Abstract
This is the fourth article in a series of exchanges with Luis Felipe Miguel about the methodology of Sentiment Analysis (SA). In his last piece, simply entitled An Answer, Miguel argues that SA is employed in the website Manchetômetro as a political intervention and not as academic work. I respond to that objection showing that academic methodologies are currently employed in several initiatives aimed at intervening in public debates. Miguel also claims that SA is not scientific or accurate. I then show that SA has been abundantly employed by Brazilian and English speaking scholars working on media and politics and that the most prominent political science periodicals frequently publish articles that use that methodology. Finally, through a hypothetical experiment I compare SA with frame analysis, a commonly used methodology in Brazilian media studies, to show that the latter must necessarily produce less accurate results than the former.
Keywords:
sentiment analysis; news media and politics; frame analysis; methodology.