November 1986

Possible Statistical Error in an Endotracheal Tube Study

Author Affiliations

Meharry Medical College 1005 D. B. Todd Blvd PO Box 126 Nashville, TN 37208

Am J Dis Child. 1986;140(11):1099-1100. doi:10.1001/archpedi.1986.02140250025022

Sir.—I read with interest the recent article by Molteni and Bumstead in AJDC.1 The authors state: "Our study demonstrated no relationship between the development of a palatal groove and the use of a hard or soft endotracheal tube."

I am writing concerning the conclusion made from this study. When the study began, a total of 106 neonates met the guidelines for inclusion and when the final comparison was made it was based on a sample of only 39 of the preexisting neonates. Due to the small final sample size, I propose the possibility of chance bias that has caused a type II error (accepting the null hypothesis when it is false). I suggest that another study be evaluated with a larger sample size before the above conclusion is accepted as being statistically significant.

First Page Preview View Large
First page PDF preview
First page PDF preview