Although there is some performs that inquiries if the step 1% API are random when it comes to tweet context such as for instance hashtags and you will LDA analysis , Facebook keeps that sampling formula try “entirely agnostic to virtually any substantive metadata” in fact it is thus “a good and proportional sign across the mix-sections” . Given that we would not expect any scientific bias to be establish about investigation considering the character of step one% API stream we consider this to be data is an arbitrary take to of your own Myspace inhabitants. We have zero a good priori cause of believing that users tweeting during the aren’t associate of one’s inhabitants so we is also thus pertain inferential statistics and you can relevance screening to evaluate hypotheses regarding whether any differences when considering those with geoservices and you will geotagging let disagree to the people that simply don’t. There will very well be pages that have produced geotagged tweets exactly who commonly obtained on the step one% API load and it will continually be a restriction of any lookup that doesn’t have fun with one hundred% of data and that is an important degree in any lookup with this databases.
Twitter small print prevent united states of publicly sharing new metadata given by the fresh API, hence ‘Dataset1′ and you may ‘Dataset2′ include just the associate ID (which is appropriate) therefore the class i’ve derived: tweet vocabulary, sex, ages and you will NS-SEC. Duplication of the data is conducted thanks to individual scientists using associate IDs to get the latest Fb-put metadata that individuals never display.
Place Attributes versus. Geotagging Personal Tweets
Thinking about all of the users (‘Dataset1′), complete 58.4% (n = 17,539,891) out-of users don’t possess venue features let even though the 41.6% manage (n = 12,480,555), thus showing that every users do not prefer this function. Conversely, the proportion of those towards the function let is higher offered you to pages need choose when you look at the. Whenever excluding retweets (‘Dataset2′) we see you to definitely 96.9% (letter = 23,058166) do not have geotagged tweets on the dataset although the step 3.1% (letter = 731,098) perform. This might be much higher than simply earlier estimates from geotagged content out of around 0.85% as the desire of the study is on the newest ratio from profiles using this feature as opposed to the proportion of tweets. Although not, it is distinguished one no matter if a substantial proportion out-of pages allowed the worldwide function, few up coming proceed to in fact geotag the tweets–hence indicating obviously that permitting metropolises qualities is a required however, maybe not sufficient standing from geotagging.
Gender
Table 1 is a crosstabulation of whether location services are enabled and gender (identified using the method proposed by Sloan et al. 2013 ). Gender could be identified for 11,537,140 individuals (38.4%) and there is a slight preference for males to be less likely to enable the setting than females or users with names classified as unisex. There is a clear discrepancy in the unknown group with a 
disproportionate number of users opting for ‘not enabled’ and as the gender detection algorithm looks for an identifiable first name using a database of over 40,000 names, we may observe that there is an association between users who do not give their first name and do not opt in to location services (such as organisational and business accounts or those conscious of maintaining a level of privacy). When removing the unknowns the relationship between gender and enabling location services is statistically significant (x 2 = 11, 3 df, p<0.001) as is the effect size despite being very small (Cramer's V = 0.008, p<0.001).
Male users are more likely to geotag their tweets then female users, but only by an increase of 0.1%. Users for which the gender is unknown show a lower geotagging rate, but most interesting is the gap between unisex geotaggers and male/female users, which is notably larger for geotagging than for enabling location services. This means that although similar proportions of users with unisex names enabled location services as those with male or female names, they are notably less likely to geotag their tweets than male or female users. When removing unknowns the difference is statistically significant (x 2 = , 2 df, p<0.001) with a small effect size (Cramer's V = 0.011, p<0.001).