Package reproducible evidence and describe your contribution

Understand the method

A portfolio reviewer should be able to move from the headline to the method and source without guessing which file is current. Package the project around that path. Include the unchanged source, prepared data, correction evidence, metric contract, editable calculations, dashboard, written note and test record. A concise README explains how these pieces relate and which file to open first.

Preserve the reproducible route

For SQLite, keep the setup script, your completed SELECT queries and expected/observed results. The setup belongs in a new empty local practice database, never a hosting application database. For a spreadsheet, keep editable formulas and clearly named source tables rather than exporting every intermediate view as an image. A reviewer may prefer a quick case study, but the reproducible route should remain available.

The supplied complete pack keeps linked files together: the worked case study can open its dashboard reference, source files and SQL scripts offline when the ZIP is extracted without changing the relative paths. Downloading only one HTML file does not bring along its linked evidence. This differs from the dashboard reference, which embeds its fictional data and can operate alone. Explain that distinction in your own folder guide.

Describe contribution precisely

Use a label such as “Guided learning project, fictional Northstar dataset supplied by DataCareerHub.” State what you actually performed: source inspection, key-based preparation, metric implementation, dashboard design, interaction tests and explanation. Identify any provided query, template or reference that you reused. A submitted file is not evidence that you authored every line inside it.

Avoid invented client names, employment roles, review endorsements, completion certificates or commercial outcomes. The case does not demonstrate that you increased a company’s revenue or saved operating costs. You can make a concrete and accurate claim about your own work: for example, identifying how changing channel weights and a sparse baseline qualify an aggregate increase.

Inspect the sharing copy

Review the actual folder or archive you intend to share. Open its links, confirm its source version and remove only your own irrelevant working copies from that sharing copy. Check for credentials, personal contact information or unrelated real data before publishing. This exercise uses fictional records and requires no personal information, so no private data is needed in the submission.

Make relative links work from the extracted folder and identify any tool dependency. Do not claim that a hosted preview is deployed if you only produced a local file. Keep a short note on what remains untested, such as an external reviewer’s environment or a full accessibility evaluation. A checked local example is a useful deliverable without exaggerated readiness claims.

Prepare a review handover

Your guide should name the question, source, method, main finding, limitations, how to reproduce and what you contributed. Include a brief presentation outline. The course does not automatically receive or publish your files. Keep the project locally until you choose where and how to share it, and use the final lesson’s independent extension to make the work more distinctly your own.