Guides · Running a lab

A new student's first fortnight

What to set up before they arrive, what to teach in week one, and the habits that are almost impossible to introduce later. Two weeks of attention that decides the next four years.

Reading time
9 min
Licence
Free to copy into your own lab documentation. No signup.

A new student's first two weeks are usually spent waiting — for an account, for a desk, for someone to have time. Then they are handed a project and quietly absorb whatever habits are nearest to hand, which are usually the habits of whoever sits beside them.

This matters more than it looks. The way someone organises their work in the first month is the way they will organise it for the whole degree. Nobody restructures a working system in year three. Two weeks of attention now is worth more than any amount of correction later.

Who this is for

A PI or lab manager expecting a new PhD student, MSc project student or postdoc. Also useful to hand to the student directly. Nothing here assumes particular software or a budget.

Before they arrive

Nearly everything that wastes a new arrival's first week is administrative and could have been done in advance. A week is four per cent of a two-year MSc.

The week before they start 0 / 6
✓Request the institutional account and network accessThe longest lead time in the whole process, and nobody starts it early.
✓Decide where their data will live before they generate anyNot their laptop. Set up the folder and tell them the path on day one.
✓Book their safety inductionIt gates bench access, and the slots fill. Everything else waits on it.
✓Name who they ask when you are unavailableOne person, told in advance that they are that person. Otherwise the answer is "nobody".
✓Pick a first experiment that will workSomething the lab does routinely. Confidence first; novelty later.
✓Put the lab's conventions somewhere they can read aloneNaming, storage, instrument booking. One page. Being told once, verbally, is not being told.

Week one: the tour that is not a tour

Showing someone around a lab teaches them almost nothing. What works is a small, complete, real task — ideally one that ends in data.

Have them run something routine end to end, with someone beside them: prepare it, run it on the instrument, get the file off, put it where it belongs, and write down what they did. It does not matter if the result is uninteresting. What matters is that they have completed the whole loop once, including the parts after the measurement, which is precisely the part that is never taught.

Most students are taught how to run an experiment and left to invent for themselves what happens to the file afterwards. Then we are surprised that everyone invents something different.

The four habits worth insisting on

Only four. A long list will be ignored, and these are the ones that are genuinely hard to introduce later.

1. Data goes to the shared place the same day

Not at the end of the week, and never only on a personal laptop. This is the habit that most often fails to form, and it is the one that costs a lab entire projects when someone leaves.

2. Write down why, not just what

“Ran plate 3” is a log. “Ran plate 3 to check whether the effect holds at lower concentration” is a record. The second takes four extra seconds and is the difference between notes that are useful in year three and notes that are not.

3. Record the failures, with a line on why

New students hide failed experiments, reasonably, because they read them as personal failure. Say explicitly on day one that abandoned work with a one-line reason is valuable to the lab — and that repeating someone's silent dead end is the waste, not the dead end itself.

4. Follow the naming convention from the first file

Trivial to start, painful to retrofit. The first file they save sets the pattern for every file after it.

Why we care about this

Habit one is the hardest to sustain by discipline alone, which is why woodle.cloud watches the folder the instrument already writes to rather than asking anyone to remember. But the four habits are worth teaching whatever you use, and three of them are about writing, not software.

See how it works →

Week two: let them find the gaps

In the second week, give them a protocol and let them run it without hovering — then ask what they had to guess.

A new person is the only reliable instrument a lab has for finding undocumented assumptions. Every question they ask is a defect in your documentation, and this is the only fortnight in which they will ask them out loud. Six months in, they will have absorbed the tacit knowledge and stopped noticing that it is tacit.

Write the questions down as they come. That list is a better to-do list for lab documentation than anything a PI produces from memory.

What to say about mistakes, on day one

Say it explicitly and early, because the default assumption is the opposite: a contaminated culture, a dropped plate or a misread protocol is normal and should be reported immediately, and the only real error is the unreported one that quietly enters the data.

Labs where this is not said out loud in the first week get students who hide small failures, and small hidden failures are how bad numbers reach papers.

The thirty-day check

Put a calendar reminder for a month out. Twenty minutes, four questions:

  • Show me where your data from last week is. (Tests habit one, honestly.)
  • What did you not understand that you did not want to ask about?
  • What has taken much longer than you expected?
  • What would you tell someone starting next week?

The last question is the useful one. A month in, they can still see the lab from outside — and that view is gone for good by month three.

Take it with you

The pre-arrival checklist, the four habits and the thirty-day questions are free to copy into your own lab documentation. Or hand this page to the student directly. No attribution needed and no signup.