User: | Open Learning Faculty Member:
I knew after designing the experiment that I would be creating a box plot due to the comparison between a continuous and a categorical variable. While a strip plot would also have been acceptable, the box plot easily presents several other pieces of information that are harder to incorporate in a strip plot: upper and lower quartiles, minimum, and maximum values in addition to the mean value. Therefore not only does a box plot show you the mean values and the range of data, but it also allows one to visually appreciate the variance in within each data group. As discussed in post 6, I revised my experimental design slightly to aggregate samples from nearby trees into unified sampling areas. The reason for the revision is that there was no way to guarantee the leaves that had fallen beneath a given tree were grown on that particular plant. I opted therefore to group sampled leaves into a “habitat areas” that included the sampling area and a control area.
Figure 1. The effects of airborne particulate pollution on premature leaf abscission were estimated based on the leaf length of dropped leaves of Prunus spp. in two areas: near a roadway and construction site with active digging (Particulate-Exposed) and approximately 70 meters East of this location (Control). Mean leaf length between Particulate-Exposed (=50.5 mm) and Control (=54.8mm) are shown in bold horizontal lines. Upper and lower bounds of the box represent the 75th and 25th quartile values, respectively. Upper and lower “whiskers” represent maximum and minimum data points, respectively. (Figure created in R Studio V1.1.414)
I really like this figure – it has a great explanation in the figure text and very clear labels