People always say correlation does not equal causation but they never mention how does one determine causation definitively? Unlike typical observational scien

People always say correlation does not equal causation but they never mention how does one determine causation definitively? Unlike typical observational science, causation must be established by having the experimenter intervene with the experiment himself. The experimenter must introduce or remove smart phones to the population being experimented on and record any changes to the data. So it's actually very possible for the author to prove his own theory beyond reasonable doubt. These are the 2 options I can think of that can be done individually or together: 1. Remove smart phones away from people who are depressed. Measure depression levels over time. 2. Pick a set of humans that don't use smart phones and choose who will have access to smart phones and who never will over a period of 5 or 10 years. Measure depression levels over time. We have procedures in place to prove things like this with medicine. I think it's very possible to conduct experiments like this with the rigor of clinical trials given that you can literally remove or introduce the predicted causative agent from any human: smart phones. It's worth it to mention here that not all things are amenable to causative experiments. I believe in climate change but it is not possible to run a causative experiment where you remove or introduce the causative agent: greenhouse gases. It is unfortunate that climate change cannot be established as true beyond reasonable doubt simply because all current evidence is correlative. But for smart phones it is very very possible for the author to go far beyond a retort and actually prove his claims correct.