It is argued that a line graph actually shows the data ‘better’ (whatever that means)
than a bar chart but you have to be careful about what you can do with the graph.

Normally a line graph should be used when both the variables (in this case the years and the sales values) are continuously changing, and then you can interpolate (i.e. read between the dots). Well you can’t do that here. The dot half way between 1992 and 1994 is clearly 1993, and certainly we had sales in 1993. But you can’t split the line between 1992 and 1993. It doesn’t split into quarters of the year, or into months, because the sales values are end-of-year totals. The information is not broken down any more than that.

But I digress. Let’s go back to bar charts.

Suppose we want to compare more than one set of similar data. An ideal vehicle for this is the compound bar chart.