good science is good data. there's no way around it. it's the same thing. sure we devise clever (and expensive) experiments to falsify bad models, but that's be

good science is good data. there's no way around it. it's the same thing. sure we devise clever (and expensive) experiments to falsify bad models, but that's because we don't have enough high quality data that you can look at and immediately tell which model is good and which is bad. but nowadays everything in science is even more data driven. (mostly because we have the infrastructure/technology to process all that data.) initially it was mostly high energy physics (run the big machine for a few years, gather data, check for 5 sigma) but when it comes to things like LIGO and the EHT (event horizon telescope) to even get the signal out of the noise they need to use a model, that of course builds on data (and its interpretation) from previous observations. many advances in the medical field are due to better microscopes, better techniques to gather data. sure, without interpretation it's just pretty images or just big "science machines" (in case of colliders and other data gathering experiments)