A reflection on project 3
This post addresses the following prompts:
Outline your project. What would you do differently? What was the most difficult part for you? What are your big take-aways from this project?
Project 3 went more smoothly than the previous two for me, as I definitely took my reflections and takeaways from the first two into account when working through this one. Yi and I started work at least two weeks before the due date and both worked on it regularly. It was an excellent partnership in my opinion and we utilized both Slack and Github to regularly check in, provide feedback, and plan out next steps. It was definitely the most collaboration I have experienced on a group assignment thus far in the Masters of Statistics program.
The most challenging part of this assignment for me personally was balancing study time for the exam with the urge to use my time to move forward with the project. As for the project content itself, I found two aspects to be of particular difficulty… The first was that none of the predictors had a strong correlation with number of shares, which made it difficult to determine which predictors to use for a regression model. In the end, we decided to use two different variable selection techniques for our linear regression models instead of testing out models with different interaction terms, higher order terms, etc. The other was the automation aspect of the project. Luckily for me, Yi felt more confident with this task and took the lead there. However, she made sure to answer my questions and to include me in the process - I definitely feel like I could make a decent start independently on a similar task in the future.
Overall, I feel like I have substantially grown several skills throughout the past few weeks working on this assignment. I finally got a good grasp on using Github and Git to make updates on the assignment and to collaborate effectively with another team member. I also feel much more confident in modeling and making predictions.