I wrote a 10 page research paper as part of a final project for a course on parasite ecology. The paper asked whether parasitic load varied between fish species across time. The dataset I used was gathered for Wood et al. (2023) and used Puget Sound fish samples spanning eight species and 130 years. Almost 700 individual fish were dissected to create this dataset. There is very limited historic parasite data, and it is extremely difficult to get in the present. Dissecting preserved samples such as those at the Burke Museum is the best way to do so. 

I used a linear mixed effects model in R to answer my research question. Going into the project, I was totally unfamiliar with this method of analysis and had to learn a lot about coding in R and how to best represent the data in order to produce results. The learning process was frustrating, and as I found out very quickly, not something I could do on my own. I was relatively new to R at the time and had no idea how many extensions were needed to carry out analysis this complex. After days of running into a wall, I realized that I had to download separate extensions to carry out the analysis, run secondary tests, and graph the data. The TAs for the course were a huge help in breaking down how the model worked in R and what additional resources I needed to use in order to use it at all. 

Using a complex form of analysis that I had to learn in real time meant that I had very limited time to conduct secondary analysis to verify my results. The data had decades-long gaps and was not uniform in sample period or volume across fish species. If I were to perform the same analysis again, I would like to consult additional resources to try and fill in the gaps in my data and conduct additional tests to reduce the chance for error.

While the results of the model did not support my hypothesis, it ended up being a blessing in disguise. By rejecting my hypothesis, I was forced to figure out how to interpret data that displays the total opposite of the central prediction I had spent weeks working around and how it affects the rest of my work. The journey left me a better problem-solver and added a new tool to my skillset.