Please explain the difference between making an observation in the lab and making an observation in an observatory.
I'm sure you can buy a $50 telescope at Walmart and discover Venus , Jupiter etc.
That's my point, I don't think there is a fundamental difference, MM thinks there is (although not always, the distinction of when there's a difference is unclear).
A lab observation is (or has the possibility to be anyway) easier because you can control variables. Move the lens, put it in a Faraday cage, factor out the earth's movement, whatever. It's harder to do when you have no control over what you are observing; coming up with different ways of looking at things to isolate variables.
How about giving me an actual example. Field "experiments" are fine as long as there is a control mechanism involved. "Observational studies" can be ok as well as long as I can be sure what you claim "did it" actually exists in nature and has some effect on nature. I can't tell if inflation once existed based *only* upon an observation.
Field experiments can involve a control, but the observations are made in the "wild" rather than in a lab. Medical studies where a control group receives a placebo.
Observational studies can lack a control, you can't run a medical study of a life saving drug by giving a placebo and seeing who dies, so these are more difficult.
You left out natural experiments... This is what astronomy and cosmology mostly utilize. Natural experiments rely on observations only, because variables can't be manipulated. So the observations are made to try and isolate the variables.
As for examples of fields that depend on natural experiments.. ecology, economics, geology, astronomy, cosmology, sociology, and medicine are all fields that will depend heavily on natural experiments where observations are the main (or only) kind of experiments that can be done.