Researchers are often interested in how an intervention might impact a population. For example, as part of the work I am involved with, we are interested in how various HIV prevention strategies will impact HIV incidence and prevalence in Mochudi. In order to do this, we need to measure the existing incidence and prevalence of HIV. However, simply measuring these things may be considered an intervention. By providing HIV testing opportunities in the community that go above and beyond the existing HIV testing services, we are altering the environment enough to likely impact peoples’ knowledge and behaviors, and subsequently, the HIV incidence and prevalence... hopefully for the better.
This example of measurement as intervention is unavoidable. In some imaginary world, researchers might like to be able to wave an “HIV detecting wand” over a community and get a reading on the HIV prevalence and incidence. However, in the real world we don’t have that capability so we will have to use our current methods and understand that by measuring the HIV prevalence and incidence we are also providing an intervention that will likely impact the variables we are interested in... which may, in fact, be a good thing.
Other examples of "measurement as intervention" raise ethical issues where there is a conflict between the ‘purity’ of the data and providing research participants with needed care. I recently attended a presentation where the researchers, in the course of their study, identified many participants who presented with symptoms of depression. Although depression was a variable the researchers were interested in studying and may have been tempted not to ‘interfere,’ they were ethically obligated to link these participants with services to help address their depression.
During the presentation, a member of the audience challenged the researchers stating that by providing treatment for these depressed participants they were contaminating the data— as if the researchers had erred in their judgment. In essence, he was saying that if the researchers really wanted to know how depression might impact the outcome variables they need to let it ‘run its course’ without intervening.
This comment made me squirm in my chair— are some people really so concerned with the ‘quality’ of their data that they would hesitate to provide treatment to participants who need it just because intervening might complicate their data analyses? Researchers routinely bemoan how long it takes to apply for and get approvals from supervisory bodies (IRBs, etc.) and resent the fact that various entities are constantly looking over their shoulders and auditing their every move. This recent exchange helped me see why it is so important. Sometimes researchers get very focused on the needs of the study, how they can get the ‘best’ data possible and seem to lose sight of the fact that the whole reason for doing research in the first place is to improve peoples’ lives.
The more I reflected on this situation, the more I appreciated the opportunity to see this example of researchers doing the right thing even though it may have been ‘interesting’ to watch and see what happened with the depressed patients in the absence of appropriate psychological assessment and care. I thought back to my ethics courses in school and subsequent trainings. In these courses we learned about cases such as the Tuskegee syphilis experiments and gasped in horror that scientists could make such blatant ethical offenses. But, the truth is, unless we closely monitor our own behavior and allow external supervision of our work, we are likely to descend the slippery slope to similar sins. The end does not always justify the means.
