Before I describe how we measure lead in drinking water, I want to discuss some of the new (to me) meta issues associated with this project. A lot of these things are uncommon for physics research, but will be familiar to folks in biology or medicine. I am a rank amateur on these issues, so consider this an imperfect discussion by someone still finding his way.


As a physicist, I was only vaguely aware of what an IRB was and the role it plays in research. IRB stands for Institutional Review Board. Under U.S. law, all research done on human subjects must be reviewed by an internal committee at the institution conducting the research to ensure that no harm or potential harm is done to subjects of the research. The IRB also reviews research on animal subjects to make sure the work is humane. Today it seems obvious that research should not needlessly endanger the subjects, but in the not-so-distant past an “ends justifies the means” attitude was taken by some scientists, particularly if the subjects were perceived as having less value than the average citizen. Racial and ethic minorities, prisoners, and the mentally impaired were all considered little better than animals by various researchers. And while it is easy to condemn the horrific work of Nazi scientists on Jews and other minorites during World War II, the United States has its own history of malicious research. A particularly vile example is the infamous Tuskegee Syphilis experiments in which unsuspecting African-American patients were deliberately infected with syphilis and intentionally denied treatment so that the scientists could study the progression of the disease. This lasted from the 1930s until it was shutdown by the government in 1972 thanks to a whistleblower leaking to the press and the subsequent public outrage.

Although IRBs were initially tasked with overseeing human research, they have expanded to cover animal subjects, and sometimes non-medical research on humans (e.g.~sociological surveys). The purpose of the latter type of review is not so much to prevent physical harm to subjects, but to prevent their information from being used against them. Typically, we require procedures to protect the identities of the subjects, for example.

One of our goals is to inform residents if the water in their homes contain lead. That seems simple enough, but we have to also consider cases where the residents do not want to know. We also have to protect our data so that no one can use it against our subjects. Here are a couple of bad situations we want to avoid:

Case 1

Suppose we test the water in a home and get a positive result. The resident is concerned but doesn’t have the resources to fix the plumbing or install filters. Although we’re trying to help them stay safe, we’ve now caused financial and emotional stress.

Case 2

Similar to the last case, but the resident is renting. Under law, the landlord has an obligation to fix the pipes or they can no longer rent out the home. The landlord could use this as an excuse to evict the resident rather than pay repair costs.

Case 3

We collect data on many homes in a neighborhood, and somehow our data get stolen. A malicious developer uses our data to get properties condemned and then purchases them to build a new apartment complex. The original residents are forced out and can not afford to live in the new housing.

These may be extreme examples, but our main point is that we don’t want anyone to be harmed by participating in our study.

We also have to be careful about what we claim with our test. Although we believe we meet rigorous scientific standards, our system does not have government certification. Therefore, our results are not “official,” and cannot, for example, be used in court. If we do not clearly state that our results do not meet legal standards, then we could be sued for making false claims. So, in the literature that we distribute to potential subjects, we have lots of weasel-words. Our results are only “suggestive” of the presence of lead, for example. If we get a positive result, we encourage people to follow up with an EPA-certified test. (Fortunately, the City of Pittsburgh will pay for testing, but the wait is long.)

We also have to protect our data in the case of potential lawsuits. Suppose we find lead in a rental property and the landlord sues us claiming that our test is invalid and that this caused unnecessary repairs to the property. The courts may then force us to reveal the names of the participants. Unfortunately, we need to take measures to protect ourselves and the participants in case something like this were to happen.

Perhaps the key concept in making sure our work is ethical is the doctrine of Informed Consent. The idea is simple: any test subjects should be aware of all possible risks of participating in an experiment, and agree to these risks. In practice this means that we are responsible for creating a document that states, in pay-person’s terms, what we are doing, what the results mean, and how they could affect the participants. In our case, there are no direct risks to the participants (in one interpretation, there are no human participants, only houses), but we feel we have at least an ethical and moral obligation, if not a legal one, to give our volunteers as much information as possible and in a clear way. We’re still working on these documents, and we’ll seek input on them before we distribute them.