Assessment of Submitted Manuscripts in Medical Sciences According to Statistical Errors

Abstract

Objective: The aim of the present study was to list errors in statistical analysis and presentation detected during the first review of submitted manuscripts, compared to the literature reviewing published articles. Material and Methods: Statistical errors determined during the first review by the biostatistics reviewer in manuscripts included in the study were presented. All studies were assessed by the same biostatistics specialist. The defined statistical errors were classified by 4 biostatistics specialist respectively. Results: No statistical errors were found in 4.42% of 181 original manuscripts; 73.99% of 173 original manuscripts with statistical errors were accepted following the correction of the statistical errors. The most common statistical errors were errors related to test in 34.59%, p-values in 17.36%, summarizing data in 15.60% and incorrect and insufficient demonstration of descriptive statistics in 15.60%. Conclusion: Considering that the errors in scientific papers are reviewed and corrected by researchers according to the suggestions of the reviewer, researchers with lack of statistical solid background are found to make errors in the process of utilizing statistics in their studies. In some cases, although researchers get statistical counseling, they make errors in the presentation while writing the papers. In conclusion, it is of major importance that the submitted manuscripts be reviewed in terms of biostatistical analyses to prevent statistical and interpretation errors.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 13 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.