Learning Data Analysis Skills in Intro Biology Labs with RLinda Forrester, Rachel Schwartz, Bryan Dewsbury, Department of Biological Sciences

Goal“While colleges should include courses in programming, data visualization and statistics, more students develop digital fluency more quickly and easily when digital tools are integrated throughout the curriculum”

ObjectivesStudents will be able to

Apply fundamental quantitative reasoning skills to analyze data and solve problems Develop professional-quality data analysis skills Visualize and analyze large data sets Conduct data analysis using software R

Rationale Writing code to requires students to

o consider the goal of their analysiso specify x and y variableso consider how changes to code impacts their analysiso think and communicate in a new language

Written code is easily repeatable and modifiable for subsequent labs. Our approach allows students to reinforce and scaffold skills. R is a data analysis program that is used in many fields.

Approach Students collect data in lab covering different biological concepts. Students analyze their collected data using R. Students utilize server-based access to RStudio, which eliminates difficulties with installation on personal computers and

allows access anywhere.

Each lab builds on previous R code. When students learn new skills, they are given explicit code. When students build skills based on previous labs, they are given suggestions but not complete code. When students practice skills, they are expected to write their own code independently based on prior labs.

Figure 1. Lab 1. Plankton diversity in Narragansett Bay.

Figure 2. Lab 2. Mechanisms of evolution.

Figure 3. Lab 3. Photosynthetic pigments in local seaweed

Figure 4. Lab 4. Seedling growth

Description of labsDevelopment of skills using R through labsText with the word “New” in front of it indicates new skillsText with the word “Build” in front of it indicates building on skills developed in previous labsText with the word “Practice” in front of it indicates practicing skills developed in previous labs

Lab 1 – Narragansett Bay Plankton Diversity

Students count plankton samples from two sites in Narragansett Bay New: Students enter counts into a shared spreadsheet to collate data learning how to organize data (long vs wide format

data) New: Students calculate mean and standard deviation of diversity for different sites New: Students learn the basics of how to use the program R New: Students create boxplots of diversity at different sites for different tidal cycles using explicit code and directions

Lab 2 – Mechanisms of Evolution -- effects on snail populations Students count changes in snail populations as affected by birds and physical processes Practice: Students enter counts into a shared spreadsheet to collate data Practice: Students calculate mean and standard deviation of diversity for different sites Build: Students create line graphs of changes in snail populations Build: Students add standard deviation error bars to plots New: Students use a t-test to compare effects of birds and physical processes on snail populations

Lab 3 – Photosynthetic pigments of Narragansett algae (seaweed) Students measured pigments in local algae using a spectrophotometer Practice: Students enter counts into a shared spreadsheet to collate data Practice: Students create line graphs representing photosynthetic pigments

Lab 4 – Seedling growth Students measured growth of seedlings under two conditions Practice: Students enter counts into a shared spreadsheet to collate data Practice: Students calculate mean and standard deviation of diversity for different sites Build: Students create line graphs of changes in snail populations New: Students add best fit regression line

Figure 5. R-Studio screen- creation of Lab 3 graph, seaweed photosynthetic pigments.

Results from student surveys after half of one semesterFigure 6. Most students already report their R skills are improving.

Figure 7. Some students already report they can use R to manipulate data and make graphs.

32% of students requested more time spent in lab on the instruction of how to write R code.

Figure 8. Some students already report they feel R makes work reproducible

Figure 9. Some students already report R is effective for making useful graphs.

Discussion Students are already reporting progress with writing code, making graphs, and working with data. Students with little computer coding experience can develop data analysis skills early in their academic career with this

approach. Using coding instruction embedded in lab to analyze data they have collected, students find coding skills to be relevant and useful.

Students are ready to continue learning applications of data analysis programs and want more instruction to improve their skills.

Future DirectionsHelp students discover the relevance of coding and data analysis program (like R) by:

We will have students propose optimal analyses and visualizations and show how these may be done in R rather than Excel.

We will have students work with large datasets that are easier to manipulate and examine in R than in Excel.Provide more time for learning so students become comfortable with analyses using code.Integrate R and data analysis in upper division biology courses.Provide explicit information to students in BIO 104 on how skills will scaffold across the curriculum.

“Wakanda forever”

Thanks to: Dean Kirby and CELS for funding to purchase lab laptops. Office of Student Learning Outcomes Assessment & Accreditation for funding to support grad student time for writing

R instruction manuals for each assignment. Chi Shen, University Computing Systems for maintaining URI R server (CELSRS.uri.edu)

References1. Cassidy & Siesing, 2017, Inside Higher Ed 2. Black Panther, 2018