Answer 2 and Question 3¶
Find which advertisement campaign is the most effective for boosting product sales
Inference
Why? You are NOT predicting product sales. You are estimating the impact of an advertisement campaign on product sales. In an ideal world, to know for sure how much the advertisement campaign brought extra sales, you need to be able to observe two counterfactuals: the sales without the campaign (control world) and the sales with the campaign (treatment world). So you are basically interested in estimating \(\beta^A\) of the advertisement campaign A on product sales. Then, you want to estimate \(\beta^B\) of the advertisement campaign B. And then, you want to estimate the difference \(\beta^A-\beta^B\) and test the null hypothesis that \(\beta^A-\beta^B = 0\).
Question 3¶
Find the price elasticity of demand for a product
Is it a prediction task or an inference task?
(See the answer in the next section)
Click the next section for another question.