eursr-writing-style

Writing Style (eursr-writing-style)

An ESR paper must be readable by a comparative-minded quantitative sociologist, formatted to OUP/ESR

conventions, and disciplined to a target length that includes endnotes and references. This skill

is about reaching ESR's audience and respecting the format — not generating claims.

When to trigger

• Drafting the introduction, framing the contribution, or final polish
• Over the ~8,000-word target and needing to cut without losing the argument
• Writing the ≤200-word abstract
• Aligning citations and format to OUP/ESR conventions (author-date, in English)

Reach the comparative quantitative audience

1. Front-load the contribution. By the end of the introduction the reader knows the question, the

mechanism and hypotheses, the comparative/longitudinal leverage, and what the paper changes.

1. Lead with the mechanism, support with models. State the macro–micro argument (from

eursr-theory-building) in words first; the estimates serve it. Avoid a results-first narration.

1. Translate the estimates. Report what a coefficient means (a predicted-probability scenario, a

marginal effect), not just its sign and stars — the reader thinks substantively.

1. Acceptable English. ESR considers only papers in good English; clarity and concision are the

author's responsibility. Spell out acronyms; define harmonization schemes (ISCED/CASMIN, ISEI/EGP)

on first use.

Format to OUP/ESR conventions

• Citations: author-date, parenthetical (e.g., "(Gans, 1962)"); one consistent style (manage with

Zotero/BibTeX). Endnotes and references count toward the ~8,000 words.

• Anonymity: no author-identifying information anywhere in the manuscript; you **may cite your own

work** but must word it so it does not identify you ("in our prior study" → "in a prior study").

• Abstract: ≤200 words, with no identifying information; keywords as requested.
• Language: English; figures/tables placed per OUP instructions.

Fit the length (incl. text + endnotes + references; tables/figures excluded)

• Tighten the literature review — engage the debate, not every paper (see eursr-literature-positioning).
• Endnotes and references count, so prune redundant citations and over-long reference strings; keep

endnotes lean.

• Move extended robustness, invariance tables, and harmonization detail to the (supplementary) appendix.
• Longer-than-8,000 papers are accepted but reviewers are asked whether the contribution justifies the

length — make every page earn its place.

Prose moves an ESR reader needs

| Weakness | Comparative reader experiences | Prose fix |

|----------|--------------------------------|-----------|

| Results-first narration | can't find the argument | lead with the mechanism + hypotheses |

| Stars-only interpretation | doesn't know the magnitude | translate to a predicted-probability scenario |

| Coding schemes undefined | stalls on ISCED/EGP | define harmonization on first use |

| Over-long reference strings | length burned, argument thin | prune; endnotes + references count |

| Self-citation in identifying voice | anonymity broken | "in a prior study," not "in our prior study" |

Worked micro-example (illustrative)

A multilevel-study introduction is rewritten to reach the ESR audience.

Before: "We estimate random-slope models of the SES gradient in test scores across PISA waves..."
  → method-first; the reader doesn't know the argument
After: "Does early tracking widen the social gap in achievement by forcing high-stakes choices on
  families under uncertainty? We argue it shifts the gap from performance to choice..." → mechanism
  first, hypothesis visible, comparison implied
Abstract: trimmed 236 → 192 words, finding plain, no institution named → within ≤200
Length: pruned 35 redundant citations + 4 long endnotes, recovering ~700 words toward the ~8,000 target

The contribution leads, the method follows, and the prune is treated as a budget item.

Referee pushback → ESR-specific fix

• "Reads as a results report." → Move the mechanism and hypotheses into the first two pages so the

argument leads the estimates.

• "What does this coefficient mean substantively?" → Translate it into a predicted-probability or

marginal-effect scenario the reader can picture.

• "Over length." → Tighten literature engagement, prune citation strings and endnotes; references and

endnotes count, tables and figures do not.

Calibration anchors

• Write for the comparative quantitative reader. ESR's reader thinks in mechanisms and magnitudes —

lead with the argument and translate the estimates.

• Endnotes and references are inside the target. ESR counts them toward ~8,000 words — budget

citations from the first draft.

• Confirm length/abstract specifics against the current OUP guidelines if exact limits are

decision-relevant; the broad shape (~8,000; ≤200-word abstract) is stable.

Anti-patterns

• A method-first or results-first intro that never states the mechanism or comparative stakes
• Reporting estimates as stars without substantive translation
• An abstract over 200 words or one that hides the finding
• Self-references worded so they break anonymity
• Treating endnotes and references as "free" — they count toward ~8,000 words

Output format

【Contribution stated by end of intro?】[Y/N]
【Mechanism-first, estimates translated?】[Y/N]
【Abstract】word count (≤200), non-identifying?
【Word count】Article ~8,000 incl. text + endnotes + references? (longer justified?)
【OUP style + anonymous + author-date】[Y/N]
【Next】eursr-transparency-and-data

Supplementary resources

• [../../resources/official-source-map.md](../../resources/official-source-map.md) — word/abstract caps, OUP citation style, anonymity, formatting