This post assumes that you have read That Is Like, So Unfair about what makes a fair test. When you are asked to comment on why a particular step is taken in an experiment it is tempting to answer, “To make it a fair test” without really thinking it through carefully. There are two other major considerations – accuracy and precision. They do not mean the same thing!
Accuracy refers to how close the outcome of an experiment is to the real or true answer. If the results of your experiment are randomly spread around the true answer then you can make your investigation more accurate by repeating the experiment several times and averaging the result. Of course, if your experiment is not very well designed then no amount of repetition is going to get you closer to the truth.
Precision describes how closely grouped the results are. You can increase precision by measuring more carefully but doing more repetitions will not improve precision.
Precision can also be used to describe how finely you are measuring your results. If you use a normal school ruler then your precision is to the nearest millimetre because that is most precise measurement that the ruler can give. If you are using our most basic balances then the maximum precision you weigh something to is 1 g. However, if you use our fancy expensive balances then you can measure to the nearest 100<sup>th</sup> of a gram – it is 100 times more precise.
.
It is possible for the outcome of an experiment to be…
- accurate and precise (the answer is correct and the data are close together)
- inaccurate but precise (the data are closely grouped but not near the true value)
- accurate but imprecise (the average of the data produces an accurate answer but they are widely spread)
- inaccurate and imprecise (the data are widely spread and far from the true answer)
When you are asked in an exam why certain steps were taken during an experiment please take the time to think it through carefully. It will probably be to increase either precision or accuracy or fairness but not all three at once!
Another good post to read about doing experiments is Tryptophan. If any links don’t work – please let me know. The site is undergoing something of an overhaul and so bits may stop working.
Questions…
- What is the single characteristic of a fair test?
- In an investigation, what is the dependent variable?
- Which variable goes on the x-axis of a graph?
- From what are proteins built?
- What biological structures give the code for building proteins?
- Enzymes are proteins. Name an enzyme and the reaction that it catalyses.
Comments