Outliers

scale by log

>>9125056
Exclude them from the chart and specifically mention them

>>9125056
Present the data in a table rahter than a chart.

>>9125056
Make 2 graphs, one with everything except "other renewables"

File: outlier2.png (8KB, 487x271px) Image search: [Google]

8KB, 487x271px

>>9125059
Like so?

Or should I have it just max out my vertical axis?

File: bar-chart-natural-axis-split1.png (14KB, 402x290px) Image search: [Google]

14KB, 402x290px

>>9125056
Use a broken bar if your software can do it.

>>9125056
Just lie! Set it to 2.5 and no one will notice.

Out of curiosity, what is the y axis?

you can let it max out the vertical axis and write the value on top of the bar

>>9125059
this is a terrible idea

>>9125090
That's very misleading to anyone who doesn't know how to read graphs. Could be used to misconstrue findings.

>>9125283
The problem here is the "doesn't know how to read graphs" part, not the graph itself, isn't it? The log scale would cause the same problem. The exclusion of the outlier ditto. What's your proposal? Leave it as it is? It could be misleading, because we cant see the shorter bars that well.

>>9125283
>could be misleading to retards

>>9125068
I like this

>>9125056
You could use a logarithmic scale

>>9125283
It could also be misleading to people who can't read. Unless your paper is supposed to be presented to children, it's safe to assume the reader can read a graph.

>>9125056
First off, your independent variable is time. This should be a line graph.

>>9125056
Label your vertical axis m8.

>>9125818
Second, what is your dependent variable? If it's an average you can include standard error bars to allow for evaluation of the degree to which that one particular measurement is an outlier.

>>9125823
Third, depending on what data you are working with, you can use an appropriate statistical test to evaluate if that particular observation is statistically different from the other measurements in the "other renewables" category. If statistically significant, it is common practice to place an asterisk over that observation to indicate this significant difference.

>>9125804
Unless a politician or someone else with an agenda presents it to laypeople to purport their bogus views. This may or may not be a problem depending on your field.