1

I have a dataset that contains historical prices of a stock and number of mentions of the stock in the news. I want to know what is the best statistical way of deducting the relation (if any) between the two variables. And how to formulate this relation?

The prices are in $ and the number of mentions go like 1,2,3,4, both over time.

I thought about probability, regression analysis and hypothesis testing. Are there other better ways?

Amr Jamal
  • 11
  • 3
  • It's an iterative process, right? You try linear regression, and see if the line fits. That's easy. If it doesn't, you can try a quadratic curve or a cubic curve. – Tim Roberts Mar 04 '21 at 20:01

0 Answers0