amarchenko26 / BA-Draft

2 stars 3 forks source link

Table 4 is DONE! #39

Closed ThiagoSchmidtResende closed 5 years ago

ThiagoSchmidtResende commented 5 years ago

(1) Edited table 4 top match the style of other summary tables. (2) Discovered why the observation numbers were different between old and new tables: the old table described the full data while the new one partially describes the full data but focuses on the data used in the regressions. NOTE: The review sentiment averages seem very similar. This is not a coding mistake. They are simply very similar between the full and reduced dta sets. (3) The previous merge branch commit was a minor merge error which I fixed on my end.