Graphs, charts and tables that include a mean must also include a measure of spread, such as standard deviation, standard error or confidence intervals. In a graph, this can be done with error bars bounding the mean.
Tick marks along axes should be reasonable (e.g., "10, 20, 30..." and not "2, 5, 8...").
Quantitative data along an axes should be spaced according to it's magnitude (e.g., 8 should be twice as far from the origin as 4).
The bottom left corner of a graph should be (0,0) unless there is a really good reason.
Axes should not be in logscale unless there is a really good reason.
All axes must be labeled (e.g., "Bitrate") and given units (e.g., "(Mb/s)").
Make sure any screenshots and graphs have readable fonts, particularly if they are shrunk to fit. As a guideline, the font in the graph should be the same size or just slightly smaller than the reports main text font.
Pie charts are almost always the wrong graph type to pick.
For comparing distributions, instead of a histogram consider a box and whiskers plot or a cumulative distribution function.
For graphs with 2+ data sets, be sure to include a legend, spaced with a readable font.
Use of colors in graphs is encouraged, but also differentiate data sets with symbols and hashing.
A title on the graph is useful for discussions, but when the graph goes in a report the title should be removed and replaced with a figure number and caption.
All Graphs, Charts (and tables) must be accompanied by have figure numbers followed by a caption, even if the caption is already written inside the figure. E.g. "Figure 4: Results of first trial run."
Each figure must be referred to at least once in the text, telling the user what to look for in that figure. E.g. "Figure 4 shows...." Capitalize "Figure" when referring to it by number (e.g., "The results in Figure 3...")
Summing this up, below is an example:
https://web.cs.wpi.edu/~imgd2905/d20/samples/analysis-example.html
Do not worry about the data - instead look at the form. Note especially the 6 guidelines at the top.