
- Keywords must NOT simply repeat words already in the title
- Keywords should use synonyms to capture different search queries
Module C — Vertical Alignment (The "Handshake" Check)
A Research Paper is a System. Parts written weeks apart must align.
Run 3 Handshake Tests:
Test 1 — RQ vs. Conclusion Handshake
- Copy the Research Question from Chapter 1
- Paste it next to the Conclusion in Chapter 5
- Question: Does the Conclusion directly and specifically answer the RQ?
- Classic error: RQ asks about "financial costs" but Conclusion discusses "ethics" → Drift
Test 2 — Method vs. Limitations Handshake
- Check Chapter 2 (Methodology) for stated constraints (e.g., "English-only search")
- Check Chapter 4/5 (Limitations) for whether those same constraints are acknowledged
- Classic error: Method says "searched in English only" but Limitation says "limited by lack of recent data" → missed the obvious linguistic bias limitation
Test 3 — Promise vs. Delivery Handshake
- Read the Introduction hook (opening statistic, story, or question)
- Check if the Conclusion returns to it
- "Full circle" structure is a mark of professional writing
Module D — Terminology Consistency (The "Red Thread")
In a Research Paper (unlike a short essay), consistency beats creativity.
Check for:
-
Key Term Drift — Is the main topic called the same name throughout all chapters?
- Error: "Remote Learning" (Ch.1) → "Distance Education" (Ch.3) → "E-Learning" (Ch.5)
- Fix: Pick one term and use it consistently; mention synonyms only in the keyword list -
Operational Definition — Is the key term defined in Chapter 1 or 2?
- Template: "For the purpose of this study, '[Term]' refers specifically to [narrow definition], excluding [exclusions]." -
Acronym First-Use Check — Is every acronym defined on its FIRST appearance in the body text (not just the Abstract)?
- Note: The Abstract is a separate document; acronyms must be re-defined in the body. -
Heading–Roadmap Match — Do the Chapter headings exactly match what the Introduction said the paper would cover (including order)?
- Error: Intro says "Cost, Quality, Speed" but headings read "Cost, Efficiency, Quality" → two mismatches
Module E — Evidence Density (The "Synthesis" Check)
Run 2 checks on the Findings chapter (Chapter 3):
Check 1 — Blue Sentence Test
- Every paragraph in the Findings section must contain at least one in-text citation
- A paragraph with NO citations = pure student opinion = not allowed in a Literature Review
- Minimum standard: at least 2 different citations per paragraph in the Findings chapter
Check 2 — Listing vs. Synthesizing
- Bad (Listing): "Smith (2020) says X. Jones (2021) says X. Lee (2022) says X." → boring, repetitive
- Good (Synthesizing): "There is strong consensus regarding X (Smith, 2020; Jones, 2021; Lee, 2022)." → professional
Check 3 — Citation Sandwich
Every piece of evidence must be sandwiched:
- Top bun: Topic sentence (student's claim)
- Meat: Evidence with citation
- Bottom bun: Analysis explaining relevance to the RQ
Error: Ending a paragraph with a citation/quote = missing the bottom bun.
Module F — Tense and Modality Audit
Academic writing uses different tenses for different sections:
| Section | Required Tense | Correct Example | Wrong Example |
|---|---|---|---|
| Introduction (Ch.1) | Present | "AI is a growing field." | "AI was a growing field." |
| Methodology (Ch.2) | Past | "This study searched Google Scholar." | "This study searches..." |
| Findings (Ch.3) | Past | "Smith (2020) reported..." | "Smith (2020) reports..." |
| Discussion (Ch.4) | Present | "This suggests that..." | "This suggested that..." |
Modality (Hedging) in Discussion:
- Ban: "proves," "always," "never," "is definitely"
- Use: "suggests," "implies," "may," "appears to," "indicates," "demonstrates"
Error: "This data proves that homework is useless." (one study cannot prove anything)
Fix: "This data suggests that homework may be less effective than previously thought."
