User: | Open Learning Faculty Member:
Figure 1: Boxplot showing median (thick line), interquartile range (box), variability (whiskers), and outliers of flower per plant measurements at each site.
Due to snow, I have changed my project and traveled several hours further afield to find sites that aren’t coated in a sheet of ice. As such, I am now investigating differences in average number of flowers on Achillea millefolium (commonly known as yarrow) at two sites near Abbotsford with different elevations.
Luckily for me, I am taking this course concurrently with statistics, and so I am becoming very comfortable with using R to analyze data and produce graphs. I decided that a boxplot was the best way to compare my data, due to the ease with which one can see the differences between the two categories in median and variability. Seeing the precise numbers of every sample is simply unnecessary.
The graph does show that Site B averages a higher number of flowers than Site A. Unfortunately, the graph also exposes that the differences are fairly slight and are not likely to be statistically significance (but maybe I’ll get lucky and they will be).
I feel like I could have used further observations to reduce the variability and be more confident that the differences aren’t chance, but unfortunately the snow shows no signs of melting and another four hour trip to Abbotsford is not particularly feasible, so I will have to settle for my poorly designed little experiment and hope to gain marks for pointing out how poorly designed it is.
