STATISTICS IN ECOLOGY WITH R /

HYPOTHESIS TESTING WITH R

Welcome to By 572- Seminar in Ecology. I am a conservation biologist by training with much of my foundations based on fundamental ecology. I am excited to discuss ecology with you. 

The relationship between ecology and statistics has gradually been growing stronger and today they are joined at the hip. Whether you are interested in understanding how species are distributed using species distribution models and investigating the main drivers of species ranges, or simply in measuring biodiversity with the myriad of diversity indices, or in investigating population dynamics of animals using mark-recapture methods and other spatially explicit models, or even in tapping into the wealth on data from citizen science programs (eBird, inaturalists ), statistic forms the backbone of much of ecology today. Conducting research in ecology thus requires a broad foundation in data management and statistical analyses. Moreover, as a graduate student understanding statistical analyses in publications and selecting the right analysis for your data is imperative. 

In this course we will use R, a free programming language, to develop your data management skills including data access, data organization, and quality control (e.g., looking for data entry errors), and analysis skills including exploring data, choosing and applying appropriate statistical tests, converting data to graphical representations, and producing reproducible results.

R is a free software for statistical computing and graphics. It is now considered one of the most popular analytics tools in the world, especially in academia. Consequently, job postings targeting candidates with a bachelor's or master's degree and requiring knowledge of R have increased over the past decade (Auker & Barthelmess 2020). R is a powerful open-source statistical tool that has applications in various fields from research to industry (Government - recordkeeping, forecasting; Finance, IT; social media - behavior analysis). This makes R an especially important skill set as a gateway to a lucrative career not limited to academia.

Coding may sound overwhelming, especially if you have little to no experience. The same may be said about statistical analyses. Learning statistics and various statistical packages in R is not just complementary but greatly enhances your ability to understand the logic behind different statistical analyses and their outputs. This course is designed for someone who has little experience in statistics and has never coded before.

Auker, L. A., and E. L. Barthelmess. 2020. Teaching R in the undergraduate ecology classroom: approaches, lessons learned, and recommendations. Ecosphere 11(4):e03060. 10.1002/ecs2.3060

COURSE BULLETIN BOARD: Canvas- I use this interface for class announcements, assignments, lecture power points, and practically everything. CHECK CANVAS ANNOUNCEMENTS AND your JSU EMAIL EVERY DAY!

Student Learning Outcomes: 

Students who complete the course will be able to

• Explain fundamental concepts in the design and analysis of ecological studies using observational and experimental studies

• Explain the logic behind statistical confidence intervals and hypothesis tests.

• Explain the logic behind parametric and nonparametric tests. 

• Able to conduct basic statistical analysis using codes in R studio.

• Compute and interpret measures of association for continuous and categorical data.

• Compare two (or more) groups based on continuous or categorical data using comparative measures and hypothesis tests.

• Compute diversity indices using R package vegan to differentiate between alpha (species richness) and beta (species turnover) diversity.

•  Convert data and analyses into graphical representations using various R packages including but not limited to ggplot 2. 

Suggestions for Success:

1. You should attend all the sessions in person: Each session will build on the previous making it imperative to do the exercises in class where you may ask questions or troubleshoot any issues.

2. Ask Questions: The primary aim of the course is to develop our statistical skills. Coding in R only supports the primary goal. If you do not understand the statistics or logic behind the code's outputs there is little sense in running the code smoothly.  

3. Do not hesitate to play around with the code: Like any programming language a small typo will hold up the entire code from running smoothly. Do not worry about it. We will have the source code saved on canvas that you may always copy and paste. Students tend to hesitate to make changes in the code. The best way to learn is to make mistakes and solve them together in class.

4.  All JSU students can schedule FREE in-person or virtual tutoring. In-person appointments will take place in the Student Success Center located at the Houston Cole Library, 2ndfloor, (256-782-8223). Virtual tutoring is available at tutor.com. To access tutor.com, log into Canvas and click the tutor.com link in the navigation pane

 HOMEWORK: Homework will be posted on canvas at 6 pm on Fridays and is due at 9 am the following Monday. Homework must be submitted on canvas. Late submissions will not be accepted unless previous accommodations are made with me before the time of submission

 FINAL PROJECT: The final project would involve analyzing secondary data on a topic/ question of your choice. You will have to mine the data from secondary sources, analyze the data in R and excel, and represent your analysis using graphs.  You will make a presentation of your project at the end of the semester. We will finalize the topic of your project after the lecture on "tests for linking data" (around the end of February).   

 DETAILED CLASS SCHEDULE