Transparency & Data (cps-transparency-and-data)
CPS enforces political-science replication norms. Papers presenting quantitative analyses will not be
granted final acceptance until replication materials (data, code, log files, etc.) are deposited at the
CPS Dataverse on Harvard Dataverse, and every paper needs a data availability statement. Build the
package as you go so final acceptance does not stall — and follow DA-RT-style transparency for
qualitative evidence too.
When to trigger
- Building the replication/reproducibility package and the data availability statement
- A manuscript is near acceptance and the editorial office expects deposited materials
- Data cannot be fully shared (privacy, ethics, provider restrictions) and you need a path
- Preparing a (optional) anonymized pre-analysis plan as supplementary material
What CPS requires (verify current wording — 检索于 2026-06；以官网为准)
- Deposit to the CPS Dataverse. Quantitative papers deposit replication materials — data, code,
log files, and documentation — at the CPS Dataverse on Harvard Dataverse
(
dataverse.harvard.edu/dataverse/cps). This gates final acceptance, not post-publication.
- Data availability statement. Every paper includes a statement saying whether data are available,
where, and — if available but not shared — why not.
- Reproducibility. Materials must let an independent researcher regenerate the manuscript's tables
and figures: master script + README + pinned versions + seeds.
- Restricted-data path. If data cannot be shared and sharing is a publication requirement, consult
the journal editorial office; explain the restriction and document how others can obtain the data.
- Preregistration (optional). Authors may submit an anonymized pre-analysis plan as
supplementary material; the journal provides it to reviewers on request. Mark registered vs.
exploratory analyses.
Qualitative / multi-method transparency (DA-RT spirit)
- Document sources, interviews, and fieldnotes so the evidentiary basis is auditable; use evidence
tables or active citation. Where confidentiality requires it, use controlled access (e.g., QDR).
- State clearly which claims rest on which evidence; do not let "the cases show" stand unsupported.
Sharing posture by evidence type
Choose the transparency route before drafting the data availability statement:
| Evidence type |
Default posture |
Documentation to include |
| Cross-national public datasets |
Share constructed data and code; cite original sources |
Source versions, merge keys, transformations, and exact download dates |
| Proprietary/admin data |
Share code, synthetic or redacted extracts where allowed, and access instructions |
License limits, access route, variable construction, and verification path |
| Interviews/fieldnotes |
Protect identities; share protocols, coding scheme, and evidence table when ethical |
Consent limits, anonymization method, and claim-to-evidence map |
| Text corpus/web data |
Share corpus identifiers or permissible text, plus scraping/cleaning code |
Collection date, inclusion rules, deduplication, language processing decisions |
| Pre-analysis plan |
Submit anonymized plan as supplement when used |
Registered vs. exploratory analyses marked in manuscript and code |
The data availability statement should mirror the table: what is shared, where, what is restricted, and
how a qualified reader can audit the claim without violating law, ethics, or provider terms.
Build-as-you-go checklist
Anti-patterns
- Treating the deposit as a post-publication afterthought (it gates final acceptance for quant papers)
- Depositing code that does not actually reproduce the printed tables/figures
- A personal URL or generic cloud link instead of the CPS Dataverse
- Omitting the data availability statement, or claiming "available on request" with no plan
- Undocumented, un-seeded, unpinned code that "works on my machine"
- Qualitative claims with no documented evidentiary basis
Output format
【Repository】CPS Dataverse (Harvard) — package staged? [Y/N]
【Reproduces tables/figures?】master script verified locally? [Y/N]
【Data availability statement】drafted? available where / why not?
【Documentation】README + provenance + seeds + pinned versions? [Y/N]
【Sharing posture】public / proprietary-admin / interview-fieldnote / text-corpus / PAP route chosen
【Restricted data?】explanation + access path + editorial-office note?
