Finally: An analytical study of "geeks" versus "nerds"

Software engineer Burr Settles from Pittsburgh studied the language used in 2.6 million tweets, and also sampled tweets that appeared when he searched for the terms ‘geek’ and ‘nerd’.

By comparing tweets in each query, Settles devised a mathematical equation that established the probability of a particular word appearing in a geeky tweet, or a nerdy one.

For example, the most nerdy subjects revolved around the words ‘cellist’, ‘neuroscience’, ‘goths’ and ‘gamer.’

On the geekier end of the scale, words included ‘culture’, ‘shiny’ ‘trendy’ and ‘webcomic’.