GIS Professional Practice, Ethics, and Project Workflow
This domain tests how a GIS professional runs a project and behaves: requirements and lifecycle, documentation and reproducibility, privacy and licensing, and the ethical obligations attached to location data about people.
Project lifecycle
A defensible workflow runs: needs assessment and stakeholder interviews, requirements definition, data discovery and quality assessment, design (data model, coordinate system, symbology), pilot, implementation, QA/QC, delivery and maintenance. Requirements come before data acquisition; choosing software before understanding the question is a classic failure. Maintenance and handover are part of scope, not an afterthought.
Documentation and reproducibility
Every deliverable should be reproducible: documented sources, processing steps, parameters, software versions and coordinate systems. ModelBuilder or scripted workflows (Python) provide an auditable record that manual geoprocessing does not.
Ethics and privacy
Professional codes (for example the GISCI code of ethics) emphasise integrity, competence, confidentiality, avoidance of conflicts of interest and respect for privacy. Individual-level location data is personal data: home addresses, patient locations and tracking traces can re-identify people even after names are removed. Standard mitigations are aggregation to a minimum unit, suppression of small counts, geomasks or coordinate jitter, and data-use agreements. Aggregation is not automatic anonymity - small areas and unique travel patterns allow re-identification.
Licensing and liability
Licences govern redistribution, derivative works and attribution; open licences are not licence-free. Maps inform decisions, so error disclosure and fitness-for-use statements limit liability and support the user.
Worked example
A public health team asks for a map of individual addresses of people with a stigmatised condition. Reasoning: the request is technically trivial but ethically and legally fraught. Identify the minimum data needed - likely counts by health district, not points. Propose aggregation to areas meeting a minimum population threshold, suppress counts below a small-number threshold, confirm the legal basis and data-sharing agreement, and document the decision and its approver. Delivering the point layer would breach confidentiality and create re-identification risk.
Traps that cost marks
- Assuming de-identified points are anonymous when geography itself identifies.
- Ignoring licence terms that prohibit redistribution or commercial use.
- Treating documentation as optional; without it the analysis cannot be defended or reproduced.