Available online 10 December 2023, 117426

Author links open overlay panel, Section snippetsThe lowest p-value doesn't guarantee the most important factor

Generally, comparing p-values to identify the most important factor is not true. In fact, the p-value measures significance, not strength of association. A true interpretation of p-value is critical to avoid misleading conclusions.

Statistically significant does not show the prediction capability

The prediction power of variables in regression models is not associated with the amount of p-value, and being significant does not show any information about the prediction power.

The lack of any prediction accuracy investigation

Although the authors claimed that higher exercise capacity predicts lower coronary artery calcium scores, the lack of any appropriate indices to show the prediction capabilities is obvious. Instead, the authors must compare the adjusted R –squares between models with and without a specific variable or use the other goodness-of-fit indexes to show the prediction capability of each variable.

Unclear variable selection method

Using different variable selection techniques in regression analysis has a critical effect on the magnitude of p-values by regulating the inclusion of variables into the final model. Due to the fact the reported p-values for age are very close to 0.05 (p = 0.047 or p = 0.048) in the study of Chedid et al. implementation of another variable selection technique may give non-significant p-values (p ≥ 0.05).

Beyond the points discussed in this note, which we hope will be useful to the researchers,

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (1)G. Chedid et al.Higher Exercise Capacity, but Not Omega-3 Fatty Acid Consumption

(2023 Jun 26)

View full text

© 2023 Elsevier B.V. All rights reserved.