Assuming that your results show trends or movements over time, such as nicotine concentration in plasma after smoking, a good way to display your data would be to construct a line graph. But do not rely on the computer to design it for you. Here are some common errors.

The line graph

This seemingly excellent line graph nevertheless appears to have two common defects: the curves are distinguished both by type of line and by type of data-point symbol – either would suffice; and the curves are identified by a separate key, obliging the reader to scan back and forth to the key to see what they represent.

In the two redrawn graphs (5.2), the curves are labeled directly and distinguished either by type of data-point symbol or by type of line.

Open and filled circles, as in the left graph, are the data-point symbols easiest to distinguish. They can also be used symbolically; for instance, if an experiment has been performed with (•) and without (O) treatment, the emptiness of the open circle suggests that nothing has been administered.

Other standard symbols for data points are open and filled squares and triangles (□ ▪ ▴). If you need more symbols, you probably have too many curves for one graph, and you should consider dividing it into two or presenting your observations in a table.