Skip to main content

I was granted the opportunity to work with Professor Berman and Professor Feit from Drexel University for 10 weeks during the summer of 2021. In 2019, Professor Berman and Professor Feit developed the latent stratification model for advertising experiments, which can estimate the average treatment effect more accurately and precisely. The goal of this project is to translate the theoretical components of the model into an R package and R Shiny online calculator for practical use. I was responsible for the package, the calculator, and the associated documentation.

Through this project, I was able to cultivate a variety of skills that make me well suited for future challenges. In order to prepare for the project, I spent the first five weeks learning how to develop R packages and Shiny apps. At the same time, I actively delved into the hardware and software foundations for code efficiency and clarity. By the end of the project, I achieved a deeper understanding of the R language as a whole.

Apart from the coding aspects, the project also fostered my literacy and design abilities. The project has trained me to think carefully about what makes a successful program. This included factors such as practical user interface and clear instructions. In the process of documenting the files, I had to accommodate both inexperienced and advanced users. Through repeated revisions, I learned to write for different audiences while keeping the instructions intuitive and interactive.

This year's PURM project was held remotely due to the covid-19 pandemic. Even though I wasn’t able to meet the professors face-to-face, the experience was still just as engaging. Furthermore, because of the scheduled hourly weekly meetings, I developed the habit of preparing an outline before every meeting for efficiency. As a result, I was able to improve my time management and organizational skills over the course of 10 weeks.

Finally, the project has also prepared me for my future in data science. In particular, the collaborative nature of the project has given me exposure to GitHub, allowing me to become familiar with the platform and polish my own portfolio. From this enriching experience, I have become a finer learner, researcher, and developer.

To see my poster, please visit Penn Presents: https://presentations.curf.upenn.edu/poster/latent-stratification-n-adv…

I was granted the opportunity to work with Professor Berman and Professor Feit from Drexel University for 10 weeks during the summer of 2021. In 2019, Professor Berman and Professor Feit developed the latent stratification model for advertising experiments, which can estimate the average treatment effect more accurately and precisely. The goal of this project is to translate the theoretical components of the model into an R package and R Shiny online calculator for practical use. I was responsible for the package, the calculator, and the associated documentation.

Through this project, I was able to cultivate a variety of skills that make me well suited for future challenges. In order to prepare for the project, I spent the first five weeks learning how to develop R packages and Shiny apps. At the same time, I actively delved into the hardware and software foundations for code efficiency and clarity. By the end of the project, I achieved a deeper understanding of the R language as a whole.

Apart from the coding aspects, the project also fostered my literacy and design abilities. The project has trained me to think carefully about what makes a successful program. This included factors such as practical user interface and clear instructions. In the process of documenting the files, I had to accommodate both inexperienced and advanced users. Through repeated revisions, I learned to write for different audiences while keeping the instructions intuitive and interactive.

This year's PURM project was held remotely due to the covid-19 pandemic. Even though I wasn’t able to meet the professors face-to-face, the experience was still just as engaging. Furthermore, because of the scheduled hourly weekly meetings, I developed the habit of preparing an outline before every meeting for efficiency. As a result, I was able to improve my time management and organizational skills over the course of 10 weeks.

Finally, the project has also prepared me for my future in data science. In particular, the collaborative nature of the project has given me exposure to GitHub, allowing me to become familiar with the platform and polish my own portfolio. From this enriching experience, I have become a finer learner, researcher, and developer.

To see my poster, please visit Penn Presents: https://presentations.curf.upenn.edu/poster/latent-stratification-n-adv…