Position
This file says what is known, what is not, and what this repository adds.
Known
- Student writing has a corpus-based genre map (Gardner 2013); the four report kinds are not named in it.
- A technical-report standard prescribes components and says organisation became user-based (NISO 2010).
- Engineers and scientists say they read the conclusion and use the summary to decide (Pinelli 1982).
- Thesis structures have been described (Paltridge 2002) and examiners interviewed (Mullins 2002).
- French schools publish their own guides and grids; they differ in what they require.
Not known
| Gap | Note |
|---|---|
| Does an executive summary raise a reader's accuracy or speed on a report? | no study located |
| Does conclusion-first beat conclusion-last for reports with comprehension as the outcome? | no study located |
| What a host company reads in an internship report, and what it wants from it | no study located |
| Whether a given reading order helps an examiner | no test located; the order examiners say they follow is reported by interview: most of the 30 open the abstract, introduction, conclusion and references first, then read in full, and form a first impression by the end of the second or third chapter (Mullins 2002, full text) |
| Move structure of a French rapport de stage | pointers unverified |
| Content of ISO 7144 | paywalled |
"Nobody has done this" is not claimed: the searches were English and French on one engine, and several pages were closed to the reader. A negative here means "not found in this pool".
Where practice runs ahead of the evidence
- Summary first, in standards and in engineering practice, has survey and interview support (Pinelli 1982, Sheffield 2017) and no experiment in this pool. The sister review (claim-audit-q9) holds Britt and Larson 2003 (claim-first, three experiments, short texts) and Garner 2013 (d = 0.81), which this review does not cite.
- Schools' weights rest on no measure of what graders value: TSE 2017 weights form 7 and substance 13; INSPE 2021 weights conduct 5 and written 15.
Where the evidence is ignored by practice
- In free recall, headings help most for unfamiliar topics (Lorch 1996); the guides read name their sections by topic (introduction, methods, results, conclusion in EOST 2026) and none asks for informative headings. Whether that matters for reports is untested.
- The standard notes that an executive summary is often prepared for reports over 50 pages, at most 10 pages when it replaces the abstract. The two French school guides read (EOST 2026, TSE 2017) mention no executive summary. Only EOST 2026 asks for a résumé, in French and English, with at most five keywords.
What this repository adds
- A section-by-section table from academic convention to a reader-ergonomics principle, with a verdict on each conflict (the academic-report shape).
- A named baseline and a measurement plan (baseline).
- A statement of what the grader, the jury, the host company and a colleague each need first, taken from the sources above and flagged where a source is missing.