Consequently, some observed interactions, particularly those with wider confidence intervals or borderline significance, should be interpreted as exploratory and hypothesis-generating rather than confirmatory, to avoid an increased risk of Type I errors.
← all excerpts
Associations between edentulism and risk of neurodegenerative diseases among middle-aged and older adults in China: a decade-long cohort study.
1
—
—