In the hierarchy of evidence, where a randomized controlled trial is considered the highest level, how does the likelihood of bias change as we ascend towards the top of the evidence pyramid?