In the last module, we discussed negative controls. They help us to know that a positive result is really positive. But there are times we need to know whether a negative result is really negative. That’s where positive controls come in. They’re the topic of this module in Experimental Design. [credits] Let’s say in our last experiment, we found that caffeinated coffee caused people to be no more active than water. We might assume, therefore, that caffeine has no effect on activity. But what if something is wrong with our measurements? What if we thought we were measuring activity with the pedometers, but it turned out that there’s something wrong with the system. How might we know? We need a positive control. This is an intervention where we know something is supposed to happen. Let’s say there’s an energy drink on the market that’s been proven to increase people’s activity. We could add in that as a control. If the coffee had no effect, the water had no effect, but the energy drink did have an effect, we could be reassured that the negative results we got for coffee are correct. If, however, the energy drink also had no effect, then something might be wrong with the pedometers. Only with a positive control, can you be sure that your negative results are real. Here’s a basic science example. If we were testing an enzyme assay to measure how much of a certain enzyme was in a set of extracts, a positive control would be an assay which contains a known amount of the purified enzyme we’re interested in. If we get the right answer on that positive control, we can be assured that the enzyme works when it’s supposed to. We can be sure that if we get a negative result, then there really isn’t any enzyme in that set of extracts. Harkening back to last module, a negative control would be a known solution with no enzyme at all. If our assay detects no activity in that solution, then we can be assured that a positive result in the set of extracts means there really must be some enzyme in there. Another form of positive controls is what we call a gold standard. Let’s use a human experiment this time. Say we want to compare a new test for bacteremia (a bacterial infection of the blood). The old test, blood culture, is really great – but it’s slow. A new test would be better. We might get a lot of people together who we think are sick and check their blood with the new, rapid test. Some will be negative and some will be positive. The way we determine if the new test is as accurate as the old one is to check it against the one we know to be correct. That would be the blood culture – a gold standard. In this case, blood culture is a positive control. It’s positive when we know it should be positive. Therefore, we can see if the new test is positive at the same time. Positive controls can also be important for the calibration of instruments. We can check our measurements against those controls with known results to make sure that our scales or tools are accurate (and precise, for that matter). One final use of positive controls is in determining dose response. We may want to see how different amounts of caffeine affect activity. By adding in the energy drink positive control, we might get a ceiling against which the different results obtained by varying amounts of caffeine can be compared. If we only use negative controls, like placebos, we’re going to miss a huge amount of data that’s necessary to ensure that our research is reproducible. Positive controls are just as important, and they should be considered before any experiment is begun.