Questions to Ask Grad Students in a Lab
Questions for a recruitment visit, a rotation, or a coffee with someone already in a lab you are considering. The students know things the principal investigator will not tell you and cannot tell you, and most of it only comes out if you ask a specific question.
20 questions, each with the reason to ask it · includes a conversation guide
The questions
Open any question to see why it works.
- 1
How did you end up in this lab, and did you look at others first?
An easy opener that quietly establishes whether people chose this lab or landed in it because a rotation fell through or funding appeared. A lab full of people who arrived by default runs differently from one people competed to join, and the students rarely volunteer which kind they are in.
- 2
What are you actually working on this week?
Ask for the week, not the project. The gap between the thesis title and the current week is where you learn whether people are running experiments, waiting on a reagent, rewriting a rejected paper, or troubleshooting an instrument that has been down since spring.
- 3
How often do you meet your advisor one on one, and how often does it get cancelled?
The scheduled frequency is on the lab website. The cancellation rate is the real number, and it tends to be highest in labs run by someone with a large grant portfolio or an administrative role. Ask what happens when a meeting is cancelled: rescheduled that week, or gone.
- 4
When you send a draft, how long until you get it back?
This single number predicts your time to degree better than almost anything else, because a paper that sits for three months twice has cost you half a year. Ask about the longest wait rather than the average, and ask whether the comments come back line by line or as a rewrite.
- 5
How does someone here decide what to work on: assigned, chosen, or negotiated?
Assigned projects get you to a defence faster and leave you without a research identity. Fully self directed projects can strand you for two years. What you want to hear is a process, including what happened the last time a student wanted to change direction.
- 6
Who taught you the techniques you use?
If the answer is a senior student or a staff scientist, the lab has a working training layer and you will be productive within months. If the answer is a protocol on a shared drive and a lot of failed attempts, budget a year. Also ask whether that person is still here.
- 7
How is authorship decided, and has anyone been surprised by the final list?
Most labs have an unwritten convention and no policy, which is fine until a project is split or someone leaves mid way. The question to actually ask is what happened when someone did leave, since that is when middle authorship and first authorship get quietly reassigned.
- 8
What happens here when an experiment does not work for six months?
Every project has this stretch. The answer tells you whether the response is more resources and a change of approach, or increased pressure and a suggestion to work weekends. Ask for a specific example rather than a general description of the advisor's temperament.
- 9
How is your stipend funded, and does that change year to year?
Grant money, a training grant, a fellowship you have to win, or teaching are four different lives. Ask when the current grant ends. Students often do not know, and the ones who do will tell you whether the lab has been anxious about renewals, which shapes everything from equipment to mood.
- 10
How much of your week goes to lab service: ordering, animal colony, equipment, undergraduates?
This is real work that nobody counts and it is distributed unevenly, often landing on whoever is most conscientious or most junior. Ten hours a week of colony management for four years is a chunk of a dissertation. Ask who does it now and how that was decided.
- 11
How long do people take to finish here, and has anyone left without finishing?
Programme averages are published; lab averages are not, and they can differ by years. The attrition question is the one that matters and it is uncomfortable, so ask it plainly and let the silence sit. Where people went afterwards tells you whether leaving was a rescue or a collapse.
- 12
Where did the last three people to graduate go?
Three specific names and destinations, not a claim that graduates do well. You are checking whether the advisor's network still places people, whether the placements match what you want, and whether the most recent one was two years ago or seven.
- 13
Does the advisor still read your writing closely, or mostly sign off?
Some senior investigators have stopped doing line edits and delegate it to postdocs, which can be good training or can mean nobody senior has read your thesis chapter. Ask who taught them to write, since in a healthy lab someone did.
- 14
What happens when someone makes a real mistake: a contaminated line, a lost dataset, a failed order?
This is the most predictive question about a supervisor's character that a student can answer safely, because it is about an event rather than a personality. Note whether the story they tell is their own or someone else's, and whether they laugh or lower their voice.
