Project Context

This file provides context for LLMs or collaborators interacting with this repository. It summarises the project background, current state, key decisions, and working constraints.

What This Repository Is

This repository publishes findings and methodology from a practice-based evaluation of the Hi-Audio web audio platform (hiaudio.fr), conducted by the Female Laptop Orchestra (FLO) between February and March 2026. The evaluation was carried out as part of the Hi-AUDiO research project, funded by the European Research Council (ERC) under Horizon Europe (grant No. 101052978).

The repository is public and hosted on GitHub Pages. It is intended to support open dissemination and reproducibility for other researchers.

Repo: https://github.com/idsinge/hiaudio_FLO_practice-based_evaluation GitHub Pages: https://idsinge.github.io/hiaudio_FLO_practice-based_evaluation/

Key People

Do not add names, email addresses, or personal identifiers to any file in this repository beyond what is already present.

Repository Files

File Status Purpose
README.md Published Main evaluation report, converted and edited from original .docx memo
docs/Practice-Based Evaluation of the Hi-Audio platform by FLO Lab .pdf Published Original evaluation memo in PDF format
docs/output.md Published Original Pandoc conversion from .docx — kept as historical reference, do not edit or delete
docs/participant_survey.md Published Working copy of the JotForm survey — data collection is complete, cleared for publication
docs/ANALYSIS_PLAN.md Published Planned approach for processing and analysing survey responses — data collection is complete, cleared for publication
docs/participant_survey_responses.csv / .xlsx Published Raw aggregate response data (9 rows), no name/email/IP columns
docs/individual_participant_responses/P01–P09.pdf Published Per-respondent submission exports, anonymised as P01–P09
docs/response_id_mapping.md Published Maps PID ↔ original JotForm submission ID/date
CONTEXT.md This file Context for LLMs and collaborators
CITATION.cff Published Structured citation for GitHub’s “Cite this repository” button
LICENSE Published CC BY 4.0
_config.yml Published Jekyll Cayman theme for GitHub Pages
images/ Published Screenshots (image1–8.png) and ERC_logo.png

Data collection status: data collection closed. docs/participant_survey.md, docs/ANALYSIS_PLAN.md, and the response data were withheld from the public repo only to avoid contaminating participant responses before they replied; that risk no longer applies, and all are now published, moved into docs/. invitation_message.md was removed from the repo entirely (not published, not withheld — deleted).

A full PII scan was run across the response CSV (all 39 columns × 9 rows, before the Q16/Q17 column correction below reduced it to 33 columns) and all 9 individual PDFs before publication: no names, emails, phone numbers, or IP addresses were found, with one accepted exception — see “Nela Brown / P09” below.

The Survey

The participant survey complements the practice-based evaluation memo. It was administered via JotForm: https://form.jotform.com/hiaudioparis/flo-lab-evaluation (Form ID: 261313989509366)

The survey title is: FLO Lab Hi-Audio Evaluation — Participant Survey

The .md file is a working copy that mirrors the live JotForm. The .md version uses [available on request] in place of the institutional email address to avoid publishing contact details in the public repo.

Survey Structure (JotForm Q1–Q23)

Consent and Ethical Notice — full GDPR-compliant text, required before proceeding to Section 1

Section 1: User Profile & Background (Q1–Q6)

Section 2: Hi-Audio Recording Technical Setup (Q7–Q15)

Section 3: Exploring Usability and Usefulness (Q16–Q23)

Q16/Q17 data correction (already applied)

Q16 and Q17 are single questions on the live form — confirmed directly in the JotForm builder, where each has exactly one matrix row with a blank row-name field (shown as JotForm’s own “Type Row Name” placeholder, not an actual label; text like “Service Quality”/”Cleanliness”/”Responsiveness”/”Friendliness” was never part of the form). The original JotForm CSV export nonetheless split each into 4 sub-columns suffixed with that bogus text — a JotForm export defect, not real data. All 4 sub-column values were identical for every one of the 9 respondents before correction, confirming nothing was lost by collapsing them. docs/participant_survey_responses.csv/.xlsx have already been corrected to a single clean column per question. See docs/ANALYSIS_PLAN.md for the full explanation.

Widget Questions (not rendered in JotForm PDF export)

Q3, Q4, Q6, Q7 use configurable list or multiple entry widgets. They are present and active in JotForm but do not appear in the PDF export. They are confirmed present in the live form.

Key Design Decisions

Framework language

The survey items for Q18–Q20 were designed for this study, informed by established usability and technology acceptance evaluation approaches. They adapt common constructs (pragmatic usability, hedonic quality, perceived usefulness, behavioural intention) to the specific context of collaborative online audio recording.

Do not describe the survey as directly reproducing or formally applying named instruments such as USE Questionnaire, AttrakDiff, UEQ, or TAM. The correct framing is “informed by” or “adapts constructs from.” This applies to all public-facing documents including README.md, docs/participant_survey.md, and CONTEXT.md.

Scale

Q16–Q20 all use the same 7-point “Strongly disagree → Strongly agree” scale. Q16 and Q17 are single-item ratings; Q18, Q19, and Q20 are multi-item matrices.

Headline metric

Q16 (overall suitability for professional/artistic use) is the primary headline metric, with Q17–Q20 as supporting evaluative items — usable as designed.

Sections and ordering

The Acknowledgements section is always last in README.md — after appendices. Do not move it.

docs/output.md

This file is the original Pandoc conversion from the .docx memo. It is kept as a historical reference. Do not edit, overwrite, or delete it.

Analysis Approach

The analysis is exploratory and descriptive. Inferential statistics are not planned given the expected small sample size.

Q17 sits at the boundary — report its mean alongside Q16, Q18–Q20, but also read it qualitatively alongside Q21 and Q22 when interpreting artistic fit.

Privacy and Anonymity Constraints

Nela Brown / P09 — dual role disclosure

Nela Brown designed the questionnaire and authored the evaluation memo, and is also P09 in the response dataset (submitted 2026-07-03). This is disclosed in README.md §3 (Methodological Approach) as a “Positionality note,” and in docs/ANALYSIS_PLAN.md under “Researcher Positionality,” which specifies how to handle her response during analysis (report n=9 as primary, additionally compute n=8 with P09 excluded for headline metrics, flag P09-sourced excerpts in thematic coding). Her name appearing in P09’s Q7 free-text answer (a track filename referencing her) is treated as consistent with this disclosure, not as a separate PII leak — she is already the named public author of this report.

What Comes Next

  1. Process the response data (docs/participant_survey_responses.csv) per docs/ANALYSIS_PLAN.md, following the “Researcher Positionality” guidance for P09
  2. Produce analysis_summary.md (not committed until reviewed and approved)
  3. Findings are documented and added to the repository

Current Session Status

This section captures the latest working state so the project can be resumed easily in a later session.

Completed in the latest review cycle

Decisions made

  1. Process the response CSV per docs/ANALYSIS_PLAN.md, applying the Researcher Positionality guidance
  2. Produce analysis_summary.md (descriptive stats, thematic coding, triangulation) — keep local until reviewed and approved