Now that you're familiar with the main concepts of diagnostic test metrics, we'll go through some examples which illustrate how you can calculate these metrics. Let's go back to our epidemiology classroom example. While the boring lecturer is teaching, 10 of the 50 students are asleep. The true prevalence of the classroom sleeping disease is 10 over 50, 0.2 or 20 percent. However, the lecturer cannot examine closely each and every student, which would be the gold standard method this occasion. He looks for the podium, trying to identify who is sleeping but he's shortsighted, so he's diagnostic method is imperfect and results in some misclassification. You can see the students that the lecturer thinks are asleep in red. Some of them are awake, they're false positives. On the other hand, the lecturer missed some of those who are asleep. These are false negatives. Let's see if you can calculate the sensitivity of the lecturer's diagnostic test. That are 10 students who are sleeping and this will be your denominator. Among those, eight have been correctly identified as being asleep by the lecturer, this is your numerator. Therefore, the sensitivity of the lecturer's test is eight over 10, 80 percent. 80 percent of the students with the disease were correctly identified as having the disease by the diagnostic test. Similarly, forty students are not sleeping, they're disease-free. This is the denominator of the specificity. Among those 40 students, the lecturer has identified 36 as being awake. Thus, the specificity of the lecturer's method is 36 over 40, 90 percent. 90 percent of the students who do not have the disease were correctly identified as healthy by the lecturer's test. Next metric is the positive predictive value. For this one, the denominator is based on the test result. The lecturer's test was positive for 12 students. He thought that 12 students are asleep. Among them, eight were truly sleeping. Therefore, the numerator is eight and the denominator is 12. The positive predictive value is eight over 12, 67 percent. This means that among those who were identified as sleeping by the diagnostic test, 67 percent were actually sleeping. To calculate the negative predictive value, you look at the number of students who are identified as non sleepers, which is 38 and use it in the denominator. Then, you focus on those people and count how many among them were truly awake. The number is 36 and this is the numerator. The negative predictive value is 36 over 38, 95 percent. Among those who were identified as healthy, not having the disease, 95 percent were in fact healthy. No matter how many different examples you use, you will always get the same results when calculating sensitivity and specificity. They refer to the test regardless of the context. But depending on the prevalence of sleeping in the classroom, positive and negative predictive values may vary widely. You should now be able to calculate and interpret diagnostic test metrics. This will help you correctly interpret results of diagnostic tests in clinical contexts, focusing on the individual who's undertaking the test, but also to consider the implications of using a diagnostic test at the population level.