Active Voice vs. Passive Voice

Whether to write a sentence in active or passive voice is not a grammar error either way — it is a stylistic choice, and in scientific and academic writing it is governed by disciplinary convention as much as by clarity. This term explains what actually distinguishes the two in a research-writing context, and when current journal and style guidance recommends each.

The distinction that matters for manuscripts

Active voice makes the sentence’s grammatical subject the agent performing the action: “We measured cortisol levels at four time points.” Passive voice makes the subject the recipient of the action, with the agent optional or omitted: “Cortisol levels were measured at four time points.” Both sentences are equally grammatical and equally precise about the procedure — the difference is who or what the sentence foregrounds, not correctness.

Why this is a live debate in scientific writing, not a settled rule

For most of the twentieth century, scientific convention leaned heavily passive, on the reasoning that removing the researcher as grammatical subject kept the writing focused on the phenomenon and procedure rather than the people doing the work — a stand-in for objectivity. That convention is still visible in most published Methods sections. Since roughly the 1990s, style authorities and a growing number of journals have pushed the other way, arguing that overused passive voice adds words, hides who did what, and can obscure whether a claim is the authors’ own finding or something asserted from prior literature. The result is a genuine, still-active split by section and by publisher, not a single universal rule.

What current style and journal guidance actually says

APA Style permits both but is explicit that writers should “use the active voice as much as possible to create direct, clear, and concise sentences,” reserving passive voice for when the sentence genuinely needs to focus on the recipient of an action rather than the actor. Author guidelines at high-impact journals, including Nature-family and Science-family titles, generally echo this: active voice is the recommended default for readability, particularly outside the Methods section. At the same time, those same journals’ published Methods sections still run heavily passive in practice, because many editors and reviewers continue to treat a depersonalized, procedure-first voice as the expected register for describing what was done to a sample, instrument, or dataset.

Section-by-section practical guidance

• Introduction, Discussion, Conclusion: active voice is now close to a universal recommendation. “This study demonstrates that X increases Y” is more direct than the passive “It is demonstrated by this study that X increases Y,” and far clearer than the agentless “It was found that X increases Y,” which leaves a reader guessing whether the finding is the authors’ own result or a claim drawn from someone else’s prior work.
• Methods: passive remains the traditional and still-common choice — “Blood samples were collected at 0, 30, and 60 minutes” — though an increasing number of journals now accept or even prefer the active equivalent, “We collected blood samples at 0, 30, and 60 minutes.” Where a journal’s author instructions are silent, following the convention used in that journal’s recently published articles is the safer default.
• Results: mixed in practice; active voice is generally clearer when attributing an analytical decision (“We excluded three outliers”) and passive is often fine for stating an outcome where the actor is irrelevant (“A significant difference was observed between groups”).

When passive voice is still the better choice

Even under an active-voice-preferring house style, passive voice remains the right call whenever the recipient of the action is genuinely what the reader needs to track, and naming the actor adds nothing. For example, “The gene was knocked out in three independent cell lines” keeps the reader’s attention on the procedure and the object of study; rewriting it as “We knocked out the gene in three independent cell lines” doesn’t add information, it just re-centers the researchers where the object under study is what matters. The test is not “is this passive,” it’s “does naming the actor here help the reader.”

Related CASRAI content

For the companion grammatical-convention question in the same manuscript, see Verb Tense Conventions in Research Papers. For other manuscript-drafting mechanics, see Manuscript Word Count Limits and How to Shorten a Submission, Justifying Sample Size in a Manuscript, and Research Abbreviations: When and How to Define Them in a Manuscript. For the broader context this term sits in, see the Scholarly Writing for Researchers pillar.