display | more...

Bar graphs, or bar charts are a way in which one can visually represent data. They consist of a set of bars (elongated rectangles), the length of which indicate the number (or frequency) of the variable being measured.

   Eggs Nogged

100 |
    |                 #
 80 |                 #
    |                 #
 60 |                 #
    |                 #
 40 |                 #
    |        #        #        
 20 |        #        #        #
    |        #        #        #
  0 |        #        #        #           
            '00      '01     '02      '03*

   *Watch this space

But bar graphs may be more interesting than you think... They come in different flavors! I will attempt to introduce you to some of the more common. If I'm missing your favorite bar graph, /msg and I'll add it.

Histogram: Not just the length, but also the width of the bars contain information (class intervals or somesuch).

Stacked Bar Chart: The bar is subdivided (usually by a change in colour), so that you can see the different subcategories within each category.

           Cars Observed

                 8 |                   ---- 
                   |                  |====|
                 6 |       ----       |====|
                   |      |====|      |====|
                 4 | Blue |====| Blue |====|
                   |      |====|      |    |
                 2 |      |    |      |    |
                   | Red  |    |  Red |    |
                 1 |      |    |      |    |
                           Mon.        Tues.

Multiple Bar Chart: Much the same as the Stacked Bar Chart, but instead of being stacked the ‘subcategories’ are placed in two (or occasionally more) separate bars, paired so that they are touching. This is used when you want emphasize the difference between the categories; with the Stacked Bar Chart, it can be hard to compare the sizes of the subcategories, as the stacked (upper) category is not starting from a fixed point (it is starting from wherever the previous subcategory left off, which often varies). These are frequently used in the social sciences. The two categories most often paired off are male and female.

Tower Chart: A three dimensional bar chart, or a bar chart presented as if it is in a three dimensional space. This is usually done so that you can place related bar charts next to each other in the third dimension (one behind the other).
http://www.causeway.co.uk/tutorial/rainpro/tutor/apl2000/chtower.htm has a nice picture.

Gantt chart: A bar graph in which the bars are not attached either edge – originally, and usually, used for representing when events take place in time.

       Eat Lunch -- Node -- Snack -- Sleep 
12:00 |               X
1:00  |   X
2:00  |   X
3:00  |               X
4:00  |               X             
5:00  |               X
6:00  |               X
7:00  |               X      X
8:00  |               X
9:00  |               X
10:00 |               X
11:00 |               X
12:00 |               X
1:00  |                      X
2:00  |                               X
3:00  |                               X

You might also want to check out the closely related Pie chart, which is good at emphasizing/comparing differences in frequency; Line graph, the most familiar of the graphs; and Table, for when a graph just isn’t needed.

Note: you’ll notice that I’m not systematic in my usage of the terms ‘bar chart’ and ‘bar graph’. This is because I have been attempting to use the term that is used most frequently (I’m using Google searches to determine this). I have no idea why it is ‘Bar Graph’ but ‘Multiple Bar Chart’ – but it is. Admittedly ‘Multiple Bar Chart’ beat out ‘Multiple Bar Graph’ by only a small amount, but it’s my node, and I’ll be pedantic if I want to.

Log in or register to write something here or to contact authors.