Corpus ID: 906845

NEJM -- Data Torturing

  title={NEJM -- Data Torturing},
  • Published 2008
"If you torture your data long enough, they will tell you whatever you want to hear" has become a popular observation in our office. In plain English, this means that study data, if manipulated in enough different ways, can be made to prove whatever the investigator wants to prove. Unfortunately, this is generally true. Because every investigator wants to present results in the most exciting way, we all look for the most dramatic, positive findings in our data. When this process goes beyond… Expand


The scandal of poor medical research
The authors need less research, better research, and research done for the right reasons, and researchers who use the wrong techniques, use the right techniques wrongly, misinterpret their results, report their results selectively, cite the literature selectively, and draw unjustified conclusions. Expand
No Adjustments Are Needed for Multiple Comparisons
  • K. Rothman
  • Computer Science, Medicine
  • Epidemiology
  • 1990
A policy of not making adjustments for multiple comparisons is preferable because it will lead to fewer errors of interpretation when the data under evaluation are not random numbers but actual observations on nature. Expand
Obstetric complications and schizophrenia.
It is proposed that obstetric complications may indeed "be the secondary consequences of yet earlier events" such as genetically determined fetal abnormality, and Lewis and colleagues seem to overestimate the extent to which their strictures might call into question this association. Expand
Cancer control in susceptible groups: opportunities and challenges.
  • F. Li
  • Medicine
  • Journal of clinical oncology : official journal of the American Society of Clinical Oncology
  • 1999
This work has studied persons at highest risk as human models of cancer susceptibility to gain new knowledge of etiology, and identified biologic mechanisms of predisposition that modify cancer risk associated with exposures to environmental carcinogens. Expand
We Need Both Exploratory and Confirmatory
This work states that exploratory data analysis is an attitude, a flexibility, and a reliance on display, NOT a bundle of techniques, and should be so taught, and confirmatory analysis is easier to teach and easier to computerize. Expand
Quality of cancer care.
  • J. Vardy, I. Tannock
  • Medicine
  • Annals of oncology : official journal of the European Society for Medical Oncology
  • 2004
Determining the right treatment requires a hierarchy of evidence from clinical trials, and high quality clinical trials to determine and supplement that evidence, which are then useful in increasing compliance with evidence-based treatment. Expand
Outcomes research in regional anesthesia and analgesia.
An overview of outcomesResearch, the current status of outcomes research in regional anesthesia, and future directions for determining the benefits of regional anesthesia and postoperative analgesia are provided. Expand
Documenting research in scientific articles: guidelines for authors: reporting research designs and activities.
The design and activities guidelines for Authors: Reporting Research Documenting Research in Scientific Articles and Chest Tom Lang Designs and Activities Guidelines for Authors are published. Expand
Modeling the relation between socioeconomic status and mortality in a mixture of majority and minority ethnic groups.
The authors illustrate conventional and Bayesian approaches to analysis with data from the 1996 census used by the New Zealand Census-Mortality Study, using spline smoothing within a generalized additive model framework as an exploratory data analysis. Expand