Optimising cohort data in Europe

transformed the way people communicate and socialise on the web, becoming more powerful as they grow since they feed off interactions among people. It can be argued that social networking shifts the balance of power in favour of social media users, partly because it increases the visibility of issues and concerns of some sectors of the general population. To some extent, this power has shifted to those categories of the population who are active on social media, i.e. predominantly younger people and organisations, lobbyists, parties ..., with social media as a platform for activists and campaigners of any kind (Liang and Zhu, 2017). In the field of health care research, SNS have been primarily focused as tools to deliver health care, to study the effect of these networks on behaviour, or to analyse Web health content. Also, it has been shown that SNS is an effective recruitment method but it may introduce a sampling bias (Alshaikn et al., 2014). Table 2 shows the strengths and weaknesses of SNS. Table 2. Strengths and weaknesses of social media data collection

Strengths

Weaknesses

− The demographic and psychological variables of users may not be known, impacting the quality of the data collected. − Users generating the information are often anonymous or use avatars. − Self-reported data may introduce self-selection bias, sampling bias or other generalisability/ reliability issues. − Data is analysed by researchers with very different backgrounds. − Collecting Big Data requires a new set of methods, as well as careful preparation with regard to compliance to General Data Protection Regulation (GDPR). − Bias related to users’ socio-demographic or socio economic characteristics (parts of societies do not use SNS at all).

− An environment where sharing information, knowledge, interest and opinion is not only meaningful and fun. − Data is usually fine-grained, real time and on a global scale. − Data can be collected automatically and unobtrusively, and without affecting social behaviour. − Increasing use has generated rich

information online (useful for recruiting and follow up cohort participants).

2.3. Passive data collection Passive data collection occurs without any overt consumer interaction and generally includes capturing user preferences and usage behaviour, including location data, from personal mobile devices. A well-known example is the use of cookies on a user’s computer to capture Internet browsing history. It is traditionally used for market or political research. Consumer data are automatically collected through their behaviour and interaction without actively notifying or asking the consumer’s permission. This