The Hidden Curriculum
The operating knowledge research training assumes you already have — named, written down, and made checkable.
What “hidden curriculum” means here
Every research field runs on two bodies of knowledge.
The first is written down: the biology, the imaging physics, the segmentation metrics, the statistics. It is what the technical units and the modules teach.
The second is not written down anywhere: which version number belongs in a figure caption, what “have a look at this” means when a supervisor says it, how long you should be stuck before you ask, how much a published result is worth arguing with, whose name goes on the paper and in what order, and what to do at 11pm when you realise the number you gave a collaborator last month was wrong.
That second body is the hidden curriculum. It is operating knowledge distributed unequally by background. Some people arrive already knowing how authorship works because someone at their dinner table did. Others reconstruct it over five years from overheard fragments and pay for the reconstruction in wasted effort, mistimed asks, and credit they never claimed.
Why this is a fairness problem and not an etiquette guide
The site’s education models page puts it directly: these are not soft skills appended to technical training. They are the operating knowledge that determines whether technical training converts into a career, and making them explicit is a fairness intervention as much as a training one.
Three things follow, and each changes what you should do:
- Norms have to be stated, not modelled. Watching a competent person work transmits a norm only to someone who already knows what to look for. Everyone else sees a person typing. So these pages state norms as sentences you could put on a checklist, not as habits to absorb.
- The cost of not knowing a norm is measurable, not social. It is a result you cannot defend in review, an authorship conversation that happened without you, six months of analysis pinned to nothing. Treat a missing norm as a defect, not as an awkwardness.
- “Ask if you’re unsure” does not fix it. You cannot ask about a norm whose existence you do not suspect. This is the reason the pages below are written as lists of named norms rather than as an invitation to enquire.
None of this means the norms are all good. Some are arbitrary, some are unfair, and a few exist because they were convenient for whoever was already in the room. Naming a norm is not endorsing it. But you cannot decide whether to comply with, push against, or ignore a rule you cannot see.
The norms are already taught on this site, usually without a name
The technical track is full of professional norms presented as technique. Read these five and notice that not one of them is labelled as a norm:
- Unit 01: “Any Bin B claim must name its assumption in the same sentence, or in the figure caption. ‘Putatively excitatory (asymmetric morphology)’ costs four words.” — a norm about how much credit a reader will extend you.
- Unit 04: “Every figure you produce records the materialization version, the query code, and the date.” — a norm about what a reviewer is entitled to check.
- Unit 05: “Negative calls are data. An annotator who never says ‘no’ is not calibrated.” — a norm about what confidence is for.
- Unit 08: “A stated 6% exclusion rate is honest and cheap; an unbudgeted tail is neither.” — a norm about disclosure.
- Unit 09: “Report how many tests you ran, including the ones you ran and did not report.” — an ethical norm with a statistical justification bolted on.
A reader who already speaks the professional register hears all five as statements about professional conduct and generalises them to the next paper, the next field, the next meeting. A reader who does not hears five pieces of connectomics-specific advice and transfers none of them. That gap is the whole reason this section exists. Each technical unit now closes with a short “the norm behind this unit” section naming its own one or two. Technical practice is where all twenty-six are collected, named, and paired with what their absence looks like in real work.
The seven pages
| Page | What it covers | Who most needs it |
|---|---|---|
| Technical practice | Twenty-six reporting and disclosure norms careful practitioners follow silently: versions, assumptions, thresholds, exclusions, negative results, and where to spend effort | Anyone about to produce a figure someone else will build on — especially people who learned analysis from tutorials rather than from a supervisor’s red pen |
| Reading and judging | How to read a paper in the order experts actually read it, read the methods for what is absent, and tell a solid result from a fragile one | Anyone who has been told to “read the paper” and does not know what would count as having done it |
| Lab norms | Asking for help, what a PI expects but will not say, how meetings work, escalation, authorship, and what to do after a mistake that affects someone else | Anyone in their first weeks in a lab, and anyone whose family did not contain an academic |
| Career mechanics | How the machinery actually works — applications, funding, references, transitions | Anyone at or approaching a transition point, who is being told to “network” without being told what that means operationally |
| Meta-learning | How to learn deliberately: what to practise, how to tell improvement from familiarity, how to run your own calibration | Anyone working hard and not getting better, which is a specific and fixable condition |
| Belonging | Identity, impostor feelings, and the difference between “I am struggling” and “this environment is hard for people like me” | Anyone deciding whether to stay, and anyone mentoring someone who is |
| Conflict | Disagreement, credit disputes, difficult supervisors, and what recourse actually exists | Anyone currently in one — and best read some months before you need it |
How to use this if you are reading alone
Do not read all seven. Read the one matching the thing that is currently going wrong, apply one norm from it this week, and come back.
A workable order if nothing is currently going wrong:
- Technical practice — because it is the one you can apply tomorrow, on work you already have.
- Reading and judging — because it makes every subsequent hour of reading worth more.
- Lab norms — before your next meeting, not after.
- The rest when the situation arrives.
Turn each page into an artifact rather than a read. After Technical practice, open your most recent notebook and add the header block. After Reading and judging, take one paper from the Journal Club list and write the four fragility questions with answers. After Lab norms, write the three scripts in your own words, which is exactly the exercise in Module 02. Reading these pages and changing nothing is the failure mode they are most vulnerable to.
How to use this if you are facilitating
The Facilitator Guide already names hidden curriculum as the predictable friction for the first-generation learner persona, and gives the standing instruction: state norms explicitly, in the first five minutes, every time. These pages are what you state.
Practical moves:
- Do not assign this section as optional reading. “Optional” reads as “for people who are struggling” to exactly the learner who needs it, and they will not open it. Put one page in the required sequence and give it session time.
- Run a norm as a 15-minute segment. Take one named norm from Technical practice, show one real figure that follows it and one that does not, and ask the cohort what a reviewer can and cannot check in each case. This works better than presenting the list.
- Attach the norms to the technical units you are already teaching, at the point where the unit teaches the technique. The norm and the technique are the same lesson; splitting them into a separate “professional skills” slot is what made the curriculum hidden in the first place.
- Say what you did not know. A facilitator narrating the norm they got wrong in their second year does more than the list does. This is the same principle as the guide’s instruction to model uncertainty in public rather than working clean examples.
- Grade the norm. If your rubric does not have a row for “version pinned, assumptions named, exclusions reported”, learners will correctly infer the norm is decorative. Module 21 already carries the reproducibility version of this rubric row.
Where this connects
- Module 02: Research Foundations and the Hidden Curriculum — the taught version, with a 90-minute run-of-show and a studio activity producing a personal navigation playbook.
- Module 21: Reproducibility and FAIR Principles — its “hidden curriculum scaffold” section is the model these pages follow: what senior reviewers expect but rarely state.
- Education models — MERIT stage 2 is where this material belongs in a mentoring relationship, and the Professional Pathways table lists the questions each page here answers.
- Facilitator Guide — session design, differentiation, and the failure modes of teaching this material.
- Learner personas — who these norms are invisible to, described concretely.
- Career and Community track — the track this section sits in.