xgao28 / election_forecast

1 stars 0 forks source link

Peer Review by Steven (Group 94) #2

Closed stevenli-uoft closed 3 weeks ago

stevenli-uoft commented 1 month ago

Summary:

This paper presents an analysis of forecasting the 2024 US Presidential Election using polling data. The authors develop a multi-linear regression model using various polling metrics to predict support for Kamala Harris or Donald Trump. The paper includes detailed appendices on YouGov's methodology and a proposed survey design.

Strong positive points:

Critical improvements needed:

Suggestions for improvement:

Evaluation:

R is appropriately cited: 1/1 Data are appropriately cited: 1/1 Class paper: 1/1 LLM usage documented: 1/1 Title: 2/2 Author, date, and repo: 2/2 Abstract: 3/4 (clear but could better emphasize significance) Introduction: 3/4 Estimand: 1/1 Data: 8/10 (good coverage but could use more visualization) Measurement: 3/4 Model: 6/10 Results: 4/10 Discussion: 4/10 (Could improve on conciseness) Prose: 2/6 Cross-references: 0/1 (no in-text citation) Captions: 0/2 (figures have to caption) Graphs/tables/etc: 1/4 (only one figure in paper) Idealized methodology: 8/10 Idealized survey: 4/4 Pollster methodology: 8/10 Referencing: 0/4 (no citation or references) Commits: 1/2 Sketches: 0/2 (no sketches in repo) Simulation: 0/4 (not done) Tests-simulation: 0/4 (not done) Tests-actual: 0/4 (not done) Parquet: 0/1 (not implemented) Reproducible workflow: 2/4 (Could add more detail to the readme for easier understanding) Miscellaneous: 1/3

Estimated overall mark: 67/126

Any other comments:

The main areas for improvement are in the technical implementation aspects (testing, simulation), and improving the paper in terms of figures/tables, in-text citation, and cross-references.

xgao28 commented 3 weeks ago

Thank you! We added citations and fixed the long table for visualizations.