Hey guys, John here again, and in lecture set eight here we're going to consider taking into account sampling variability through confidence intervals where we're looking at measures that compare two populations using data from two samples. We'll find out the general drill for creating confidence intervals for things like mean differences or differences in proportions is the same as what we saw for single measures. We take our estimated difference and then subtract two estimated standard errors and we'll show how to do that. Then for ratios and we've alluded to this before, we're going to need to take things to the log scale before doing their confidence interval computations. So, it'll be a two-step process where we first do the computations for the confidence interval in the log ratio scale and then anti-log or exponentiate our results back to the ratio scale. So, in this process we'll see that generally again, the procedure is similar across all these computations. We take our estimate of interests and add and subtract two standard errors, but we'll also focus on interpreting the range of possible values for association measure whether it be a mean difference, difference in proportions, relative risk etc. Interpreting that in the scientific context and paying attention to something called a null value whose presence would indicate no association at the population level. We're going to be very interested as to whether that number appears in our confidence interval or not. So for example, if we were computing a difference between two populations for example, mean difference, there are really no difference at the population level then the true mean difference would be zero. So, we're going to be concerned about whether zero appears as a possible value in our confidence interval or not, and if not we're going to define the concept called statistically significant and talk about that as well. So, this will be a really interesting lecture set, so we'll start seeing how the results and inclusions were made for some of the seminal studies we've looked at and other studies as well.