[Intentional and Non-Intentional Distortion of Research Findings]

Polim Med. 2015 Jan-Jun;45(1):7-9.
[Article in Polish]

Abstract

The flood of scientific information increases the risk of research findings distortion. The authors brought attention to the possible pitfalls of statistical inference in biomedical research. Due to intentional or non-intentional errors in study design and size, data collection and inappropriate statistical analysis some of research findings are false or exaggerated. Researchers excessively motivated to publish all possible study results sometimes seek any possible way to achieve p-value less than 0.05. False-positive study results can be obtained in underpowered studies, excessive flexibility in study designs and participants recruitment, probing different control groups, creating special subgroups or changing statistical tests.

Publication types

  • Editorial
  • English Abstract

MeSH terms

  • Data Interpretation, Statistical*
  • False Positive Reactions
  • Observer Variation
  • Periodicals as Topic / standards*
  • Publication Bias*
  • Research Design / standards*