Common errors in statistics and methods

Peter Flom; Katie Harron; Javier Ballesteros; Chester Kalinda; Eirini Koutoumanou; Jeremy Miles; Sarah Jane Nevitt; Peter Rohloff

doi:10.1136/bmjpo-2024-002755

Common errors in statistics and methods

Peter Flom, Katie Harron, Javier Ballesteros, Chester Kalinda, Eirini Koutoumanou, Jeremy Miles, Sarah Jane Nevitt, Peter Rohloff

Centre for Reviews and Dissemination

Research output: Contribution to journal › Article › peer-review

Abstract

As statistical reviewers and editors for BMJ Paediatrics Open (BMJPO), we frequently see methodological and statistical errors in articles submitted to our journal. To make a list of these common errors and propose suitable corrections, and inspired by similar efforts at other leading journals, we surveyed the statistical reviewers and editors at BMJPO to collect their ‘pet peeves’ and examples of best practices.(1, 2) We have divided these into seven sections: graphics; statistical significance and related issues; presentation, vocabulary, textual and tabular presentation; causality; model building, regression and choice of methods; meta-analysis; and miscellaneous. Here, we present the common errors, with brief explanations. We hope that the guidance provided here will help guide authors as they prepare their submissions to the journal, leading to higher quality and more robust research reporting.

Original language	English
Article number	e002755
Number of pages	7
Journal	BMJ Paediatrics Open
Volume	8
Issue number	1
DOIs	https://doi.org/10.1136/bmjpo-2024-002755
Publication status	Published - 15 Sept 2024

Bibliographical note

Access to Document

10.1136/bmjpo-2024-002755Licence: CC BY-NC

e002755.full
© Author(s) (or their employer(s)) 2024
Final published version, 401 KBLicence: CC BY-NC

Cite this

@article{6857652f28bf4e26871918e60b457b87,

title = "Common errors in statistics and methods",

abstract = "As statistical reviewers and editors for BMJ Paediatrics Open (BMJPO), we frequently see methodological and statistical errors in articles submitted to our journal. To make a list of these common errors and propose suitable corrections, and inspired by similar efforts at other leading journals, we surveyed the statistical reviewers and editors at BMJPO to collect their {\textquoteleft}pet peeves{\textquoteright} and examples of best practices.(1, 2) We have divided these into seven sections: graphics; statistical significance and related issues; presentation, vocabulary, textual and tabular presentation; causality; model building, regression and choice of methods; meta-analysis; and miscellaneous. Here, we present the common errors, with brief explanations. We hope that the guidance provided here will help guide authors as they prepare their submissions to the journal, leading to higher quality and more robust research reporting.",

author = "Peter Flom and Katie Harron and Javier Ballesteros and Chester Kalinda and Eirini Koutoumanou and Jeremy Miles and Nevitt, {Sarah Jane} and Peter Rohloff",

note = "{\textcopyright} Author(s) (or their employer(s)) 2024",

year = "2024",

month = sep,

day = "15",

doi = "10.1136/bmjpo-2024-002755",

language = "English",

volume = "8",

journal = "BMJ Paediatrics Open",

issn = "2399-9772",

publisher = "BMJ",

number = "1",

}

TY - JOUR

T1 - Common errors in statistics and methods

AU - Flom, Peter

AU - Harron, Katie

AU - Ballesteros, Javier

AU - Kalinda, Chester

AU - Koutoumanou, Eirini

AU - Miles, Jeremy

AU - Nevitt, Sarah Jane

AU - Rohloff, Peter

PY - 2024/9/15

Y1 - 2024/9/15

N2 - As statistical reviewers and editors for BMJ Paediatrics Open (BMJPO), we frequently see methodological and statistical errors in articles submitted to our journal. To make a list of these common errors and propose suitable corrections, and inspired by similar efforts at other leading journals, we surveyed the statistical reviewers and editors at BMJPO to collect their ‘pet peeves’ and examples of best practices.(1, 2) We have divided these into seven sections: graphics; statistical significance and related issues; presentation, vocabulary, textual and tabular presentation; causality; model building, regression and choice of methods; meta-analysis; and miscellaneous. Here, we present the common errors, with brief explanations. We hope that the guidance provided here will help guide authors as they prepare their submissions to the journal, leading to higher quality and more robust research reporting.

AB - As statistical reviewers and editors for BMJ Paediatrics Open (BMJPO), we frequently see methodological and statistical errors in articles submitted to our journal. To make a list of these common errors and propose suitable corrections, and inspired by similar efforts at other leading journals, we surveyed the statistical reviewers and editors at BMJPO to collect their ‘pet peeves’ and examples of best practices.(1, 2) We have divided these into seven sections: graphics; statistical significance and related issues; presentation, vocabulary, textual and tabular presentation; causality; model building, regression and choice of methods; meta-analysis; and miscellaneous. Here, we present the common errors, with brief explanations. We hope that the guidance provided here will help guide authors as they prepare their submissions to the journal, leading to higher quality and more robust research reporting.

U2 - 10.1136/bmjpo-2024-002755

DO - 10.1136/bmjpo-2024-002755

M3 - Article

SN - 2399-9772

VL - 8

JO - BMJ Paediatrics Open

JF - BMJ Paediatrics Open

IS - 1

M1 - e002755

ER -