When the time comes to analyse the experimental data, a researcher must decide on the characteristics of any graph plotted so as to maximise the potential for extraction of further data. This means judicious choice of axes, scale and careful point plotting. |

Choice of axes

In principle the choice of axes is usually the easiest part of graphical data analysis. The independent (input) variable is chosen as the x-axis and the dependent (output) variable as the y-axis.

However, it is quite possible that a student performs an experiment and, understanding the theory behind the process, decides to plot either variable as an inverse, or logarithmic value. In this case the concept derived from the independent variable still goes on the x-axis.

Choice of scale

The scale is chosen to maximise the data extraction potential of the graph. A decision must be taken as to whether or not to start the axis scale at zero. There are many cases where it is simply not relevant to do so. However, there are also cases where it is imperative. The decision depends entirely on the graph plotted.

Example: If you are measuring pressure
in Nm If the units being plotted are inconveniently small or large, the axis can have a factor built in to make the numbers of the axis scale more comfortable. In this case the scale measures from 1 - 10 and the units are Nm |

On a purely common sense note, the scale should be chosen with plotting in mind. Graph paper is divided into squares usually subdivided into 10 smaller units. If your scale sets the value of each 'big' square at 3 units, it will be difficult to plot points.

Plotting the points

Once your scale has been chosen it should be easy to plot the points carefully for every x and y value. At this stage you should consider whether to include error bars to cover uncertainties in the event that you wish to draw a line of best fit. (For lines of best fit and error inclusion see the next section)