EC-17.5 The Portfolio as Evidence

What this is and why it exists

This topic is about the evidence a degree should leave behind, which is a different question from getting hired. The Interview preparation area covers that, and the two are cross-linked rather than merged.

The argument here is narrower and simpler. Unrecorded work is indistinguishable from work never done.

And the recording is cheap, if it is done as you go and expensive if it is not.

The vocabulary

• Portfolio — the assembled evidence of what you can do.
• Artefact — a concrete output: code, a board, a measurement set, a written report.
• Decision record — a note of what you chose and why.
• Attribution — stating which parts were yours and what you built on.
• Curation — choosing what to include and what to remove.

The mental model

What belongs in a portfolio is anything with a result somebody else could judge.

Projects that reached an outcome. Code that runs. Measurements with their method. Writing that explains a decision.

What does not belong is coursework that everyone in the year submitted in the same form. It demonstrates attendance rather than ability, and a reader who has seen the same submission forty times learns nothing from the forty-first.

Writing up a project means making the decisions visible, not describing the finished thing.

What the problem was. What you chose, and why. What went wrong, and what you did about it.

A reader learns far more from the decisions than from a description of the result. The decisions are what would transfer to a different problem. This is also why the design module insisted on recording them: the record you kept then is the write-up now, almost unchanged.

Evidence has to be verifiable. A repository someone can open. A measurement with its conditions. A photograph of the actual build rather than of a similar one.

An assertion that something worked, with nothing attached, carries almost no weight. A portfolio made entirely of such assertions carries less than none, because the pattern is visible.

Attribution within your own work is the failure mode that damages people most.

Say clearly which parts were yours in a team project. Cite the libraries, reference designs and tutorials you built on. Overstating a contribution is the most common way a portfolio harms its author. The overstatement is usually discovered in conversation with someone who knows the field.

Understating is a smaller error and a recoverable one. Overstating is not.

Curation means removing things. Three strong pieces read better than nine mixed ones, because a reader judges by the weakest item they see rather than the best.

Removing early work as later work replaces it is maintenance rather than modesty. A first-year project alongside a capstone invites a comparison that helps nobody.

Finally, keeping it current. Twenty minutes at the end of each project, while the details are still in your head.

Assembling a portfolio from four years of memory in a single week is the version everyone regrets attempting. The artefacts are scattered, the measurements are gone, and the decisions have been forgotten. It is the same argument the laboratory topic made about records, at the scale of a degree.

What you should now be able to explain or do

• Decide what belongs in a portfolio and why identical coursework does not.
• Write up a project so that the decisions are visible rather than the outcome alone.
• Provide evidence a reader can verify rather than an assertion.
• Attribute your own contribution and what you built on, accurately.
• Curate a portfolio down to its strongest pieces.
• Maintain it with a small effort at the end of each project.

Check yourself