Wednesday, November 5, 2014

Census with cellphones

Census scientists have been greeted with an interesting development as they can locate calls through identification of cellphone towers that send and receive them. In turn, call density can be used around the cell towers in the estimation of the local population density. This particular strategy can be attributed to the discovery that 96% of the world’s people have cellphone subscriptions that are active. Additionally, the number of mobile subscribers in developing countries has gone beyond the total population as some individuals have more than one phone while rates related to subscriptions continue to increase reaching heights of about 90%.
            Secondly, through the use of call records, researchers can estimate population density around every tower from call density while taking note of variations in the usage of the phone between areas of high and low coverage. The results are indicative of dynamic population trends across weeks and seasons which cannot be reflected in a traditional survey-based census. An example of this can be drawn from the holiday season where populations in cities dropped while transit sites saw a sharp increment.  However, during the course of a week, weekdays are when people travelled to cities for work and to rural areas during weekends.
            Furthermore, in low income countries, mobile records are representative of an efficient and easy alternative. This is because census data may be outdated or unreliable. Take for example the Democratic Republic of Congo where census was last conducted in 1984. Interestingly, about 70% of the people have mobile subscriptions.
            Another important issue to consider is the recent Ebola outbreak in which cellular records could prove to be a vital tool for population movement tracking. Better access to updated data from affected countries could lead to more accuracy in information on the movement of populations. This in turn would be helpful to governments to coordinate outbreak responses.


No comments:

Post a Comment