Statistics has always been something that I have enjoyed doing. I do not always love math but when it is framed in a scientific or biological light I tend to drift towards it. In high school I took an AP statistics course and I really liked the concepts but I had trouble consistently using formulas correctly. Once I got to Longwood I was placed in an Intro Statistics course which helped this problem tremendously. I then really got to use these skills when I took Biology 251, Intro to Ecology and Evolution. In this course, studying large populations was commonplace and I knew that statistics would come into the picture soon enough. While collecting a lot of cemetery data for our survivorship curve project, my group and I quickly realized we would need some sort of statistics to be able to draw valid conclusions. We wanted to see if local cemeteries in Farmville and cemeteries from around the world had similar survivorship. We utilized our knowledge of ANOVA and T-tests to see if our data was significant. I was also able to use my excel skills to create graphs for our presentation. These tests are so helpful after collecting as much data as we did. We were able to conclude that our data followed international trends, like women living longer than men. Here is a link to the presentation, which includes slides for our statistical tests.
Additionally, I was able to even further build on these skills as I took animal behavior where I had to calculate all my equations for statistical tests on my own. As I had only used calculators and programs before this class, it was difficult to relearn equations and computing difficult math. I do understand now though where exactly the numbers are coming from and why our conclusions make sense after using these tests. While challenging, it was super beneficial to my understanding of data analysis. For this class, I needed to compare two populations and see if their behavior was different or the same for a mating response. Since we had reviewed every test beforehand it was easy for me to choose the Kruskal-Wallis test which works for multiple experimental groups. We did not find statistically significant data, however our sample size was small so we speculate that affected our outcome. Either way, being able to interpret data like this is much easier after this course and I am confident I can continue to draw these sorts of conclusions from data. Here is a link to the final paper, where you can find detailed statistical analysis of our data.