What turns out in practice: there is a Phone Number Database lot of noise. About half of the messages fall into the 'Other' category. Tweets about 40 characters that are not about anything. 10 percent of the messages Phone Number Database are about leisure, 7 percent about sports and media and 5 percent about work. Although Piet sees opportunities to use big data from social media for general statistics, he also states that there is a lot of useless ballast on Twitter. A Disruption Factors: Demographics Erik Tjong Kim Sang ( Meertens Institute ) talks at Phone Number Database the annual meeting about his attempt to predict the outcome of the elections via tweets.
To do this, he searched for Phone Number Database tweets with specific characteristics (eg location). In the end, the mass of tweets he was left with turned out to be too small to make any actual predictions. The biggest disturbing factor Phone Number Database was the fact that the demographics of the Netherlands do not match the demographics on Twitter. At the same time, it can be seen that the age distribution on Twitter has changed Phone Number Database a lot over the past five years. Erik also showed that the number of tweets has decreased, while the number of bots and spammers has increased.
Demographics Twitter versus Phone Number Database the Netherlands Demographics Twitter versus the Netherlands Using Facebook for your survey panel Arnoud Wijnant ( CentERdata ) ends the day with a vision on how you can use Facebook for your survey panel. He sees three levels: Simple support : here you use Facebook Phone Number Database as login for your panel. Advantages of this are: it is simple for the respondent (no extra password usernam safe (more secure than many current protocols), there are many updates and extras (geo-location, you can prevent abuse.