HOW STATISTICS CAN MISLEAD/BLEICH ETAL. RESPOND

[...]the conditions differ only negligibly when effects are presented as risk ratios or absolute percentages (Table 1); the ARRs were 0.96, 0.96, and 0.94 for absolute calorie, relative calorie, and exercise equivalents, respectively. Sainani criticizes our article for "a needless emphasis on o...

Full description

Saved in:
Bibliographic Details
Published in:American journal of public health (1971) 2012-08, Vol.102 (8), p.E3
Main Authors: Sainani, Kristin L, Bleich, Sara N, Herring, Bradley J, Flagg, Desmond D, Gary-Webb, Tiffany L
Format: Article
Language:English
Subjects:
Beverages
Calories
Diet
Editors
Food
Health risk assessment
Low income groups
Medical research
Public health
Ratios
Statistics
Studies
Teenagers
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:[...]the conditions differ only negligibly when effects are presented as risk ratios or absolute percentages (Table 1); the ARRs were 0.96, 0.96, and 0.94 for absolute calorie, relative calorie, and exercise equivalents, respectively. Sainani criticizes our article for "a needless emphasis on odds ratios" but we present the change in sugar-sweetened beverage rates (Table 2 in original article), odds ratios (ORs; Table 3 in original), and the predicted probability of beverage purchases pre- and postintervention (Figure 1 in original). Because we acknowledge that the use of odds ratios in certain contexts may be undesirable, we deliberately reported the unadjusted frequencies and predicted probabilities to give an undistorted sense of the effect size. Human Participant Protection The Johns Hopkins Bloomberg School of Public Health institutional review board approved the study.
ISSN:0090-0036