Block 1: Can a face predict an election?
Week 8 · Prediction II · Imai §4.2.1 · about 3 min
NoteHow this works
Run each cell with Run Code, look at the output, then answer the question at the end of the block. Every cell is editable; Start Over resets it. Errors are information: read them.
face.csv: one row = one US Senate/Governor race (2000, 2002, 2004). d.comp / r.comp are competence scores of the Democratic / Republican candidate, based on their faces, rated by participants who did not know the candidates.
1. Load and compute the margin
How many races? diff.share is the Democratic margin (Democratic minus Republican two-party vote share).
2. Scatter plot and correlation
Conclusion: What do the plot and the correlation show?
The correlation is about 0.43: candidates whose faces look more competent do better. Not strong, but surprising for a judgement made in a second. The correlation says how strong, not how large: that needs a slope.