"Non-Markovian" and "directional" errors inhibit scientific self-correction and can lead fields of study astray: an illustration using gardening and obesity-related outcomes

Jon Agley; Sarah E Deemer; David B Allison

doi:10.1186/s12874-025-02590-6

"Non-Markovian" and "directional" errors inhibit scientific self-correction and can lead fields of study astray: an illustration using gardening and obesity-related outcomes

BMC Med Res Methodol. 2025 May 24;25(1):137. doi: 10.1186/s12874-025-02590-6.

Authors

Jon Agley¹, Sarah E Deemer², David B Allison³

Affiliations

¹ Department of Applied Health Science, School of Public Health - Bloomington, Indiana University Bloomington, Bloomington, IN, USA.
² Department of Kinesiology, Health Promotion & Recreation, The University of North Texas, Denton, TX, USA.
³ Department of Epidemiology and Biostatistics, School of Public Health, Indiana University Bloomington, Bloomington, Bloomington, IN, USA. allison@iu.edu.

Abstract

Herein we coin the terms non-Markovian error and biased Markovian error to describe categories of scientific errors for which general progress within a field of study may be unlikely to result in scientific self-correction. We provide examples of such errors and show their impact on studies about gardening and obesity.

Keywords: Non-Markovian errors; Reproducibility; Rigor; Scientific errors; Scientific self-correction; Transparency.

Publication types

Letter

MeSH terms

Bias
Gardening* / statistics & numerical data
Humans
Obesity*
Research Design*