Robustness Issues of Clinical Significance and Clinical Importance

Four articles address issues related to the measurement challenges and the relationship between statistical significance and clinical importance. A commentary by Li, Walter and Thabane [20-00282] comment on a series of challenges of over-emphasis on statistical significance: i) studies that are not statistically significant but are clinically relevant e.g. due to poor statistical power; (ii) failure to distinguish between absence of evidence, versus evidence of absence; (iii) the fragility of statistical significance where a difference of a handful of outcomes makes or loses statistical significance; (iv) the temptation of p-hacking; (v) the temptation for research misconduct; (vi) the temptation to ‘spin’ exaggerating and overemphasising the importance of statistical significance of a study.
Source: Journal of Clinical Epidemiology - Category: Epidemiology Authors: Tags: Editors Choice Sept 2021 Source Type: research