Module G — Citation Audit (APA "Accounting")
Run 2 matching checks:
Check 1 — Text-to-List (Orphan Check)
- Every name cited in the body text MUST appear in the Reference List
- Missing from list = technically plagiarism
Check 2 — List-to-Text Check
- Every source in the Reference List MUST be cited in the body text
- Sources listed but never cited = "Bibliography Padding" (delete them)
Check 3 — Et al. Rule
- 1–2 authors: always list both → (Smith & Jones, 2020)
- 3+ authors: use "et al." immediately → (Lee et al., 2020)
- Correct punctuation: "et al." (period after "al" only; NOT "et. al." or "et al")
Check 4 — Reference List Formatting
- Hanging indent: first line flush left, second line indented
- Italics: Journal/book title italicized, NOT the article title
- Article: regular text ("The history of AI in education")
- Journal: Journal of Computer Science (italicized)
Module H — Visual Data (Tables and Figures)
For a Literature Review, the expected visual is a Summary of Included Studies Table (usually in Chapter 2 or 3).
Check the table against APA rules:
| Requirement | APA Rule |
|---|---|
| Structure | Author/Year |
| Lines | Horizontal only (top, bottom, under header) — NO vertical lines |
| Label | "Table 1" above the table (Bold) |
| Title | Italic title below the label |
| In-text reference | Must mention the table in the text: "As shown in Table 1..." |
Required Deliverable Structure (Final Paper Checklist)
Before declaring the paper ready, verify all 8 sections are present:
- Title Page (APA Style)
- Abstract (Structured, ~200 words, following B-O-M-R-C)
- Introduction (Context + Research Question)
- Methodology (Search Strategy: database, N, date range)
- Findings (3 Themes with citations)
- Discussion (Interpretation + Limitations with hedging language)
- Conclusion (Summary + Implication)
- References (APA format, hanging indent, correct italics)
Summary Self-Audit Checklist
Use this for a rapid pass/fail scan. Read the full module above for any FAIL item.
Part A — Content Quality
- [ ] Alignment Test: RQ (Ch.1) and Conclusion (Ch.5) match perfectly?
- [ ] Density Test: Every paragraph in Findings contains at least 2 different citations?
- [ ] Method Test: Methodology explicitly states the database(s) used and N (number of papers)?
- [ ] Logic Test: Discussion uses hedging words (suggests, implies, may) not absolutes (proves, always)?
- [ ] Honesty Test: Limitations section acknowledges the specific weaknesses of this review?
Part B — Technical Accuracy
- [ ] Orphan Test: Every in-text citation name appears in the Reference List (and vice versa)?
- [ ] APA Test: References use Hanging Indent? Journal titles are in Italics?
- [ ] Visual Test: If a table is included, does it have a Label (Table 1), italic title, and no vertical lines?
- [ ] Abstract Test: Does the Abstract contain the sentence "This review analyzed [N] studies..."?
Reference Files
references/unit14-concepts.md— Full teaching notes and examples for each concept (A–H)references/apa-quick-reference.md— APA formatting rules for in-text citations and reference lists
Output Format
REMINDER: All output below MUST be in English (see Language Policy above).
When running a full paper review, structure the response as:
# EAS Term Paper Self-Check Report
## Overall Verdict: [READY TO SUBMIT / NEEDS MINOR REVISIONS / NEEDS MAJOR REVISIONS]
## Module-by-Module Results
### A. Abstract — [PASS / NEEDS REVISION]
...
### B. Title — [PASS / NEEDS REVISION]
...
### C. Vertical Alignment — [PASS / NEEDS REVISION]
...
### D. Terminology Consistency — [PASS / NEEDS REVISION]
...
### E. Evidence Density — [PASS / NEEDS REVISION]
...
### F. Tense and Modality — [PASS / NEEDS REVISION]
...
### G. Citation Audit — [PASS / NEEDS REVISION]
...
### H. Visual Data — [PASS / NEEDS REVISION / N/A]
...
## Priority Fix List (Top 3 Issues)
1. [Most critical issue + specific fix]
2. ...
3. ...
## Final Submission Checklist
[Filled-in checklist]