【Qualitative transparency】sources/evidence documented? [Y/N/NA]
【Pre-analysis plan】anonymized + registered/exploratory marked? [Y/N/NA]
【Next】cps-review-process
Supplementary resources
1---2name: cps-transparency-and-data3description: Use when preparing the replication and transparency materials for a Comparative Political Studies (CPS) manuscript. Quantitative papers cannot be finally accepted until replication materials are deposited at the CPS Dataverse; a data availability statement is required. Prepares the package; it does not waive requirements.4---56# Transparency & Data (cps-transparency-and-data)78CPS enforces political-science replication norms. **Papers presenting quantitative analyses will not be9granted final acceptance until replication materials (data, code, log files, etc.) are deposited at the10CPS Dataverse on Harvard Dataverse**, and every paper needs a **data availability statement**. Build the11package as you go so final acceptance does not stall — and follow DA-RT-style transparency for12qualitative evidence too.1314## When to trigger1516- Building the replication/reproducibility package and the data availability statement17- A manuscript is near acceptance and the editorial office expects deposited materials18- Data cannot be fully shared (privacy, ethics, provider restrictions) and you need a path19- Preparing a (optional) anonymized pre-analysis plan as supplementary material2021## What CPS requires (verify current wording — 检索于 2026-06；以官网为准)22231. **Deposit to the CPS Dataverse.** Quantitative papers deposit replication materials — **data, code,24 log files, and documentation** — at the **CPS Dataverse on Harvard Dataverse**25 (`dataverse.harvard.edu/dataverse/cps`). This gates **final acceptance**, not post-publication.262. **Data availability statement.** Every paper includes a statement saying whether data are available,27 where, and — if available but not shared — **why not**.283. **Reproducibility.** Materials must let an independent researcher regenerate the manuscript's tables29 and figures: master script + README + pinned versions + seeds.304. **Restricted-data path.** If data cannot be shared and sharing is a publication requirement, **consult31 the journal editorial office**; explain the restriction and document how others can obtain the data.325. **Preregistration (optional).** Authors *may* submit an **anonymized** pre-analysis plan as33 supplementary material; the journal provides it to reviewers on request. Mark registered vs.34 exploratory analyses.3536## Qualitative / multi-method transparency (DA-RT spirit)3738- Document sources, interviews, and fieldnotes so the evidentiary basis is auditable; use evidence39 tables or active citation. Where confidentiality requires it, use controlled access (e.g., QDR).40- State clearly which claims rest on which evidence; do not let "the cases show" stand unsupported.4142## Sharing posture by evidence type4344Choose the transparency route before drafting the data availability statement:4546| Evidence type | Default posture | Documentation to include |47|---------------|-----------------|--------------------------|48| Cross-national public datasets | Share constructed data and code; cite original sources | Source versions, merge keys, transformations, and exact download dates |49| Proprietary/admin data | Share code, synthetic or redacted extracts where allowed, and access instructions | License limits, access route, variable construction, and verification path |50| Interviews/fieldnotes | Protect identities; share protocols, coding scheme, and evidence table when ethical | Consent limits, anonymization method, and claim-to-evidence map |51| Text corpus/web data | Share corpus identifiers or permissible text, plus scraping/cleaning code | Collection date, inclusion rules, deduplication, language processing decisions |52| Pre-analysis plan | Submit anonymized plan as supplement when used | Registered vs. exploratory analyses marked in manuscript and code |5354The data availability statement should mirror the table: what is shared, where, what is restricted, and55how a qualified reader can audit the claim without violating law, ethics, or provider terms.5657## Build-as-you-go checklist5859- [ ] One **master script** regenerates **every** table and figure from raw/constructed data60- [ ] **README** documents data provenance, construction, and how to reproduce each exhibit61- [ ] **Seeds** set and reported for every stochastic step62- [ ] Software/package **versions pinned** (`renv.lock` / `requirements.txt` / recorded installs)63- [ ] Exhibit numbers in the manuscript **match** the package output exactly64- [ ] **Data availability statement** drafted (available where / why not shared)65- [ ] Restricted data: explanation + access instructions + editorial-office consultation noted66- [ ] Anonymized pre-analysis plan attached where applicable; registered vs. exploratory marked6768## Anti-patterns6970- Treating the deposit as a post-publication afterthought (it gates final acceptance for quant papers)71- Depositing code that does not actually reproduce the printed tables/figures72- A personal URL or generic cloud link instead of the CPS Dataverse73- Omitting the data availability statement, or claiming "available on request" with no plan74- Undocumented, un-seeded, unpinned code that "works on my machine"75- Qualitative claims with no documented evidentiary basis7677## Output format7879```80【Repository】CPS Dataverse (Harvard) — package staged? [Y/N]81【Reproduces tables/figures?】master script verified locally? [Y/N]82【Data availability statement】drafted? available where / why not?83【Documentation】README + provenance + seeds + pinned versions? [Y/N]84【Sharing posture】public / proprietary-admin / interview-fieldnote / text-corpus / PAP route chosen85【Restricted data?】explanation + access path + editorial-office note?86【Qualitative transparency】sources/evidence documented? [Y/N/NA]87【Pre-analysis plan】anonymized + registered/exploratory marked? [Y/N/NA]88【Next】cps-review-process89```9091## Supplementary resources9293- [`../../resources/official-source-map.md`](../../resources/official-source-map.md) — CPS data policy + CPS Dataverse URL94- [`../../resources/external_tools.md`](../../resources/external_tools.md) — reproducibility tooling and qualitative-transparency options (QDR)
brycewang-stanford