Health data you can trust Personal proposal by Per-Arne Andersen H3 · Spring 2027 https://thesis.uya.no/proposals/can-you-trust-this-number/ Show conflicting values, outdated observations and missing fields when two health records are combined. TECHNICAL - Create two CSV exports for the same fictional patients. - Build a merged record view with source labels and data-quality warnings. DATA (planned) 50 fictional patients in two CSV files Create two CSVs with 50 observations each for 50 fictional patients: patient ID, event ID, measurement, value, unit, time and source. Plant 10 duplicate events, 10 incompatible measurement–unit pairs and 10 missing timestamps in disjoint records. Include clean repeated measurements and convertible units. METHOD Automated checks against all 30 planted faults; then 8 health-information students complete 8 record-review tasks. Report detection precision, recall and task errors. OUTPUT Record viewer, test data and error-detection report. BACHELOR Task: Import the two health-record CSVs and flag duplicate events, bad units and missing times using fixed rules. Data: 50 fictional patients in two CSV files (planned) Create two CSVs with 50 observations each for 50 fictional patients: patient ID, event ID, measurement, value, unit, time and source. Plant 10 duplicate events, 10 incompatible measurement–unit pairs and 10 missing timestamps in disjoint records. Include clean repeated measurements and convertible units. Requires: Recruit students; use fictional records only. Method: Test against the independent defect key; observe 4 students reconciling records. Output: A working data-quality viewer, tests and user instructions. Use relevant literature to justify the established approach; a new research contribution is not the aim of this proposal. MASTER Investigate whether quality warnings help reconciliation equally for missing, conflicting and duplicate data. Derive separate predictions from data-quality and cognitive-fit research; test them by defect type in the same viewer. Intended contribution: Design principles relating defect type, warning format and reconciliation decisions. STARTING PAPERS Wang & Strong (1996) — Beyond Accuracy: https://www.tandfonline.com/doi/abs/10.1080/07421222.1996.11518099 Distinguish accuracy, completeness, representation and fitness for the task. Vessey & Galletta (1991) — Cognitive Fit: https://pubsonline.informs.org/doi/10.1287/isre.2.1.63 Test whether a representation helps one kind of task more than another. Search Scopus or Web of Science and ACM Digital Library using the topic query, then follow citations to the thesis start date. Record searches and compare methods, data, findings and limitations in literature-matrix.csv. Use that review to confirm or revise the gap and choose a current comparator. The linked papers are starting points. START HERE Tools: Python, pandas, Streamlit 1. Write 5 fictional patients in both exports; preserve a clean reference copy. 2. Add one fault of each type and record its expected detection in the answer key. 3. Test the merge on those records before expanding to all 50 patients. Literature search: health records data provenance information quality trust Study controls: - Before collecting participant data, agree consent, storage and withdrawal handling with the supervisor. Use participant codes, not names, in study files. - Define duplicates by event identity, not equal values. Preserve both original records. Score warnings by defect type against an independent hidden key; do not silently repair records. - For the master’s study, use the research task above to define the factors and comparisons in this pilot plan. Preregister one primary outcome and feasible scope after the literature review; do not add every possible model or interface variant. REQUIRES Recruit 8 students; use fictional records only. -------------------- NON-TECHNICAL - Prepare merged records with and without source and quality labels. - Ask participants which values they would trust and why. DATA (planned) 12 fictional merged records + 8 sessions Create 12 paper records: 6 clean controls and 2 each with a unit conflict, missing time or stale value. Prepare source-labelled and source-hidden versions with identical values, units and times; never show the defect key. METHOD Within-person think-aloud study with balanced order. Count correctly identified faults; code reasons for trusting a value. OUTPUT Annotated record examples and findings on trust in data. BACHELOR Task: Identify which record labels users need to resolve conflicting values. Data: 12 fictional merged records (planned) Create 12 paper records: 6 clean controls and 2 each with a unit conflict, missing time or stale value. Prepare source-labelled and source-hidden versions with identical values, units and times; never show the defect key. Requires: Recruit health-information students. Method: Use the paper records in 4 task sessions; describe problems using established data-quality categories. Output: A record-layout recommendation and data-quality checklist. Use relevant literature to justify the established approach; a new research contribution is not the aim of this proposal. MASTER Explain when provenance changes trust and when task-relevant completeness matters more. Compare matched records using data-quality dimensions; examine contradictory trust/verification decisions. Intended contribution: A refined explanation of how users judge record fitness for a task. STARTING PAPERS Wang & Strong (1996) — Beyond Accuracy: https://www.tandfonline.com/doi/abs/10.1080/07421222.1996.11518099 Distinguish accuracy, completeness, representation and fitness for the task. Orlikowski & Gash (1994) — Technological Frames: https://dl.acm.org/doi/10.1145/196734.196745 Compare how roles interpret the purpose, operation and use of the same system. Search Scopus or Web of Science and ACM Digital Library using the topic query, then follow citations to the thesis start date. Record searches and compare methods, data, findings and limitations in literature-matrix.csv. Use that review to confirm or revise the gap and choose a current comparator. The linked papers are starting points. START HERE Tools: LibreOffice Writer/Calc, audio recorder with consent; no programming required. 1. Prepare a pilot with 2 examples from: 12 fictional merged records + 8 sessions. Write the task questions and a reference answer sheet. 2. Write a recruitment message, information sheet and consent form for the participants named above. Agree privacy handling with the supervisor before contact. 3. Pilot one session after approval; revise unclear questions, freeze the task sets and coding categories, then recruit the planned sample. Literature search: health records data provenance information quality trust qualitative scenario study Study controls: - Before collecting participant data, agree consent, storage and withdrawal handling with the supervisor. Use participant codes, not names, in study files. - Pilot separately, then freeze the questions and coding plan. Check objective answer keys independently; keep an audit trail of coding, including disagreements. - For comparisons, counterbalance order and case assignment; do not show a person both versions of one case. Report participant-level findings, not repeated tasks as independent people. - Show/hide source-only labels on matched records; never reveal fault labels. Keep measurement, unit and time information identical. Measure false alarms as well as identified problems. REQUIRES Recruit 8 health-information students. Bachelor: apply established methods and evaluate a practical solution or study. Master: position a research question in current scientific literature, investigate a mechanism or unresolved problem, and explain the contribution. Final scope is agreed with me. Study sizes are proposed; participant recruitment and planned materials are not already arranged.