Curiel et al (2020) published a report on an analysis involving over 32 million tweets "from within the 18 Spanish-speaking countries from Latin America were initially retained, while retweets were not considered."
The degree to which we can learn from social media is highly controversial. Understanding 'fake news' can have a direct influence on public health measures and thus the safety of many people. In the context of forensic psychology, it can be useful to understand where people are now getting their news and facts about crime and how this information differs from the statistics in reality.
Some original notes on the method:
Of the 32,513,684 tweets:
27% of the tweets were from Mexico
23 % from Argentina
12% from Columbia
The rest came from smaller countries
The researchers utilized hashtags and keywords to classify the tweet.
Findings:
As with regular media such as print, social media was heavily biased towards specific crimes. In the tweets collected - 28.3% were related to murder though murder in Mexico murder accounted for 0.072% of crime. This can be shown to be similar in newspapers in the USA where 30% of the crime stories were related to murder when murder accounts for 0.02% of crime.
Violent deaths such as homicide, suicide and terrorism are overrepresented in both social media and regular media.
Also, social media reported crime is not useful for assessing trends as the tweets often represented events that happened months before.
To be newsworthy, social events must capture the attention of the viewer/reader and so they have to be rare, or timely, or unexpected, or alternatively have some special significance (Chermak and Gruenewald, 2006 as cited in Curiel et al, 2020)
Link to paper:
https://www.nature.com/articles/s41599-020-0430-7