Bar Graph? Pizza Graph? No graph?
I have two data tables, but the numbers it's too similar between both, a LITTLE difference occurs in my data. I want to show this for the public. Show can i do?
(I have Microsoft Excel and R installed)
***Edited***
Now i have Statistica 7, too
*********
Thanks
Thanks for this nice comment, Luiz :-)
When it comes to plotting I agree with Tukey. Pizza should be eaten - not plotted. I personally also dislike bar charts (see for instance this poster for arguments: http://biostat.mc.vanderbilt.edu/wiki/pub/Main/TatsukiKoyama/Poster3.pdf; I also find this policy noteworthy: http://biostat.mc.vanderbilt.edu/wiki/Main/StatisticalPolicy).
For presenting data from "investigative research" I recommend to show the data in scatter plots rather than summaries. This is more explorative and the eye is best in recognizing patterns. If simple numbers are to be presented (either because the data are simple numbers or they are summarized in single numbers), I prefer dot charts over bar carts (http://www.b-eye-network.com/view/2468). If there are really a lot of values by group, (1D-)scatter plots may be over-done. Here, Tukey's Box-and-Whisker plots are a good summary graph (http://en.wikipedia.org/wiki/Box_plot).
(btw for R there is the beeswarm package to produce "beeswarm plots", a kind of 1D scatter plots with a systematic way to prevent overplotting)
Maybe it would be better to present the numbers in tables actually, in order to the reader check the real differences. If you have some grouping variable, you should also test for the statistical significance of the differences and present the results of the test as well.
As always, Dr. Wilhelm has given a nice suggestion. As a complementary approach, you cold present these normalized differences or ratios in a bar plot , to improve clarity. The great Professor Tukey once said "There is no data that can be displayed in a pie chart, that cannot be displayed BETTER in some other type of chart."
I personally agree with him.
Cheers,
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics