Hi, I was wondering what was the best way to compare T-statistics from 5 t-tests of different metrics (i.e. each metric contributed independently to the development of a disease, so I compare the difference in the metric for a person with the disease vs. a healthy person and find the significance level using a t-test).
What I want to do is come up with a weighting of these 5 metrics based on their level of significance in the t-tests to arrive at a final score. So how would I determine how to weight a metric that had let's say a t-test with a t-statistic of 3 vs another metric with a t-statistic of 6. Would I weight the latter at twice the first metric?