Identify additional journals where people in WASH research my publish

[larnsce]

Which other journals could be assessed? Can we use filters to specifically identify WASH research articles that are not in PLOS Water, WASHDEV, UNC Newsletter. What about ES&T?

Prepare a list of keywords related to WASH (e.g. faecal sludge management).

[larnsce]

Proposal for identifying additional journals and filtering WASH articles.

Candidate journals, grouped by how well we could assess data availability

Programmatically accessible DAS (best next candidates):

• PLOS Global Public Health, PLOS ONE, PLOS NTDs (WASH subsets): same API and XML pipeline as PLOS Water (New data source: PLOS Water via the PLOS API (pure R, no scraping) #14), mandatory DAS. Filter by keyword within the journal.
• npj Clean Water (Nature/Springer): DAS section standard; full text on nature.com, TDM API available.

IWA stable (same scraper as washdev, #11):

• Journal of Water and Health
• H2Open Journal
• Water Practice & Technology

High-relevance, harder platforms (DAS in HTML but per-publisher work):

• Environmental Science & Technology and ES&T Water (ACS)
• Water Research, International Journal of Hygiene and Environmental Health, Science of the Total Environment (Elsevier)
• Water Resources Research (AGU/Wiley)
• American Journal of Tropical Medicine and Hygiene
• Tropical Medicine & International Health (Wiley)
• Waterlines (Practical Action; historically weak DAS practice, interesting as contrast)

Finding WASH articles outside fixed journals

Rather than scraping journal by journal, query OpenAlex (open API, no key, polite pool): filter works by title/abstract keyword match, publication year, and exclude the sources already covered. OpenAlex also returns OA status, funder, and country fields we do not currently collect. Crossref works as a fallback.

Draft keyword list (seed for data-raw/wash-keywords.csv)

water sanitation and hygiene; WASH; drinking water; water supply; water treatment; household water treatment; point-of-use treatment; water quality; water safety plan; sanitation; safely managed sanitation; faecal sludge management; fecal sludge; excreta; latrine; pit latrine; container-based sanitation; open defecation; sewerage; wastewater treatment; faecal contamination; E. coli; handwashing; hand hygiene; hygiene behaviour; menstrual hygiene; menstrual health; water security; water scarcity; groundwater contamination; rural water supply; water point; borehole; water kiosk; greywater; water reuse; diarrhoeal disease; cholera; schistosomiasis; soil-transmitted helminth; SDG 6

Happy to turn this into: (1) a keywords CSV in data-raw, (2) an OpenAlex query script that reports how many candidate articles each journal/keyword yields, so the next source is chosen on evidence rather than guesswork. Say the word and I will open that as its own scoped issue.

[larnsce]

Researcher-first mapping: prominent WASH & FSM researchers → their journals → how "WASH" those journals really are

This complements the journal-first proposal above by inverting the question: instead of guessing which journals matter, start from the people who dominate the field and see where they actually publish. Plan first, then best-effort results, then a judgement on opening up the scope.

The plan (reproducible via OpenAlex)

Note: outbound API calls are blocked in the sandbox this analysis was drafted in, so the journal-share percentages below are informed estimates, flagged as such. The plan makes every number reproducible; the script belongs in data-raw/.

1. Seed + expand the researcher list. Start from a curated seed (~30 names, below). Expand empirically: query OpenAlex works matching WASH/FSM keywords (list in previous comment) for 2011–2026, group_by=authorships.author.id, keep the top ~50 authors by works count and separately by citation count. Union with the seed; drop false positives manually.
2. Index their publications. For each author ID: GET /works?filter=author.id:A...,from_publication_date:2011-01-01 (paginated). Store as a tidy table: author, work, year, venue, DOI, OA status.
3. Rank venues. Group the union of those works by primary_location.source.id → a ranked list of journals weighted by how many prominent-researcher papers they carry.
4. Compute the WASH-share per journal. For each candidate journal: share = (works in journal matching WASH/FSM keyword search, 2011–2026) / (total works in journal, 2011–2026). This replaces my estimates below with real numbers.
5. Decision rule for inclusion in washopenresearch:
   • share > ~50% → index the whole journal (like washdev);
   • share low but absolute WASH count > ~100 papers and DAS programmatically accessible → index a keyword-filtered slice;
   • otherwise skip the journal and rely on the article-level OpenAlex sweep.

Step 1 draft: the prominent researchers (2011–2026)

WASH epidemiology / health (the trial generation): Thomas Clasen (Emory), Stephen Luby (Stanford), Amy Pickering (UC Berkeley), Jack Colford & Benjamin Arnold (Berkeley/UCSF), Oliver Cumming (LSHTM), Sandy Cairncross (LSHTM), Val Curtis (LSHTM, †2020), Matthew Freeman (Emory), Joe Brown (UNC), Wolf-Peter Schmidt (LSHTM), Daniele Lantagne (Tufts — the most frequent WASHDEV author per Research.com's journal profile).

WASH systems, policy, monitoring: Jamie Bartram (UNC/Leeds), Barbara Evans (Leeds), Juliet Willetts (UTS), Guy Howard (Bristol), Jenna Davis (Stanford), Robert Bain & Rick Johnston (JMP/UNICEF/WHO), Mark Sobsey (UNC).

FSM / on-site sanitation engineering: Linda Strande (Eawag-Sandec, 6k+ citations), Damir Brdjanovic & Konstantina Velkushanova (IHE Delft), Mariska Ronteltap, Chris Buckley (UKZN, †2021), Thammarat Koottatep & Chongrak Polprasert (AIT), Charles Niwagaba (Makerere), Kara Nelson (Berkeley), Francis de los Reyes III (NC State), Björn Vinnerås (SLU). The field's anchor references are the IWA books Faecal Sludge Management (2014) and Methods for Faecal Sludge Analysis (2021) — a reminder that a chunk of FSM output is books/reports, not journal articles, and will never be caught by journal scraping.

Steps 2–4 draft: where they publish, and how much of each journal is truly WASH/FSM

Estimated share = fraction of the journal's 2011–2026 output that is genuinely WASH/FSM (to be replaced by the step-4 computation).

Journal	Who publishes there	Est. WASH/FSM share	Verdict
WASHDEV	everyone above	~95–100%	already indexed
Waterlines	practice-side authors	~90%	whole journal, but weak DAS
Journal of Water and Health	Bartram, Sobsey, Lantagne	~40–60% (also high-income recreational/drinking water)	whole journal, IWA scraper reuse
PLOS Water	Pickering, Brown, Strande	~25–40%	whole journal (small), already planned (#14)
H2Open / npj Clean Water	mixed	~15–30%	filtered slice
Int. J. Hygiene & Environmental Health	Clasen, Cumming, Bartram (former EiC)	~15–25%	filtered slice — underrated candidate
Environ. Sci.: Water Research & Technology	Nelson, Pickering	~10–15% (WASH explicitly in scope)	filtered slice
Trop. Med. & Int. Health / Am. J. Trop. Med. Hyg.	Clasen, Luby, Freeman	~5–10%	filtered slice
ES&T + ES&T Water	Pickering, Brown, Nelson, Strande — e.g. the WASH Benefits environmental papers	~2–5% of a huge journal	filtered slice only — never whole-journal
Water Research	Strande, Nelson, Buckley (FSM treatment)	~2–5%	filtered slice only
Lancet Global Health / PLOS Med / BMJ Global Health	the flagship trials (WASH Benefits Kenya)	<2%, but the highest-impact WASH papers of the decade	article-level capture only
Sci. Total Environ. / Environ. Sci. Pollut. Res. / Environ. Monit. Assess.	long tail	<2% each, though high absolute counts (STOTEN tops WASH bibliometrics by volume)	article-level capture only
J. Environmental Management / Water SA	FSM treatment papers	low	article-level capture only

Judgement: should the scope open further?

