What Previum returns on a published sports science paper: every critique pinned to its sentence, p values recomputed, figures that do not add up and references checked against eight databases.
Sample report · sports science · real, already published paper
A paper published in a peer-reviewed journal, run through the Basic report exactly as it is, nothing edited: the manuscript with every critique pinned to the sentence that triggers it, and the code checks underneath.
The paper has already been peer reviewed and published. It is reproduced here under its Creative Commons Attribution 4.0 license (the full attribution, authors included, is at the linked DOI; they are deliberately not named here) to show what Previum returns on a real text from your field. The objections are the ones one more reviewer could have raised; that a published paper receives them says how demanding the tool is, not how weak the paper is.
Comparison of resistance training using barbell half squats and trap bar deadlifts on maximal strength, power performance, and lean mass in recreationally active females: an eight-week randomised trial. BMC Sports Science, Medicine and Rehabilitation, 2024. DOI: 10.1186/s13102-024-00911-8. License: CC BY 4.0.
The manuscript presents a well-intentioned randomized comparison of two popular lower-body exercises in an under-studied population. The two most consequential problems are the very small sample (n=22) without power justification, producing wide confidence intervals that cross zero, and language in the abstract and conclusions that claims statistically supported efficacy and practical equivalence beyond what the non-significant between-group tests actually demonstrate. Minor issues include lack of training verification and ambiguous table reporting. These problems are correctable with textual revisions and an expanded limitations paragraph; after those changes the work would be suitable for a mid-tier sports-science journal.
Quoted sentence: «Limitations of our sample size likely contributed to the somewhat imprecise estimates in our results. Given our somewhat limited sample, this study likely had a low statistical power to detect small and medium effect sizes which increases the type II error probability.»
Why it matters: Without a power calculation or sensitivity analysis, the non-significant between-group results cannot be interpreted as evidence of equivalence; reviewers will question whether the absence of difference reflects true similarity or insufficient statistical power.
What would fix it: Add a post-hoc sensitivity analysis (e.g., using the 'pwr' package or G*Power) stating the minimum detectable effect size given n=22, two groups, α=0.05 and 80% power for an ANCOVA on change scores, and discuss this limit explicitly in the limitations section.
Quoted sentence: «The present study suggests that twice-weekly RT sessions with high external loads (≥ 85% 1RM), using either half squats or trap bar deadlifts, can significantly increase lower-body maximal strength and enhance power performance in recreationally active females over a period of eight weeks.»
Why it matters: Overstatement of findings in the abstract and conclusions is a common reason for revision requests or rejection at the first editorial screen; the current wording risks misleading readers about the strength of evidence for the practical implications listed.
What would fix it: Revise the abstract conclusions and final paragraph to state only what the data show: 'no statistically significant between-group differences were detected' and 'both exercises were associated with within-group improvements whose magnitude cannot be statistically tested against baseline in the present analysis'. Remove the word 'significantly' from efficacy claims.
In total the reviewer found 7 major and 9 minor issues; the Basic report shows the two that decide.
With the same engine the tool uses, on 11 September 2026: the full text of the paper, tables and references included, went into the Basic report with no added instructions. The checks below are arithmetic on the text and queries to eight bibliographic databases, no model involved. Not a line of the output has been edited.
Upload your manuscript (Word or PDF) and in a few minutes you have this same report on your text. Free, with your account. The manuscript is not stored.