Four best practices for measuring news sentiment using ‘off-the-shelf’ dictionaries: a large-scale p-hacking experiment

We examined the validity of 37 sentiment scores based on dictionary-based methods using a large news corpus and demonstrated risk generating spectrum results with different levels statistical significance by presenting an analysis relationships between U.S. presidential approval. summarize our findings into four best practices: 1) use suitable dictionary; 2) do not assume that reliability dictionary is ‘built-in’; 3) check for influence content length 4) multiple dictionaries to test same hypothesis.