Yes on topics, no on whole journals. Three conclusions fall out of the researcher-first view:

1. No unindexed journal is majority-WASH except Waterlines and (arguably) Journal of Water and Health. Everything else prominent researchers use is a generalist venue where WASH is a small slice. Adding whole journals beyond the IWA stable buys mostly noise.
2. The field is split across three publication cultures — environmental science/engineering (ES&T, Water Research), global health (Lancet GH, TMIH, AJTMH), and core WASH (WASHDEV, JWH). A journal-only strategy anchored in the WASH core systematically misses the highest-impact WASH research, which lands in the first two cultures. So yes: open the topic scope to WASH-health epidemiology and on-site sanitation/FSM engineering, and capture those via article-level keyword filtering (OpenAlex), not journal indexing.
3. FSM specifically is small and concentrated: a few hundred journal articles total, centred on WASHDEV, Water Research, J. Env. Management, plus books. It needs its own keyword block (faecal/fecal sludge, on-site sanitation, pit latrine emptying, container-based sanitation, sludge dewatering/drying beds, co-treatment, resource recovery from excreta) rather than its own journal list.

Proposed next step: I turn steps 1–4 into an R script in data-raw/ (openalexR is on CRAN) that outputs (a) the empirical researcher ranking, (b) the venue ranking, (c) real WASH-shares per journal to replace the estimates above — then we apply the decision rule in step 5. Can open that as a scoped issue/PR on request.

Sources: Research.com WASHDEV profile · ESPR WASH bibliometric · Strande Google Scholar · IWA FSM book · WASH consensus statement, BMC Medicine · ES&T WASH Benefits paper · ES:WR&T scope

---

Generated by Claude Code

[larnsce]

Decision: analysis window set to 30 years (1996–2026) — and what a bibliometric paper from this could look like

Why 30 years, not 15

The 15-year window in the researcher-first plan above is the wrong window for measuring prominence and identifying journals, though it remains the right window for the data-availability indexing. Three reasons:

1. A 15-year window excludes the field's canon. The most-cited works in WASH are older: Esrey's reviews (1991), Curtis & Cairncross on handwashing (2003), Fewtrell & Colford's meta-analysis (2005), and most of the careers of Cairncross, Feachem, and Duncan Mara. Ranking researchers on 2011–2026 output systematically places the trial generation above the people who built the field — a recency artifact, not a prominence measurement.
2. The field had no dedicated journals before ~2011 (WASHDEV Vol. 1). Pre-2011 WASH research lived entirely in generalist venues — Tropical Medicine & International Health, Bulletin of the WHO, Water Research, Waterlines, WEDC proceedings. The question "which journals carry WASH research?" is hardest and most interesting exactly in 1996–2011, which the 15-year window discards.
3. The DAS baseline argument. Data availability statements became mandatory around 2014 (PLOS) and widespread after ~2017. If washopenresearch wants to show change in open-data practice rather than just describe current practice, we need the pre-mandate baseline: a 30-year window gives ~20 years of "before" against ~10 of "after."

Two-tier design going forward:

• Tier 1 — prominence/venue mapping (steps 1–4 above): 1996–2026. Fully automated via OpenAlex.
• Tier 2 — DAS/open-data indexing: unchanged (journal inception or ~15 years), because the open-data question is only answerable there.

Known cost: OpenAlex/Crossref metadata degrades before ~1995 (missing abstracts, weak author disambiguation, unindexed grey literature), which is one reason to stop at 30 years rather than 50. The 1980s grey-literature era (World Bank TAG, IRC, WEDC) is a historical project, not a metadata harvest.

---

What "a publishable bibliometric paper" concretely means

Since this affects whether a separate dataset is worth building, here is the genre, the realistic angles, and what each requires.

The genre. Bibliometric / science-of-science papers define a corpus (keyword + venue query in OpenAlex/Scopus/WoS), then analyse: growth over time, geography of authorship, venue distribution, co-authorship networks, keyword co-occurrence and topic evolution, citation structure. WASH already has routine examples — the ESPR analysis of global WASH research and a JWH bibliometric on drinking-water research in LMICs. These are publishable but low-impact: they describe the corpus without answering a question nobody else can answer. If we do this, we should not write that paper.

