Eliminating systematic errors Since systematic errors are biased, they must be avoided if experimental data is to be reliable. You will recall from the Measurement Corner #20 that systematic errors occur when equipment is faulty or incorrectly calibrated, when equipment is used incorrectly, or when the experimenter is biased. To eliminate systematic errors in your experiment, you should check each one of these possible sources. Firstly, it is important to know how your equipment works and how to use it correctly. Faulty calibrations will always cause systematic errors, so it is best to carefully calibrate equipment before any project. (We have often skipped this step in this class to save time.) Some measurands are directly measured and some are inferred from measurements. You should check that your equipment is actually measuring what you intend. For example, we say that we "weigh" an object using a balance, but a balance doesn't actually measure weight; balances measure normal force, which is often, but not always, equal to weight. Just as important is setting aside any personal biases while you are performing an experiment. Before we experiment, we make predictions. It is natural to hope that your predictions are correct, so we are easily tempted to skew our measurements so that they match our predictions. This is bias and a source of systematic errors. In fact, this bias can be a form of science fraud. Systematic errors are often unintentional and subtle, so they are difficult to identify and fix. Habitually checking your results can be of great help in finding systematic errors. You can check your experiment against theoretical models; this is one major reason for making quantitative predictions before experiments. You can also check your experiment against results of another independent experiment. The other experiment, especially if it uses a different measurement technique, is likely to not experience the same systematic errors as your experiment. The disagreement with theory (if the theory is correct) or an independent experiment points out that a systematic error occurred somewhere. In case of disagreement, you should not change the results of your measurements in order to agree, but keep a keen eye out for possible problems with measurement technique or biases.

1. Think of an example of a systematic error. Propose a way to fix the systematic error.