Sport Geography Review Bibliometric

This skill supports systematic reviews, scoping reviews, critical reviews, and bibliometric/scientometric manuscripts in sport geography and urban health geography. It owns literature-evidence synthesis before manuscript prose: corpus construction, paper-level extraction, Evidence Trace, cross-paper argument mapping, and review synthesis QC. It is designed for work that combines sports facilities, parks, green/heat exposure, accessibility, equity, spatial justice, and planning implications.

Use it to move from a broad topic and paper corpus to a journal-positioned review manuscript with reproducible search logic, defensible coding, meaningful bibliometric maps, and a policy agenda.

Core Contract

Do not treat bibliometrics as the contribution. The contribution must be one of:

• a clarified concept boundary;
• a systematic evidence map;
• a method evolution framework;
• a justice/equity coding framework;
• a journal-fit planning or policy agenda;
• a combined review + bibliometric route that explains why the field matters.

Default argument chain:

public health and active living pressure -> unequal spatial opportunity -> exposure/accessibility measurement gap -> evidence synthesis -> method and justice framework -> planning agenda

Fast Workflow

1. Position the manuscript

   • Decide whether the paper is a systematic review, scoping review, bibliometric review, or hybrid review.
   • Choose one target route: SCS/Cities, landscape-green-space journals, health-exposure journals, or broader geography/planning journals.
   • Read references/00_skill_design_principles.md when updating the skill itself.
   • Read references/01_positioning_and_journal_route.md.

2. Define scope and search logic

   • Separate object terms, exposure/accessibility terms, equity/justice terms, health/activity terms, and exclusions.
   • Use WoS + Scopus by default for bibliometrics; add PubMed/SPORTDiscus/Dimensions only when the research question needs health or sport coverage.
   • Read references/02_search_prisma_protocol.md.

3. Build the evidence corpus

   • Document search date, database, query, filters, de-duplication, and screening reasons.
   • Keep OA/non-OA status, DOI, journal, partition label, and manual-download status separate from inclusion decisions.
   • Use assets/PRISMA筛选记录模板.xlsx and assets/文献批判性编码模板.xlsx.
   • Record paper-level findings in the Evidence Trace with a full-text status and source locator. ABSTRACT_ONLY records must not receive invented section, page, table, figure, or supplement locations.

4. Run bibliometric analysis only after scope is stable

   • Use bibliometrix for performance analysis and thematic evolution.
   • Use VOSviewer for co-authorship, keyword co-occurrence, citation/co-citation, and overlay maps.
   • Use CiteSpace for burst references, timeline clusters, and turning points.
   • Read references/03_bibliometric_workflow.md.

5. Add critical manual coding

   • Code facility type, exposure metric, spatial scale, method family, population group, justice dimension, health pathway, and planning implication.
   • Use coding to answer what bibliometrics cannot answer.
   • Read references/04_critical_coding_framework.md.

6. Synthesize across papers before drafting

   • Build the Cross-paper Argument Map from Evidence IDs, preserving supporting and contradictory evidence and recording methodological, population, exposure/concept, spatial-scale, and temporal heterogeneity.
   • Create a paragraph synthesis contract before drafting Results, Discussion, or review Introduction paragraphs that make factual or synthesis claims.
   • Do not use citation count, paper count, semantic similarity, bibliometric co-occurrence, RF/SHAP output, or feature importance as evidence strength or causal mechanism.

7. Draft the manuscript

   • Make the review argument-led, not article-by-article.
   • Use bibliometric figures as evidence for field evolution, not decoration.
   • Translate findings into planning, governance, and equity implications.
   • Read references/05_manuscript_structure.md and references/06_figures_tables_policy_agenda.md.
   • For reusable writing prompts, prompt chains, reviewer audits, and policy-agenda prompts, read references/09_prompt_library.md.
   • For target-journal review and bibliometric paper patterns, read references/10_top_journal_review_patterns.md.

8. Audit before completion

   • Verify every citation and DOI used in the manuscript.
   • Check whether journal partition labels are current using the user's institution access to JCR/CAS.
   • Make sure the final narrative does not overclaim health benefit from exposure or accessibility.
   • Read references/07_quality_checks.md.

For end-to-end projects, this skill owns literature-search reproducibility, screening, high-impact paper classification, paper-level extraction, literature matrix, Evidence Trace, Cross-paper Argument Map, evidence gap taxonomy, PRISMA records, bibliometric figures, critical coding, and review synthesis QC. Handoff stable synthesis claims and boundaries to agent-auto-sci-methodology and agent-auto-sci-scicomm; those skills do not reconstruct evidence from prose memory.

Local Corpus

When available, use the user's curated local corpus as evidence:

• <your-literature-library>/00_表格与清单/中文分类整理总表_含PDF匹配与下一步路线.xlsx
• <your-literature-library>/02_PDF文本提取与匹配索引/全文结构索引_48篇.csv
• <your-literature-library>/03_下一步精读与Skill建设路线.md
• <your-literature-library>/02_PDF文本提取与匹配索引/prompt手册_提取文本.txt

Read references/08_local_corpus_guide.md before using these files.

Scripts

• scripts/create_review_templates.py: create PRISMA, critical coding, Claim-Evidence Map, Evidence Trace, Cross-paper Argument Map, paragraph synthesis contract, and journal-positioning templates in a chosen output directory.
• scripts/validate_review_templates.py: deterministically validate workbook sheets, headers, duplicate columns, reopenability, and controlled-vocabulary bindings without judging scientific evidence.
• scripts/audit_review_library.py: audit a local review library folder for expected tables, PDFs, extracted text, and DOI-match counts.

Must Not Do

• Do not use "sports park exposure" as the only topic if the corpus is actually green/open/sport space equity.
• Do not equate accessibility with actual use.
• Do not equate exposure with health benefit.
• Do not equate equality with equity or justice.
• Do not rely only on VOSviewer/CiteSpace screenshots for high-impact journals.
• Do not present JCR/CAS partition labels as confirmed unless the user has verified them from a current official source.
• Do not write a review as one-paper-at-a-time summaries; synthesize by concept, method, population, and planning implication.