Many stories in science aimed both at the general public and technical stories between scientists hinge on understanding probability, but our brains aren't really built for comprehending probability. However, it's not really that hard to grasp on a basic level, so we can talk about the relative chances of a particular statement being "right" -- and avoid insulting anybody's intelligence in the process. (I did something of this sort in my Science Vs. Pseudoscience class last year. We actually tested telepathy statistically.) Trained scientists know (on an intellectual level at least) that absolute certainty isn't known, and working with error bars or other measures of uncertainty is standard. However, as narratives often focus on conflict and seeking out the rare dissenting voice on matters where there is a great deal of consensus (e.g. global climate change), it's essential to get an idea of levels of uncertainty. This session might involve learning to read (or learning to explain for those who know how to read) plots and other figures that have error quantified in them.
