 Health economics researchers want to have information about people's health. For this they collect biomedical data in the form of biomedical measurements during surveys. This data collection and also having to notify someone of their test results because of ethical reasons could actually change people's behavior. In and of itself this is not a bad outcome because you do want these poor people to go to the doctor to get healthier but from a research perspective where you're maybe interested in isolating or identifying a different effect, you could have the problem that that effect is not identified because of this measurement that you did. My research is a randomized experiment of blood pressure measurements during a household survey in Tanzania, in the Kilimanjaro region to look at the response of people to getting this measurement and also getting the test results, so whether they have high blood pressure or normal blood pressure and so I look at whether they go to the doctor more often and whether they take health insurance more often and this is in a setting where health insurance is being offered to one half of the sample but not to the other half of the sample so this can be nicely identified. So what I find is that people who have this high blood pressure measurements during the survey they do go to the doctor more often but actually are not more likely to take health insurance. It's important to know whether healthcare interventions actually have a beneficial effect on the people that they're meant for. Now to do that a researcher needs to collect the data. If you indeed want to get to this effect of the healthcare intervention you have to be able to disentangle that from the effect of the biomedical measurement and the way to do that is to randomly exclude a sub-sample from your biomedical measurement. Sometimes the test itself and not the healthcare intervention changes behavior and so you want to know what is really contributing for the effect that you're finding. The main takeaway of my research is that rapid biomedical testing indeed can change people's behavior and I think that researchers should do more to account for this behavioral change.