Metasophi currently applies an algorithm in order to identify how one variable is correlated with another, while controlling for other possible influences. It is not fool-proof, so you need to exercise common sense when interpreting the results. Correlation does not always imply causation. The algorithm will be improved over time, so you can expect the result to change slightly every now and then.

Some questions you might want to ask yourself after a result is: Is there something untracked that could account for this variable? Could you start tracking this? Was there a variable not controlled for in this specific regression that could have had an influence? You can check this information in the "More detail" link on the specific result. If this variable was not controlled for, check that it has been tracked for at least 10 days in the Data Evolution page.

We can take no responsibility for the quality of the data, which is dependent on the devices you use, how you use them, and your own accuracy when manually tracking data. Therefore we cannot provide any quality assurance for the statistics, which is based upon said data.

If you publish any graphs, tables, or statistics from Metasophi, or otherwise act on said information, you are considered as the researcher as only you can verify the quality of the data supplied and therefore the quality of the statistics. Consequently, you also assume all liability emerging from the use of such statistics, either by you or someone else.

Some other important points.

First, it is not permitted under our Terms of Service to use Metasophi to either 1) ascertain or verify/compare the efficacy of a medicine or 2) ascertain or verify/compare the safety of a medicine. Efficacy is defined as the concept of demonstrating scientifically whether and to what extent a medicine is capable of diagnosing, preventing or treating a disease.

Second, this service must not be used to track adverse events emerging from the use of medicinal products.

An approach to answering the above two questions requires a much more rigorous approach to data quality and analysis which is beyond the scope of Metasophi to provide.

If you are linked or affiliated to an institution receiving federal funding from the US government, then we advise you to look into the regulations relating to the Common Rule. If what you are doing qualifies as research, then you may need to get approval from a relevant IRB. For the purposes of IRB review, research is considered as "a systematic investigation, including research development, testing, and evaluation, designed to develop or contribute to generalizable knowledge", where "contribute to generalizable knowledge means that the purpose or intent of the project is to test or to develop scientific theories or hypotheses, or to draw conclusions that are intended to be applicable and/or shared beyond the populations or situations being studied." See here for more information.

According to some sources, "Federal regulations do not distinguish between self-experimentation and research on subjects recruited for a specific project. Faculty, staff, and students who wish to act as participants in their own research are considered human subjects. All self-experimentation human research activities require IRB review and approval prior to commencing the study." See here for more information.

For more information on the Common Rule and the institutions to which it applies, please consult the following:

Revised Common Rule Regulatory Test

Federal Policy for the Protection of Human Subjects ('Common Rule')