- 15
Can you take a holiday, and did you take one last year?
Every lab says yes to the first half. The second half produces the truth, along with the informal rules: whether you ask or tell, whether experiments are ever paused, and whether the person who took two weeks off heard about it afterwards.
- 16
How does this lab get on with the rest of the department?
Isolated labs limit your committee options, your collaborations and your recommendation letters. Ask whether they attend other groups' meetings, whether anyone co-supervises, and whether there is a dispute with a neighbouring group that a new student would inherit without knowing it.
- 17
What would you ask if you were sitting where I am, that you did not think to ask?
Hands the conversation to them and often produces the most useful minute of the visit. People who have been in a lab for four years know exactly which question they wish they had asked, and it is usually about funding length, the advisor's travel, or an interpersonal situation nobody mentions on a visit day.
- 18
Is there anyone here I should know about before I join?
Deliberately open, and phrased so the answer can be positive. A helpful senior student is worth knowing about; so is a lab member who takes credit or blocks equipment. Watch the pause before the answer more than the answer itself, especially if others are in earshot.
- 19
Would you join this lab again?
The blunt version of every other question, and it should come near the end when they have already said a lot. Qualified yes is normal and honest. An immediate enthusiastic yes from every person in a lab is worth being slightly suspicious of, since no lab is right for everyone in it.
- 20
Who else should I talk to, including someone who has left?
Former members are the most candid source available and there is usually no reason a current student cannot pass on a name. If nobody will give you one, that itself is information. Email the alumnus afterwards rather than during the visit, and ask them the same short list.
Getting real answers out of a lab visit
Practical guidance for the conversation itself.
Where the honest version happens
Where the honest version happens
- Never ask the hard questions with the principal investigator in the room, or in a group tour where everyone can hear who answered.
- Walking somewhere is better than sitting somewhere. The route to lunch, the walk to another building, and the corridor after a meeting are where students say the thing they were choosing not to say.
- Ask the same three questions of three different people. Consistent answers are usually true; a wide spread usually means the experience depends on who you are in this lab.
- Email a follow up question a week later. People answer more candidly in writing after the performance of a visit day is over, and they have had time to decide they liked you.
- Talk to the most junior person and the most senior person. First year students know what recruitment promised; fifth years know what it turned into.
Translating the standard answers
Translating the standard answers
He is very hands off
Sometimes this means genuine independence with support when you ask. More often it means you will not get feedback on a draft for months and you will need a second mentor. Test it by asking how quickly he responds when a student asks for help, rather than how often he checks in.
She is intense, but you learn a lot
Usually accurate on both halves. Follow up by asking what intense looks like on a bad week, and whether anyone has been shouted at in front of others. People will answer that specific question when they would not criticise a supervisor in general terms.
We are like a family here
Can mean a genuinely supportive group. Can also mean the boundaries between work and life have dissolved, that weekends are assumed, and that leaving is treated as disloyalty. Ask what happens when someone says no to something, since families with no exit are the ones this phrase should worry you about.
Funding is fine
Ask for the end date of the grant that pays the stipend. Students often do not know it, and finding out together is a reasonable thing to do on a visit. A lab with a renewal pending in your second year is a different offer from one with five years secured.
If this is a rotation rather than a visit
If this is a rotation rather than a visit
A rotation gives you access no visit day can match, so use it for the things you cannot ask about directly. Notice who eats lunch together and who eats alone. Notice whether people work in the lab or from home, and whether the reason is preference or avoidance. Ask the same question in week one and week ten and see whether the answer changes once you are treated as an insider. Sit in on a lab meeting and watch how a student is treated when their data disappoints, which is the only reliable preview of how you will be treated when yours does. Before the rotation ends, ask the advisor directly whether they intend to take a student next year and how many, because a warm rotation followed by no slot is the most common bad outcome and it is avoidable with one question.