Three angles where openwashdata has a real edge:

1. Open-data practice in WASH research (strongest — on-mission and unique). Question: how has data sharing in WASH research actually changed, 1996–2026, and does stated availability (DAS) match real availability? Nobody has done this for WASH. The manually curated DAS fields in washdev (and PLOS Water / New data source: PLOS Water via the PLOS API (pure R, no scraping) #14) are data no one else holds, and they serve as a hand-validated gold standard against which the automated 30-year corpus is benchmarked. This is meta-research; venues: PLOS Biology (meta-research section), PeerJ, PLOS ONE, Scientific Data, Quantitative Science Studies, or WASHDEV itself for sector visibility. This paper is essentially the research output the package name promises.
2. Field-formation history. How WASH became a field: venue migration from generalist journals to a dedicated core, the three publication cultures (env-eng / global health / core WASH), generational turnover. Venues: WIREs Water, Water Alternatives, npj Clean Water (perspective), H2Open Journal. Interesting, but it's the paper anyone with OpenAlex could write; our edge is modest.
3. Authorship equity. LMIC authorship representation and "helicopter research" patterns in WASH, 1996–2026 — first/last-author affiliation countries vs. study-site countries. Topical, on-mission, and the affiliation-country fields already exist in the washdev schema. Venues: BMJ Global Health, PLOS Global Public Health, WASHDEV.

What the separate dataset would be. A second, fully script-generated package (working name washbibliometrics or a data-raw corpus in this package): one row per work in the WASH corpus 1996–2026 — DOI, title, year, venue, authors, affiliations + countries, OA status, citation count, funder, concepts/keywords — harvested via openalexR. Rough size: 30–60k works depending on keyword strictness. Crucially it is cheap in the dimension where washdev is expensive: zero manual curation, fully reproducible, refreshable with one script run. It is a different artifact from the hand-curated DAS datasets and should not be mixed into them.

Low-lift publication regardless of the paper: the corpus itself can be published as a data descriptor (Scientific Data, Data in Brief) — openwashdata already mints Zenodo DOIs, so a data paper is a small step and gives the dataset a citable identity even if the full bibliometric analysis never happens.

Decision criteria for "is a separate dataset worth it":

• Yes, if we want angle 1 (open-data meta-research) or a citable data paper — the corpus is the substrate for both, and its marginal cost is small because steps 1–4 of the prominence mapping already require harvesting most of it. The dataset is nearly a by-product of work we've committed to.
• No, if the only goal is choosing the next journals to index — the intermediate outputs of the mapping script answer that without curating and documenting a full corpus.

Effort estimate: harvest + tidy script ~1–2 weeks part-time; data descriptor ~1–2 weeks more; the full meta-research paper (angle 1) is a multi-month project and the only genuinely expensive item.

Recommendation: build the 30-year OpenAlex corpus as a scripted by-product of the prominence mapping, publish it as a data paper, and decide on the bibliometric paper (angle 1 preferred) after seeing the corpus — that ordering front-loads the cheap, reusable work and defers the expensive commitment until the data can justify it.

---

Generated by Claude Code

[larnsce]

Open-data practice in WASH research

That's in preparation and builds on this dataset. The bibliometric dataset needs a new issue and additional data package. Consider publishing both datasets in data journal, then build manuscript on that data.

Authirship equity

We have discussed this idea to be used as an indicator against our own research. It's a new idea to write a paper that shows historic data and changes over time. Paper needs a separate repository and goes beyond the research around data stewardship and open data practices in the wash sector. It could be a follow-up paper based on an updated dataset in 2027.

[larnsce]

Scoped out and staged

Actioned the agreed direction:

• New data package: WASH & FSM bibliometric corpus (1996–2026) from OpenAlex #22 — new data package: the WASH & FSM bibliometric corpus (1996–2026, OpenAlex). Separate package, as decided.
• Placeholder: authorship-equity analysis of WASH research (possible 2027 follow-up) #23 — placeholder for the possible 2027 authorship-equity follow-up, with the schema constraint recorded so it survives until then.
• Branch claude/wash-fsm-research-analysis-xc5fpn now carries data-raw/washbiblio/: wash-keywords.csv (blocked keyword list), SCHEMA.md (works + authorships), and harvest_corpus.R (OpenAlex skeleton).

This issue stays open until harvest_corpus.R runs on a networked machine and produces journal_wash_share.csv — the real per-journal WASH-shares that replace the estimated table above. At that point the decision rule (whole journal / filtered slice / article-level only) gets applied to real numbers and this closes. The keyword list here is now the seed in wash-keywords.csv; refine it there.

---

Generated by Claude Code