The next step is Method. We determined that because the problem involves estimation, sample data, one numerical variable, and two paired samples, a potential method is the paired-samples t confidence interval. The next step is Check. There are two requirements for the paired-samples t confidence interval. The first requirement is that the subjects in the sample are either randomly selected or are representative of the population. We are given that the subjects can be considered representative of the population, so this requirement is met. The second requirement is that the sample size is large or the distribution of differences for the population of healthy adult males is approximately normal. Recall that we are given a set of data from 10 healthy men who were exposed to diesel exhaust for 1 hour. The sample size is only n = 7, so n is less than 30, and we need to consider whether the distribution of differences is approximately normal. Choose the boxplot of the ten sample difference.