New concepts about so-called common sense are always challenging, not disregarding that it´s true observational studies likely provide the best available evidence or “the identification of harms when randomised controlled trials would seem unethical or impractical”(1).
In statistical hypothesis testing, a type I error is the incorrect rejection of a true null hypothesis. so is a "false positive", while a type II error is incorrectly retaining a false null hypothesis, i.e, a "false negative".
Simply put, a type I error is the false detection of an effect that is not present, while a type II error is the failure to detect an effect that is present. Emphasizing that in this study authored by Topiwala et al, because of the small number of subjects involved, and the always large 95% confidence intervals of alcohol-related brain health effects presented; a type 1 statistical error(2), could have happened, namely when using proxies or surrogates markers of diseases, in the occasions where exist today all kinds of non-alcoholic beverages for testing of pragmatic randomized clinical trials,(3).
Table of error types(2).
Table of error types Null hypothesis (H0) is
Decision About Null Hypothesis (H0) Reject Type I error
(False Positive) Correct inference
Fail to reject Correct inference
(True Negative) Type II error
1. S. Barton. Which clinical studies provide the best evidence?
The best RCT still trumps the best observational study. BMJ. 2000 Jul 29; 321(7256): 255–256.
2. Sheskin, David (2004). Handbook of Parametric and Nonparametric Statistical Procedures. CRC Press. p. 54. ISBN 1584884401.
3. Laura A. Young, MD, PhD; John B. Buse, MD, PhD; Mark A. Weaver, PhD; Maihan B., et al, for the Monitor Trial Group . Glucose Self-Monitoring in Non–Insulin-Treated Patients With Type 2 Diabetes in Primary Care Settings. A Randomized Trial. JAMA Intern Med. doi:10.1001/jamainternmed.2017.1233 Published online June 10, 2017.
Competing interests: No competing interests