Online lyric databases are skewed toward contemporary rappers, biasing hip hop data analysis

Evidence

I am now drawing on data generously provided to me by Ian Lin, a software engineer and hip hop fan who published an article in Cuepoint that uses data visualization to compare various metrics in hip hop, including the popularity of car brands. His data partially comes from Genius, the previously mentioned website that operates “Rap Stats,” as well as Discogs, an online depository of music release information, and other sources. A clear limitation Ian points to is that contemporary rappers and their lyrics are far more prevalent in the available online data than rappers from the 1980s and 1990s. That is partly due to there just being more rap music produced in the past decade, but also, contributors to online databases are more likely to contribute contemporary data. — Andrew deWaard, Derivative Media, 873

Categorization

• QUESTION: A methodological caution directly relevant to whether the book can demonstrate causation: findings built on Genius/Discogs data systematically underrepresent pre-streaming catalogs, which may inflate apparent post-consolidation shifts.