What are ways to improve the accuracy, precision, reliability and validity of an experiment?
Adding onto Chocolatepistachio:
Accuracy (how similar your results are to the expected/correct results) and precision (how similar the results are with each other) are linked with reliability (whether the experiment can be reproduced and repeated to get the same/similar results), and these contribute to validity (whether the experiment actually tests what it claims to be testing).
To fix accuracy, you want to do these:
- calibrate and maintain instruments
- repeat the experiment
- use equipment with less uncertainty (like if you're measuring 5mL, you want to use a measuring cylinder with 0.5mL intervals rather than a beaker with 10mL intervals)
To fix precision you can take some from accuracy:
- calibrate and maintain instruments
- if using live organisms such as mice make sure they're all the same age/species/health in each sample to minimise random errors. What I mean is that if you have 15 adult aged mice in a vaccination experiment and 5 old mice and then the old mice die after taking the vaccine, your results aren't really that precise and that may prevent the vaccine from being trialled on humans (because a 66% success rate is not good) but in reality your vaccine IS successful, it's just that 5 mice are old. I suppose these come under controlled variables.
- have a large sample size to minimise effect of outliers
- use equipment with less uncertainty (like if you're measuring 5mL, you want to use a measuring cylinder with 0.5mL intervals rather than a beaker with 10mL intervals)
To ensure validity have controlled variables, control groups (either with a positive or a negative control), specific measurements for everything, and ensure to test only one independent variable.
Controlled variables to include in your experiment
- environmental conditions: temp, access to light, light intensity, light wavelengths (light is usually only relevant for enzymes and photosynthesis experiments)
- species of whatever
- number of everything
- duration of experiment
- if using live organisms then have the same diet for them given at the same time of the day (very important for cellular respiration)
- also with live organisms, whatever you're doing to them, do it at the same time. Vaccination, measuring glucose concentration, or whatever, do it at the same time every day/whenever you're doing it.
Anyone feel free to correct me as I may be wrong