Supervisors & feedback

What supervisors mean when they say the work is not ready

What the research on feedback and supervision shows, and a clinician-supervisor’s read on it.

There is a substantial literature on feedback, feedback literacy, supervisory relationships and doctoral attrition, and it is fairly consistent about what helps. This guide builds on that research, and uses my own time in both chairs, as a student and a supervisor, to illustrate it rather than to prove anything.

By Heidi Andersén, MD, PhD · Updated 18 July 2026 · 7 min read

“Do you ever feel like you’re a disappointment to your supervisor?” And, from another thread: “losing confidence due to my supervisor’s harsh criticism.” — two students, describing the same wound

What research showsFeedback shapes the next draft, and the confidence to write it.

Feedback can influence motivation long after it is written. Studies of feedback literacy and doctoral supervision consistently report that comments perceived as vague, personal or dismissive are less likely to improve the next draft, and more likely to reduce confidence, than comments that identify a concrete revision. The cost is not only emotional: ambiguous feedback often delays revision, because the student cannot translate the comment into an action.

Research on doctoral supervision and feedback repeatedly identifies emotional responses alongside technical ones. Students describe uncertainty, loss of confidence and reduced motivation when feedback is perceived as personal or insufficiently specific. The online discussions behind these guides echo the same themes, though they should not be treated as representative evidence.

The direction of the evidence is clear. Feedback that names an actionable next step tends to work; comments such as “be more critical” or “needs more depth” ask the student to infer the intended revision, whereas naming a specific analytical gap is more likely to improve the draft that follows.

✎Thirty seconds for me. Two weeks for them.

Why it happensExpertise, time, and an unequal desk.

Little of this is bad intent. Expertise compresses: a supervisor sees a whole diagnosis and writes its two-word summary, “more depth”, which a novice cannot unpack. Time competes with clarity, and a long delay interrupts the student’s writing momentum and raises uncertainty. And the supervisory relationship carries a real power imbalance, so the same words weigh more when they come from the person who signs off on progression and authorship. Vague or harsh feedback is usually the ordinary shape of expert-to-novice communication under time pressure and unequal stakes.

Common mistakesFour failure modes the evidence names

Critiquing the person, not the work. Research on effective feedback recommends directing comments at the manuscript, not the student’s ability or identity.
Delaying feedback. Long turnaround times interrupt writing momentum and increase uncertainty.
Shorthand a novice cannot decode. Experts compress complex judgements into phrases like “more depth” or “be more critical”; an inexperienced researcher cannot translate these into concrete revisions.
Ignoring the power imbalance. The same words carry different weight when written by someone who evaluates progression or authorship.

What supervisors can doSpecific, timely, actionable

The research points one way: make feedback specific, timely and actionable. In practice that is three habits. Pair every criticism with a concrete next revision, so “this is descriptive” becomes “this is descriptive, so add a sentence on what it means for your question”. Keep the comment on the work separate from any signal about the person, because a page-level problem stated without “and you are not the problem” is easily read as a self-level one. And protect the turnaround: a rough set of comments on time beats a polished memo the student has stopped waiting for.

Name the fix, or don’t send the comment.

Worked exampleThe same note, twice

Adapted from my own supervision practice

The version I might once have sent: “This literature review is superficial. Not PhD level. Be more critical.” The version I write now: “The review summarises each study well but does not yet weigh them against each other, and that is what pushes it to doctoral level. Pick the three most relevant studies, say where they disagree, and tell me which you find more convincing and why. The writing itself is clear; this is about the argument, not your ability.”

Same judgement, same bar. The second names the revision, keeps the person out of it, and a student can start it tomorrow.

What I take from itNot softer. More specific.

The evidence does not support making feedback softer. It supports making it more specific, timely and actionable. My own experience, as both a student and a supervisor, has led me to the same place: most difficult conversations get easier when every criticism is paired with a concrete next revision. I have come to treat a comment the way I treat any instrument in a study, and ask whether the person receiving it can actually use the reading. I am one supervisor in one field, and I still get this wrong. The direction, though, is not in doubt.

ChecklistBefore I send feedback

Four checks before it goes

Does every criticism name a concrete next revision?
Is the comment on the work kept separate from any signal about the person?
Is anything vague (“more depth”) unpacked into the specific gap I mean?
Is it on time, and could the student start work in the morning?

Related guidesRead next

There is no product on this page. Supervision is the hardest wound the students behind these guides describe, and the research says the fix is mostly specificity and timing, not softness. If you are on the receiving end, the practical companion is decode vague supervisor feedback.