Good science should involve randomized, well controlled experiments, the use of power calculations to determine appropriate sample and effect sizes, appropriate statistics that allow statements to be generalized to the whole population (random field theory), sufficient evaluation periods, publication of negative or unexpected results and independent replications.
Negative findings and reproducibility
'Research cannot be self-correcting when information is missing' (Nature 551, 414, 2017). We wish to undo some of the damage done to the field of science by academic publishing, which has favored novel and positive findings to to detriment of reproducibility and reliability.
Stewarding the Data Commons
We are piloting a brand new way of dealing with data - the abolition of private ownership of data, replaced by collective decision making over the use of collectively produced data. All collected data and any intelligence that springs from it should be collectively owned by the community it is about and put to use for this community. A loss of privacy should never result in a loss of (social) power. In this way, we go beyond traditional models of informed consent for data.
We believe that all people can contribute to the body of knowledge on which our civilization is founded. This project allows people to come and participate in the experiments (‘socialize for science’), to be part of the analysis process and to have those data and findings visualized for them in a way that hopes to encourage an understand of both the scientific process and human behavior.
We acknowledge the importance of both hypothesis and goal driven science, but also value blue skies exploration of phenomena. Whilst we value replications and negative findings, we pledge to ensure that at least 25% of our studies explore unexpected avenues also.