Science needs philosophical insight, as shown by the history of science:
According to Chalmers ("What Is This Thing Called Science"), theory precedes observation.
As these preceding theories are fallible and incomplete; they may be misleading.
The idea that real knowledge occurs when theories are falsified doesn't always work. The data that falsified the theory may be wrong; as the experiment relied on assumptions (theories) that may have been wrong.
Heinrich Hertz presumably didn't record his shirt colour, what he had for breakfast, the dimensions of the lab, and numerous other seemingly irrelevant observations; when testing Maxwell's idea about generating radio waves from electromagnetism. Hertz couldn't understand why the waves he got didn't travel at the speed of light.
Turned out that one of the 'irrelevant' potential observations WAS relevant: the waves were bouncing off the lab walls and interfering and affecting the velocity reading. The selectivity that led to ignoring the dimensions of the lab was an example of the fallibility of the theoretical assumptions that underpin experiments.
Copernicus's theory of the planets going around the sun predicted parrallax; that near stars would appear to move against the star background over 6 months (half an earth orbit around the sun). The observations 'falsified' Copernicus as no parrallax was found for centuries. But the theory lived on beside the 'falsification'; eventually the mistaken assumption (theory) that the stars were closer than they are now known to be, was exposed by better instruments that detected the small parrallax.