Can You Use “I” in a Thesis?

In most disciplines, yes — and in some, you are expected to. The rule that first person is banned from academic writing is one of the most persistent pieces of misinformation students carry from secondary school into postgraduate work, and following it produces exactly the contorted, agentless prose examiners complain about. APA 7 states the position bluntly: use first person to describe your own work, because the alternative is ambiguous about who did what.

What is true is that the rule is not universal, it changes by chapter, and the substitutes students reach for instead — the author, the researcher, we in a single-authored thesis — are usually worse than either option. This guide sets out what your style authority actually says, which chapters take first person, where the passive still earns its place, and how to check your own department’s convention in fifteen minutes.

What the style authorities actually say

Before you ask your supervisor, know what the manuals say, because most departmental guidance is downstream of one of them.

• APA 7 (psychology, education, health, much of the social sciences). First person is preferred when describing your own actions. The manual specifically discourages referring to yourself in the third person as the author or the researcher, on the grounds that it is ambiguous and can read as evasive. Use I for a single-authored thesis; we only when there genuinely are multiple authors.
• MLA and Chicago (literature, history, philosophy). No prohibition. First person is standard when you are staking an interpretive position, which is most of what a humanities thesis does.
• IEEE and numbered engineering styles. Mixed, and increasingly permissive, but the impersonal construction remains common: In this work, a method is proposed… Many engineering supervisors still expect it.
• Vancouver, AMA and most medical journals. Methods are conventionally written impersonally or in the passive; discussion sections increasingly use we.

Notice what none of them say: that first person is unscholarly. The concern behind the old rule was never grammar. It was that a personal pronoun invites personal opinion — and the actual fix for that is calibrating your claims, not deleting yourself from the sentence.

Objectivity is a property of evidence, not of grammar

This is the misunderstanding worth dismantling first. Removing I does not make a claim more objective; it only hides who is making it.

Compare “the data were interpreted as indicating a shift in practice” with “I interpret these data as indicating a shift in practice.” The second is more objective in the only sense that matters, because it is transparent about the fact that an interpretation is being made by a person who can be questioned about it. The first quietly presents a judgement as though it fell out of the data unaided.

Examiners read agentless prose as evasion precisely when the stakes are highest — around decisions you made and could have made differently. “It was decided to exclude twelve responses” invites the obvious question: decided by whom, on what basis? “I excluded twelve responses because they failed two attention checks” answers it before it is asked. Where you made a choice, owning it in the first person is the stronger position, not the weaker one. The related discipline of matching your claim strength to your evidence is covered in our guide to overclaiming, boosters and what your data licenses.

Where first person belongs, chapter by chapter

The convention is not uniform across the document. This is the pattern that holds across most empirical theses in the social and health sciences.

If you take one rule from the table, take this one: use first person for decisions and interpretations, and let the data or the procedure be the subject everywhere else. That single heuristic resolves the great majority of cases without consulting anything.

When the passive voice is the right choice

The passive has a bad reputation it only partly deserves. It is a legitimate tool with three clear jobs in academic prose, and deleting it everywhere is as much a style error as overusing it.

• When the actor is irrelevant or obvious. “Samples were stored at −80°C.” Nobody needs to know which pair of hands moved them.
• When the recipient is the topic. Placing the important thing first is what the passive is for: “Ethical approval was granted by the faculty committee” keeps approval in the subject position, where the sentence is about it.
• For cohesion between sentences. The passive lets you begin a sentence with information the previous sentence just introduced, which is what makes a paragraph read as connected rather than as a list.

Where it fails is when it conceals agency you should be claiming. “It was determined that the framework was inappropriate” hides a decision. “A modified version of the scale was used” hides who modified it and why. These are the passives worth hunting — not the ones in your procedure section.

There is also a purely mechanical cost. Passive and expletive constructions (“there is”, “it was found that”) add words without adding meaning, which matters when you are over your limit; the mechanics of that trade-off are covered in our guide to cutting thesis word count without losing substance.

The substitutes that are worse than either option

Students who believe I is forbidden reach for four replacements, and all four create problems.

“The author” and “the researcher.” APA discourages these explicitly. In a literature review the ambiguity is real — a reader meeting “the author argues” cannot immediately tell whether you mean yourself or the source you just cited. Where you need the distinction, write I and let the cited authors keep their names.

The royal we. “We interviewed thirty participants” in a single-authored thesis raises the question of who else was present. Some supervisors accept it as a disciplinary convention, particularly in mathematics and physics where we is a genuine inclusive convention addressing the reader. Outside those fields, check before using it.

“This researcher” or “the present author.” Every objection to the author, plus stiffness.

“One.” “One might argue that…” is register-mismatched in most contemporary academic English and reads as archaic. It also hedges the source of the argument out of existence.

If a hard prohibition genuinely applies in your department, the workable route is not a pronoun substitute but a change of subject: make the thesis, the chapter, the study, or the analysis the agent. “This chapter argues…”, “The study adopted a purposive sampling strategy…”, “The analysis proceeds in three stages…” These are impersonal without being agentless, and they read cleanly.

How the rule shifts by discipline

• Humanities. First person is normal and often expected. Interpretive claims are yours, and hiding that weakens them.
• Social sciences. First person for decisions and interpretation, following APA. Qualitative work in particular requires it, because reflexivity cannot be written impersonally — see our guide to writing a positionality statement.
• Laboratory sciences. Traditionally impersonal, but journals in many fields now accept and even prefer active first person in methods. Departmental practice lags journal practice here, so check locally.
• Engineering and computing. The most conservative in practice. Impersonal constructions remain common, though we is widely accepted in conference papers.
• Medicine and nursing. Impersonal methods, first-person plural discussion. Clinical reflective assignments are a separate genre and require first person throughout.

These are tendencies, not rules, which is why the fifteen-minute check below beats any general statement including this one.

The fifteen-minute check that settles it

Do not guess, and do not rely on a general guide. Establish the local convention directly.

1. Read your programme handbook and the thesis submission regulations. Some state a rule explicitly. Most do not, which is itself informative — an unstated rule is rarely enforced.
2. Download three recent theses from your own department through your institutional repository. Same department, same degree, passed within the last three years.
3. Search each one for “I ” and “we “. Note not just whether the pronoun appears but where — methodology only, or discussion too, or throughout.
4. Do the same with three recent articles from your target journal if you intend to publish from the thesis.
5. Ask your supervisor once, specifically. Not “can I use I?” but “I plan to use first person for methodological decisions and interpretation, and keep procedure in the passive — does that match what the department expects?” A specific proposal gets a usable answer; an open question gets a shrug.

Whatever you settle on, apply it consistently. Drifting between I, the researcher and we across chapters is more damaging than either convention applied throughout, and it is the version examiners actually flag.

Rewriting in practice

Four before-and-after pairs covering the cases that come up most.

The last row is the one to remember. A rewrite pass that converts every passive to active produces prose as mannered as the problem it set out to fix. The target is agency where agency matters, not active voice everywhere. The wider conventions governing the sentences around these choices — hedging, signposting and paragraph structure — are set out in our academic voice style guide, and the closely related question of which verb to use when introducing someone else’s work is covered in our guide to choosing the right reporting verb.

Write in the voice your department expects

Tesify drafts in the register you set — first person or impersonal, APA or MLA — and keeps it consistent across every chapter, so you are not fixing pronoun drift the week before submission.

Start writing your thesis with Tesify

Frequently Asked Questions