For accuracy talk about controlled variables, equipment used to collect data like calibration maybe and of course how close your results are to the expected value. Reliability talk about whether the experiment was repeated at least 3 times retrieving similar results. Validity is a combination of accuracy reliability as well as if your experiment addressed your aim and tested your hypothesis.
By repeating the experiment AND retrieving similar results reliability to achieved. so things like minimising random errors like having an unstable environment that affects the controlled variables or bad readability of the measuring instruments.