Language Use: Choosing Tense and Modality in a Literature Review Sentence
A candidate is revising a literature review paragraph. The original sentence reads: "Smith (2018) proves that early feedback improves writing, and this must be true for online courses too." The candidate's supervisor has flagged two problems: overclaiming and unwarranted generalisation. The candidate must produce a revised sentence that (a) reports Smith's finding accurately, (b) does not treat one study as proof, and (c) avoids extending the claim to online courses without evidence. The candidate also wants to keep a critical stance, since a later sentence will present a contradicting study by Ali (2021). She has three minutes and no access to Smith's original article.
1. Explain why "proves" and "must be true" are inappropriate here, and what the candidate can and cannot verify without the original article.
Read the worked answer
"Proves" attributes certainty that a single empirical study rarely delivers; findings are typically probabilistic and bounded by sample and context. "Must be true" adds a second layer of necessity, treating a logical deduction as if it followed from the data, when the extension to online courses is an empirical question Smith may never have tested. Without the original article, the candidate cannot verify Smith's sample, design, or effect size, so she should report only what she can attribute to the cited work. Safe reporting verbs are "reports," "found," or "suggests," depending on how strong the cited evidence is; since she cannot check, the more cautious "reports" is defensible. The generalisation to online courses should be flagged as untested rather than asserted.
2. Write the revised sentence and a follow-up sentence that sets up Ali (2021) critically, keeping the academic register.
Read the worked answer
Revised: "Smith (2018) reports that early feedback is associated with improved writing outcomes; whether this association holds in online courses remains untested." Follow-up: "Ali (2021), however, found no such association in a blended-learning setting, suggesting that the effect may depend on how feedback is delivered." The revision uses "reports" and "is associated with" instead of "proves" and "improves," which keeps the causal question open, and it converts the unwarranted extension into an explicit gap ("remains untested"). The follow-up uses "however" to signal contrast and "suggesting that" to interpret rather than declare, so Ali's null result is neither dismissed nor overgeneralised. Together the two sentences create a critical stance grounded in the cited evidence rather than in the candidate's assumptions.