Limits...
Quantifying regional differences in the length of Twitter messages.

Alis CM, Lim MT, Moat HS, Barchiesi D, Preis T, Bishop SR - PLoS ONE (2015)

Bottom Line: Previous studies of utterances in books, movies and Twitter have shown that most of these utterances, when transcribed, are much shorter than 140 characters.Furthermore, the median length of Twitter messages was found to vary across US states.We find that the median message length, depending on grouping, can differ by up to 2 characters.

View Article: PubMed Central - PubMed

Affiliation: Department of Mathematics, University College London, London, United Kingdom.

ABSTRACT
The increasing usage of social media for conversations, together with the availability of its data to researchers, provides an opportunity to study human conversations on a large scale. Twitter, which allows its users to post messages of up to a limit of 140 characters, is one such social media. Previous studies of utterances in books, movies and Twitter have shown that most of these utterances, when transcribed, are much shorter than 140 characters. Furthermore, the median length of Twitter messages was found to vary across US states. Here, we investigate whether the length of Twitter messages varies across different regions in the UK. We find that the median message length, depending on grouping, can differ by up to 2 characters.

No MeSH data available.


Median message length, in characters, by latitude.Partitioning the tweets into 20 latitude bins of about 105 tweets as in Fig. 3 similarly did not show any North-South division. The median message length per latitude bin measured by characters and by available space are also almost perfectly correlated (Spearman ρ = 0.97, n = 20, p < 0.001).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4390308&req=5

pone.0122278.g006: Median message length, in characters, by latitude.Partitioning the tweets into 20 latitude bins of about 105 tweets as in Fig. 3 similarly did not show any North-South division. The median message length per latitude bin measured by characters and by available space are also almost perfectly correlated (Spearman ρ = 0.97, n = 20, p < 0.001).

Mentions: When measuring the message lengths in terms of the available space, we do not find any obvious North-South division in the districts of the entire UK (Fig. 4a) or even in the boroughs of Greater London (Fig. 4b). The median message lengths of each district measured by characters and by available space are almost perfectly correlated (Spearman ρ = 0.99, n = 404, p < 0.001). Both Northern UK (Fig. 5a) and Northern Great Britain (Fig. 5b) have smaller median message lengths than that of Southern England but longer than that of the Midlands. Unlike in Fig. 2c where the median message length of Northern England is longer than that of Southern England, in Fig. 5c the median message length of Northern England is the same as that of Southern England. Unlike before, the median message length (Fig. 5d) of Scotland is within the same range (36–37% available space) of Wales instead of the Midlands. Binning the tweets into latitudes (Fig. 6) does not reveal any qualitative difference from Fig. 3. The median message length per latitude bin measured by characters and by available space are almost perfectly correlated (Spearman ρ = 0.97, n = 20, p < 0.001).


Quantifying regional differences in the length of Twitter messages.

Alis CM, Lim MT, Moat HS, Barchiesi D, Preis T, Bishop SR - PLoS ONE (2015)

Median message length, in characters, by latitude.Partitioning the tweets into 20 latitude bins of about 105 tweets as in Fig. 3 similarly did not show any North-South division. The median message length per latitude bin measured by characters and by available space are also almost perfectly correlated (Spearman ρ = 0.97, n = 20, p < 0.001).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4390308&req=5

pone.0122278.g006: Median message length, in characters, by latitude.Partitioning the tweets into 20 latitude bins of about 105 tweets as in Fig. 3 similarly did not show any North-South division. The median message length per latitude bin measured by characters and by available space are also almost perfectly correlated (Spearman ρ = 0.97, n = 20, p < 0.001).
Mentions: When measuring the message lengths in terms of the available space, we do not find any obvious North-South division in the districts of the entire UK (Fig. 4a) or even in the boroughs of Greater London (Fig. 4b). The median message lengths of each district measured by characters and by available space are almost perfectly correlated (Spearman ρ = 0.99, n = 404, p < 0.001). Both Northern UK (Fig. 5a) and Northern Great Britain (Fig. 5b) have smaller median message lengths than that of Southern England but longer than that of the Midlands. Unlike in Fig. 2c where the median message length of Northern England is longer than that of Southern England, in Fig. 5c the median message length of Northern England is the same as that of Southern England. Unlike before, the median message length (Fig. 5d) of Scotland is within the same range (36–37% available space) of Wales instead of the Midlands. Binning the tweets into latitudes (Fig. 6) does not reveal any qualitative difference from Fig. 3. The median message length per latitude bin measured by characters and by available space are almost perfectly correlated (Spearman ρ = 0.97, n = 20, p < 0.001).

Bottom Line: Previous studies of utterances in books, movies and Twitter have shown that most of these utterances, when transcribed, are much shorter than 140 characters.Furthermore, the median length of Twitter messages was found to vary across US states.We find that the median message length, depending on grouping, can differ by up to 2 characters.

View Article: PubMed Central - PubMed

Affiliation: Department of Mathematics, University College London, London, United Kingdom.

ABSTRACT
The increasing usage of social media for conversations, together with the availability of its data to researchers, provides an opportunity to study human conversations on a large scale. Twitter, which allows its users to post messages of up to a limit of 140 characters, is one such social media. Previous studies of utterances in books, movies and Twitter have shown that most of these utterances, when transcribed, are much shorter than 140 characters. Furthermore, the median length of Twitter messages was found to vary across US states. Here, we investigate whether the length of Twitter messages varies across different regions in the UK. We find that the median message length, depending on grouping, can differ by up to 2 characters.

No MeSH data available.