Research Ethics

In this class, you will learn about the importance of research ethics. You will also learn about computer simulations, and you will conduct some simple experiments using a simulation.

Preparation

First, watch/read these resources about human subjects research:

• Henrietta Lacks, the Tuskegee Experiment, and Ethical Data Collection
• Facebook Tinkers With Users’ Emotions in News Feed Experiment, Stirring Outcry

Computer simulations are sometimes used as an alternative to real-world experiments. One benefit of computer simulations is that they avoid harm to research subjects.

This video shows examples of computer simulations. You are welcome to watch the whole video, but it is sufficient to watch from 5:45 to 10:37.

• Computer Simulation: Exploring Nature with a Computer

Optional: Consider also reading:

• What models can and cannot tell us about COVID-19
• Study Confirms Climate Models are Getting Future Warming Projections Right
• Science relies on computer modelling – so what happens when it goes wrong?
• Surprising things happen when you put 25 AI agents together in an RPG town

In Class

Activity 1: Discussion

We will start class by discussing the readings.

Activity 2: Explore the Disease Spread Simulation Project

First, read page 1 of BJC’s Simulations lab, which explains why simulations might be preferred over real world experiments.

As described on page 2 of the Simulations lab, it would be unethical to intentionally spread a disease to measure how quickly it spreads. Instead, a program could be used to simulate disease spread. We have already completed the Disease Spread Simulation project for you. You will use this project to conduct several experiments.

Next, inspect the code of the completed Disease Spread Simulation project. In particular, review the code for the “average steps for disease spread,” “simulate disease,” “initialize person,” and “move person” blocks.

Run a single simulation by clicking the gray “simulate disease” block. This block reports the number of simulated steps before all individuals became infected. Click the gray “simulate disease” block several more times, and record the number of steps reported by each run.

Discuss with a neighbor: Why does the gray “simulate disease” block report a different number each time?

Next, run a series of simulations by clicking the green “average steps for disease spread” block. This block runs 20 simulations, then reports the average number of simulated steps before all individuals became infected. Click the green “average steps for disease spread” block several more times, and record the number of steps reported by each run.

Discuss with a neighbor: What is the advantage of using the green “average steps for disease spread” block instead of the gray “simulate disease” block?

Note: You can speed up the simulation by enabling “Turbo mode” from the settings menu.

Note: In BJC’s starter code, the blocks report the time taken for all individuals to become infected. In our code, the blocks report the number of steps required by the simulation. What is the advantage of reporting the number of steps, instead of the amount of time taken?

Activity 3: Experiments with Simulations

Next, you will conduct experiments to test the relationship between population size, percent initially infected, and movement speed (independent variables) and the number of simulation steps needed to infect the entire population (dependent variable).

Run experiments to fill in the blanks in the template below. For each category of experiment, one independent variable is changed while the others are fixed.

Hint: To evaluate the relationship between variables, I recommend creating a scatterplot with the independent variable on the x-axis (e.g., population size), and the dependent variable on the y-axis (i.e., average steps).

Hint: Instead of running each experiment manually, you could store the results directly into a variable in Snap!

Submit

Upload a PDF containing:

• Your name(s)
• A filled-in version of the template shown below, including at least 300 words total responding to:
  • The disease spread reflection questions
  • The “Other Computer Simulations” research activity
• Thoughts on how this activity could be improved in the future

Template

Population Size Experiments

Population Size	Initially Infected	Movement Speed	Average Steps
25	5%	2
50	5%	2
100	5%	2
200	5%	2
400	5%	2
800	5%	2

Percent Initially Infected Experiments

Population Size	Initially Infected	Movement Speed	Average Steps
100	1%	2
100	2%	2
100	4%	2
100	8%	2
100	16%	2
100	32%	2
100	64%	2

Movement Speed Experiments

Population Size	Initially Infected	Movement Speed	Average Steps
100	5%	1
100	5%	2
100	5%	4
100	5%	8
100	5%	16
100	5%	32
100	5%	64

Disease Spread Simulation Reflection Questions

• Based on your data, explain the relationship between the number of simulation steps needed to infect the entire population and each of these variables:
  • Population Size
  • Percent Initially Infected
  • Movement Speed
• What real-world phenomena do you think this simulation shows?
• List the ways in which this model oversimplifies real-world disease spread.

Other Computer Simulations

Computer simulations are used in many areas of scientific research. Read and link to an article about a computer simulation created for scientific research. Answer these questions:

• What real-world phenomena does the simulation try to capture (e.g., disease spread, erosion, climate change, etc.)?
• What claims or predictions do the researchers make based on their simulation?

Note: If you’re unsure where to start, search for “computer simulations in science.”

Learning Goals

Students will:

• Learn about research ethics
• Conduct experiments using computer simulations
• Reflect on the strengths and limitations of computer simulations
• Develop written communication skills

Acknowledgements

Thanks to Stanford’s BJC materials, which served as a basis for this assignment.