International A Level Biology / Practical Skills in Biology I
Accuracy, precision, validity and repeatability
Edexcel International A Level Biology (YBI11), chapter 13: Practical Skills in Biology I, specification point U3.2, U3.3. Free lesson.
Read the revision notes · International A Level Biology past papers
Accuracy, precision, repeatability, reproducibility and validity describe different qualities of experimental results. Knowing which quality is weak helps you identify errors and choose the right way to improve an investigation.
What the five terms mean
Each term describes a different part of the evidence. Repeating a test can show whether results are consistent, but it does not automatically make them closer to the true value. Accurate results are close to the true value. Calibration, more colour standards or a colorimeter can improve accuracy. Precise results have little spread: repeated readings are close together. Tighter control of variables can reduce the random variation that makes readings spread out. Repeatable results are much the same when you repeat the investigation using the same method and equipment. Repeat at…
How random and systematic errors differ
The type of error determines how to deal with it. A random error varies unpredictably, while a systematic error shifts readings in the same direction each time. Random error makes repeated readings disagree, so it reduces precision. Beetroot discs cut by hand may vary in thickness and leak different amounts. Stopping a stopwatch a little early or late is another example. Reduce random error by controlling variables more tightly, repeating readings and calculating a mean. Systematic error makes readings consistently too high or too low, so it reduces accuracy. A colorimeter…
How to deal with an anomalous result
An anomalous result does not fit the pattern of the other results. On a graph, it may appear well away from the line. Repeat the reading to check it, then leave it out when calculating a suitable mean if it remains anomalous.
Why repeats do not always improve accuracy
In a test using colour standards, the unknown colour can only be placed between two standards. Repeating the estimate with the same standards gives the same limited range, so it does not make the estimate more accurate. Use more colour standards at smaller intervals to narrow the range. A colorimeter measures colour as a number, which can give a more accurate estimate.
How controlling variables makes a test valid
Protease acts on the surface of squid meat. Cutting pieces to the same size gives them the same surface area for the enzyme to act on. This controls another factor, so only the independent variable affects the result and the investigation is valid. Accuracy, reliability and precision do not describe this reason.