For questions 5–8, assume we have a dataset with 1.4 million…
For questions 5–8, assume we have a dataset with 1.4 million rows of the following form: Treatment sales past_sales gender Age saw_ads 1 $30.05 0 female 36 1 0 0 0 female 25 1 1 0 0 female 67 1 0 $7.18 0 female 52 1 0 0 $9.21 female 72 1 1 0 0 female 54 0 0 $17.07 0 female 23 1 … … … … … … Question #8: Suppose you estimate the following experimental differences: -$0.03 ($0.24 st. err.) = [Avg past_sales if Treatment=1] – [Avg past_sales if Treatment=0] -$0.44 ($0.31 st. err.) = [Avg past_sales if Treatment=1 & saw_ads=1] – [Avg past_sales if Treatment=0 & saw_ads=1] Which of the following would be your preferred ad effectiveness estimator?
Read Details