Area charts show the relative contributions over time that each data
series makes to a whole picture. For example, an area chart would be
good to show how much the relative amounts of the principal and
interest change over time of a mortgage.
Bar charts compare distinct items or show single items at distinct
intervals. Usually, a bar chart is laid out with categories along the
vertical axis and values along the horizontal axis. In other words,
the bars are horizontally placed on the page. Bar charts are useful for comparing data items that are in
competition, so it makes sense to place the longest bars on top and
the others in descending order beneath the longest one.
Column charts are like bar charts because they compare distinct items or show single items at distinct intervals. However, column charts have the categories arranged along the horizontal axis and the values along the vertical axis, so the bars are vertical on the chart.
A very common use for column charts is to display how values change
over discrete units of time (monthly or yearly change, for example).
a line chart plots the value of the variable as a specific point, then
'connects the dots' in order to give you some idea of the relationship
of consecutive points.Line charts may also be used to show how the value of a variable changes over time. Unlike bar and column charts, line charts imply continuous change rather than a number of discrete points. For this reason, line charts are better at implying a trend. For example, if you are doing an experiment about the number of fish in a certain pond, you might be interested in the number of fish in the pond at a certain time, but you may also be very concerned with the trend of the fish population. Is it increasing or decreasing?
Just because a line chart implies trends does not necessarily mean they are there! Be careful when interpreting such charts that you do not automatically assume intermediate values by the line placement.
A pie chart is used to show proportions of a whole. It is very useful for figures that relate to a larger sum, such as demographic data or budget information. It is easy to get a feel for the relationship between component values when they are placed in a pie chart. Be careful that you do not have too many slices in the pie, or they will become meaningless.
Also, note that a pie chart is usually used as a snapshot of ONE
moment in time. If you want to show relationships as part of a whole
over time, you would use an area chart. If you want to look at a
number of pie charts at once, you might consider a doughnut chart. (look it up in online help or just play around with it!)
A scatter plot is the simplest type of graph. It simply plots the
data points against their values, without adding an connecting lines,
bars or other stuff. This is visually the least appealing type of
graph, but the lack of bells and whistles can actually be an
advantage. Since all the other types of graphs tend to have some kind
of psychological implication built in (eg bar charts imply comparison,
line graphs imply continuity), scatter diagrams are devoid of this
type of clutter. If you are searching for the patterns and meaning in
a graph, you may find the scatter diagram the clearest representation
of the data. Once you understand what it means, you can use one of
the other types of graph to give your readers whichever impression you
choose.
Many of the above charts can be created in 3-dimensional forms. The charts work pretty much the same way, but they can be a little more complex to work with. A three dimensional chart can be rotated so that it can be seen from other perspectives. 3-D charts are often used for the added dramatic impact they provide to a presentation.