A p -value cut-off of 0.2 was used to detect variables that may not be significantly associated to the outcome in the univariate analysis and may be significant in a multiple model after adjustment of other covariates.
← all excerpts
Long-Term Effects of Exposure to Ionizing Irradiation on Periodontal Health Status - The Tinea capitis Cohort Study.
2
—
—
The sentences
Variables that did not reach statistical significance in the regression models were excluded using a backward elimination method.