![]() In your case, I suspect that red-cell velocity and vessel diameter might be inversely related (wider vessels, slower velocity at constant volume flow) so they aren't really independent. The significance of one entire regression model is examined at the first step rather than 5 separate independent correlations, improving the multiple-comparison problem too. Very often the association of a predictor variable with an outcome variable becomes clearer when other variables are taken into account in this way, enhancing your ability to find true relationships. Instead of comparing red-cell velocity against each of the 5 predictors separately, examine how all 5 predictors are related to red-cell velocity when they are all considered together. You might, however, get more useful results by doing a multiple linear regression. The correction in the Holm method depends on the particular p values that were found for the individual tests. The Holm method still keeps the same protection against making false positives but can be less likely to discard true positives. The strictest Bonferroni correction with 10 independent comparisons would require a p-value cutoff of (0.05/10) or 0.005 for significance at a false-discovery rate of 0.05. If you do 10 independent comparisons when there are no true correlations and accept a p value of 0.05 as the criterion for statistical significance, then you have about a 40% probability of at least 1 false positive finding by chance. You measured all of these in several instances and examined the 10 correlations of each of the 2 outcome variables against each of the 5 predictors.įor that type of analysis you certainly are running into the multiple comparisons problem. You have 5 potential predictors: mean arterial pressure, hemoglobin level, left ventricular pressure, serum sodium, and serum potassium. ![]() You have 2 outcome variables: red-cell velocity and vessel diameter.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |