An admissions or scholarship committee reads hundreds of letters and is looking for one thing: comparative evidence. Not whether the student is good — everyone applying is good — but where they sit against everyone else this professor has taught. Three samples: a course instructor, a thesis supervisor, and a scholarship reference.
The weakest position a referee can be in — you taught them in a class of eighty. The letter works by being honest about that and precise about what was visible.
I taught Ahmed Nabeel in two courses — Data Structures in his second year, a cohort of 120, and Data Engineering in his third, a cohort of 50. He was subsequently my teaching assistant for Data Structures, which is where I came to know his work properly.
In Data Structures he placed third of 120 with a final mark of 91. In Data Engineering, which I designed and which is the more demanding of the two, he placed first of 50. Across eleven years of teaching I would put him in the top two or three per cent of undergraduates I have taught at this university.
What distinguishes him is not the marks. As my teaching assistant he wrote six practice problem sets which I now use across all sections — I did not ask him to, and they are better sequenced than the ones they replaced. He also maintains an open-source visualisation library for teaching data structures which has 140 stars on GitHub and which two colleagues have adopted for their own sections.
I should be clear about the limits of what I can say. I have not supervised him on extended independent research, so I cannot comment on how he would handle a long project without structure. On everything I have observed — technical ability, initiative, and the willingness to build things other people can use — he is exceptional.
I recommend him strongly and without qualification for admission to your programme.
The strongest possible referee. A supervisor has seen the student fail at something, which is the part a committee most wants to hear about.
I supervised Zainab Mirza's MS thesis from 2022 to 2024 and have since employed her as a research associate on an HEC-funded project. I have supervised thirty-one MS students in eighteen years and she is among the two best.
Her thesis examined fouling behaviour of ceramic membranes in textile effluent. She designed and ran a 200-day continuous filtration experiment, which is longer than any MS project I have supervised, and she ran it alone. Two months in, contamination invalidated the first dataset entirely. She identified the cause herself, rebuilt the rig, and restarted without being told to and without asking for an extension. That episode tells you more about her than her results do.
She has published two papers as first author, one in the Journal of Water Process Engineering, and has a third under review. She has also trained three junior students on the analytical procedures, and they now work independently.
On weaknesses, since a letter without them is not useful to you: she writes slowly and revises more than necessary, and I have had to push her to submit work she considered unfinished. This is a common failing in careful researchers and it improves with a supervisor who sets firm deadlines.
She has identified a genuine limitation in her own MS work — the inability to separate organic and colloidal fouling fractions — and wants a doctorate to address it. That is the right reason to do a PhD and it is rarer than you would think. I recommend her in the strongest terms.
A scholarship committee assesses merit and, often, circumstance. This letter addresses both without turning the student into a case study in hardship.
I have taught Rabia Nawaz physics for four years and been her form tutor for two. I write in support of her application for a full undergraduate scholarship.
Academically she is the strongest science student I have taught in eleven years at this school. She achieved four A* grades at O-Level including an A* in physics, in a cohort of 140 where nine students achieved four or more. In A-Level mock examinations she placed first in physics and mathematics.
She has done this while working. Her father died in her first year of A-Levels and she has since taught mathematics tuition to seven younger students, six hours a week, contributing to the household. Her attendance across that period was 96 per cent. I mention this not for sympathy but because the committee should know what her results were achieved alongside.
She also runs our physics club, which she started, and which has taken three students to national olympiad placings in two years.
Without funding she will not be able to continue to university, and she has told me so directly and without drama. I have no hesitation in recommending her as the strongest candidate I have put forward for this scholarship.
An admissions committee is not deciding whether a student is good. Everyone in the pile is good — that is why they applied. The committee is ranking, and a letter that does not help them rank has not helped at all.
So the most valuable sentence any academic referee can write is a comparison with a denominator.
The denominator is what makes it usable. "Excellent student" appears in every letter. "Second of thirty-one over eighteen years" appears in one, and the committee knows exactly what to do with it.
This feels counter-productive and is the opposite. Experienced committee members treat an entirely positive letter as either uninformed or perfunctory, and discount all of it.
A genuine, proportionate weakness does three things. It proves the referee knows the student well enough to have seen them struggle. It makes the praise credible by contrast. And it is frequently useful to the receiving supervisor, who has to work with this person.
Choose one that is real, specific and improvable. "Writes slowly and over-revises; needs firm deadlines" is honest and helpful. "Works too hard" is a non-weakness and reads as evasion, which is worse than saying nothing.
A referee who taught someone in a class of 120 cannot honestly comment on their independent research capability, and a committee knows it. Pretending otherwise damages the whole letter.
Stating the boundary explicitly — "I have not supervised him on extended independent research, so I cannot comment on how he would handle an unstructured project" — has the opposite effect to the one people fear. It tells the committee this referee distinguishes between what they have observed and what they assume, which makes everything they do claim more trustworthy.
It also, usefully, tells the student which referee to ask for the other letter.
All three samples are free to read and copy. Want them as editable Word files, or a letter drafted from your actual record for a supervisor to check and sign? Both take a minute.
Build mine with Hexa → Download all three · FreeThe people above are illustrative examples written for teaching purposes. Names, employers, registration numbers and figures are invented and should not be copied into a real application — use them as a shape to pour your own verified numbers into.