Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Project Argus: Re-creating a web application for interactive a priori power analysis
In this master project, you will learn about statistical power analysis to implement an interactive web application with D3.
Keywords: R, JavaScript, Shiny, D3, Power Analysis
Controlled experiments are essential for establishing a cause-effect relationship between stimuli and measurements. Such experiments are widely used to evaluate new interaction techniques or applications. When planning a controlled experiment, researchers need to decide the required sample size, i.e. the number of participants. A priori power analysis helps to inform this decision.
Conducting an a priori power analysis is a non-trivial task as researchers need to take several parameters into account that all influence the resulting sample size. Wang et al. (2020) created Argus, a web application prototype that researchers use to perform such power analysis more easily. With Argus, researchers are able to explore the dynamic relationship between the input and output parameters more easily to better plan the sample sizes for their experiments. Argus is currently available here: https://argus.shinyapps.io/project-argus/ (use control + 1 for a preset).
The goal of this master project is to implement a production-ready version of the existing Argus web application. With this project, you will advance your skills in Javascript, R, and statistics. A deeper understanding of power analysis in the realms of controlled experiments can be very useful if you intent to work in fields close to HCI or UX.
WORK PACKAGES
1. Understand a priori power analysis and how the widgets in Argus work based on the original paper.
2. Learn the technologies (R, Shiny, D3, Javascript) used in the application.
3. Inspect the current code base as preparation for the implementation.
4. Implement Argus production-ready open source project that will be shared with the HCI community.
REFERENCES
Wang, X., Eiselmayer, A., Mackay, W. E., Hornbæk, K., & Wacharamanotham, C. (2020). Argus: Interactive a priori Power Analysis. arXiv preprint arXiv:2009.07564. URL: https://arxiv.org/abs/2009.07564
ADMINISTRATIVE INFORMATION
Duration of project: 6 months
Availability: Master project for a team of 2–3 students (18 ECTS). The work packages will be scaled or partitioned proportionally to the number of students.
Supervisor: Alexander Eiselmayer, Prof. Dr. Chat Wacharamanotham
Controlled experiments are essential for establishing a cause-effect relationship between stimuli and measurements. Such experiments are widely used to evaluate new interaction techniques or applications. When planning a controlled experiment, researchers need to decide the required sample size, i.e. the number of participants. A priori power analysis helps to inform this decision. Conducting an a priori power analysis is a non-trivial task as researchers need to take several parameters into account that all influence the resulting sample size. Wang et al. (2020) created Argus, a web application prototype that researchers use to perform such power analysis more easily. With Argus, researchers are able to explore the dynamic relationship between the input and output parameters more easily to better plan the sample sizes for their experiments. Argus is currently available here: https://argus.shinyapps.io/project-argus/ (use control + 1 for a preset). The goal of this master project is to implement a production-ready version of the existing Argus web application. With this project, you will advance your skills in Javascript, R, and statistics. A deeper understanding of power analysis in the realms of controlled experiments can be very useful if you intent to work in fields close to HCI or UX.
WORK PACKAGES 1. Understand a priori power analysis and how the widgets in Argus work based on the original paper. 2. Learn the technologies (R, Shiny, D3, Javascript) used in the application. 3. Inspect the current code base as preparation for the implementation. 4. Implement Argus production-ready open source project that will be shared with the HCI community.
REFERENCES Wang, X., Eiselmayer, A., Mackay, W. E., Hornbæk, K., & Wacharamanotham, C. (2020). Argus: Interactive a priori Power Analysis. arXiv preprint arXiv:2009.07564. URL: https://arxiv.org/abs/2009.07564
ADMINISTRATIVE INFORMATION Duration of project: 6 months Availability: Master project for a team of 2–3 students (18 ECTS). The work packages will be scaled or partitioned proportionally to the number of students. Supervisor: Alexander Eiselmayer, Prof. Dr. Chat Wacharamanotham
Not specified
Prepare your application package as detailed on zpac.ch/projects. Then, e-mail your application to eiselmayer@ifi.uzh.ch. If you apply as a team, send only one email. Team applications will be prioritized over applications from individuals.
Prepare your application package as detailed on zpac.ch/projects. Then, e-mail your application to eiselmayer@ifi.uzh.ch. If you apply as a team, send only one email. Team applications will be prioritized over applications from individuals.