Should Quantified Self Researchers Protect Themselves with Disclaimers?
When I made my September self-tracking data available for download, I thought I should include some sort of disclaimer. Not to limit access to data, but rather protect myself from potential liabilities associated with use, analysis and interpretation of my data by other parties. I turned to the Internet to find out how this disclaimer should look like, but to no avail: most of the discussions around Quantified-Self data so far have focused on data portability and privacy issues. I then turned to Twitter, Facebook, and Quantified-Self forum. My post on Quantified Self forum actually resulted in small but somewhat heated discussion, so I thought I should elaborate more on this issue in a separate post. So today I would like share my thoughts on some potential legal pitfalls associated with publishing your quantified-self research and data, and how they could be avoided by including a disclaimer.
Again, just to be clear, the point of the disclaimer that I have in mind is not to restrict access to data or research. I am 100% for making QS data and studies as open and free as possible, and believe that just like in the “big” science, peer-reviews and open exchange and communication of ideas and studies by citizen scientists are the only way to ensure quality of research. What I am really concerned about is those rare but likeable instances when our data and findings could lead to liabilities or involuntary involvement in legal disputes, due to misuse or misinterpretation.Below are examples of just two liabilities that I could come up with from the top of my head.
Liabilities Associated with Personal Injuries Caused Directly or Indirectly by Source of Information
In our attempt to improve our health conditions, performance and other aspects of our lives, we often experiment with various remedies. When we find something that eases our pain, or helps to boost our mental performance or make us sleep better, we share our findings with others. More than often, we back our findings with our own data, making it look like a real scientific study. As a result, we come across as “experts”, which we are most likely not. Other people, be it patients with similar ailments, or simply curious self-experimenters trying to improve their life, then follow our example, often without questioning us. What we often forget, however, is that what was helpful to us, may be detrimental to others. Standing on one leg may have helped me to improve my sleep, but may result in severe back pain and potential injuries for you. Chugging coffee with additives may boost my mental performance, but result in a serious allergic reaction or kidney failure in other people. Such is the nature of the “n=1” research. Thus, it is important to remind people that what worked for you, won’t necessarily work for them. It is also important to disclose whether you have a sufficient knowledge or experience in this area, and always advise potential readers to consult with medical, fitness or other professionals before replicating your experiment or using the recommended treatment.
Defamation and other “harm to reputation” liabilities
In our self-tracking studies, we often rely on certain tools, like gadgets and apps, and when published, such studies inadvertently become reviews and testimonies to the quality of these tools. When backed by the data and statistical analysis, such reviews may seem more like endorsements or rejections by experts. The problem, however, is that we are not real experts, and our findings are often based on our personal observational data and thus are not generalizable. It is important to explicitly state that when releasing data or self-tracking research findings in order to avoid potential problems with manufacturers and developers. I will use my September dataset as an example. In that dataset, I released three different metrics of sleep quality: sleep efficiency scores provided by 100+ dollars tracking device, a sleep tracking/alarm clock app, and my personal subjective assessment of sleep quality. The problem is that based on the correlations, one of these metrics does not seem to actually reflect sleep quality; the .99 cents app seems to be doing a better job at capturing sleep quality than a 100+ dollars tracking device. When summarizing my analysis, I address findings with caution and consider them inconclusive: after all, analysis is based on 18 days of data from a single subject. Still, what can stop the manufacturer of a 100+ dollar device to file a lawsuit against me, accusing me of making statements that are “harmful to reputation” of their product? Or what can stop someone from citing me and my data as a source to support a claim that the .99 cents app better measures sleep quality than more expensive devices?
And these are just two kinds of potential legal issues that I could come up with. Of course, there may be many more. It is my understanding that neither Creative Commons license, nor even First Amendment sometimes could sufficiently protect you from these liabilities. What is needed is a standard detailed disclaimer that could specify and delimit the scope of all the rights, risks and obligations for all the parties that could be potentially affected by your research and data.