Reliability: repeat the experiment and get identical/similar results (depending if you're getting quantitive or qualitative data)
Validity: make sure you have the same conditions except for the variable you want to study (independent), reduce as many extra variables as possible
Accuracy:it depends on the experiment e.g. use electrical equipment such as a data logger
Validity is also having an appropriate experimental method for the experimental aim, using the right equipment (don't use a thermometer for measure distance etc).