Sometimes one of the most complicated parts of research is the analysis. In order to properly analyze data, the researcher must know what they collected and what they are trying to prove. There are many different statistical tests that can be run to determine significance but choosing the correct one can be challenging. At Longwood, many biology courses require some form of statistical analysis to allow students to understand the importance of collecting good, reliable data. Some of the projects that required statistical analysis came from classes such as Introduction to Ecology and Evolution and Immunology.
One of the first mini projects that I remember that involved analysis was in my Introduction to Ecology and Evolution course. We were given 5 sets of spider data from different sites in Sir Lanka in order to determine which site should be the main focus of conservation. This project challenged us to learn the specific tests and processes within excel needed to analyze this data by determining the total number of species, the species richness, heterogeneity, and any endemic species. Simpson’s Index of Dominance was used to find species richness and Shannon-Wiener Index of Species Diversity was used to find heterogeneity. Based on the results site 1 was determined to be the best choice for conservation due to its high abundance and diversity of spider species. Below is my mini report for this project which I wrote in the Fall of 2017.
Also, in my Introduction to Ecology and Evolution course, we conducted a larger project that involved the students to develop their experiment. For this project, my project revolved around the assessment of the daily activity patterns of birds in Longwood Lancer Park. This required many hours of watching and recording the abundance and diversity of bird species that visited over the course of a day over many weeks. Atmospheric data was also collected to identify any correlations. This project accumulated a ton of data that needed to be analyzed a few different ways. Using the statistical analysis program R, a one-way ANOVA test was run to find any significant differences between abundance, diversity, and time of day. After that to further analyze the significance time of day has, a Tukey HSD test was run to compare the times of day against one another. Diversity and dominance were determined by the Shannon-Wiener Index and Simpson’s Index. The last analysis performed were linear regression graphs for all the atmospheric parameters. Based on this analysis, it was determined that bird species diversity was highest during the morning and evening hours and diversity was negatively impacted by temperature and wind speed. Below is my research report for this project which I wrote in the Fall of 2017.
An Assessment of Daily Activity Patterns of Birds in Longwood Lancer Park
Unlike my previous projects in Introduction to Ecology and Evolution as shown above, my project in Immunology had a simpler method for analysis. This course is a higher level one usually taken by juniors and seniors so the project itself became more complex, but the analysis only required two-way t-tests to determine significance for each experiment. This project investigated the effect of ethyl and butyl parabens to mimic estrogen to alter the differentiation of myeloid derived suppressor cells (MDSCs). For each experiment, the different parabens were compared to the two controls, media and estrogen to find any significance between the treatments. Based on the analysis, both of these parabens mimic estrogen to induce the differentiation of MDSCs which are associated with cancer. Below is my poster that I presented at Longwood University’s Spring Showcase for Research and Creative Inquiry in 2019.
While I have only listed a few of my projects that highlight data analysis, many more of my courses at Longwood use some form of analysis during 1-day labs, semester projects, as well as my own undergraduate research. Data analysis is a crucial part of an experiment because it allows you to confidentially state if your data is credible or not.