Privacy and ethics in social media research

Manfred Mareck

Best practice to preserve consumer trust and researchers’ reputations.

User-generated content has become ubiquitous and it is not surprising that marketers are focusing more attention on social media as an online channel to communicate with consumers and to commission market research to gain insight into consumers’ opinions, attitudes or behaviour. This has created new opportunities for researchers to observe, interact and gather information. It has also led to new techniques including the use of community panels, co-creation, netnography, blog-mining and web scraping.

Last year things went horribly wrong when Nielsen’s BuzzMetrics were found to be systematically scraping comments from a forum on PatientsLikeMe.com, where registered members exchanged their experiences of medical and emotional problems, clearly assuming they did so in a private space.

According to Matt Anchin, Senior VP of Global Communications, Nielsen stopped the practice immediately upon being contacted by PatientsLikeMe.com. Months later, when the story broke in the Wall Street Journal, The Nielsen Company quickly issued a mea culpa. “We no longer include data beyond what is freely available to anyone online via search engines or through normal browsing from any website requiring establishment of an individual account for access without first obtaining consent from the website operator.” According to Anchin, BuzzMetrics has established data usage agreements with a number of websites that require some form of login “and we will continue to add others so that we can carry on to provide clients with the most robust data coverage in the industry”.

That researchers should ask the website’s permission is not a new concept – after all, researchers would first ask a store’s permission if conducting ethnography in that store. Online however, it is also connected with copyright issues as most websites hold the copyright to content generated on that website.

New guidelines
Social Media Research (SMR) is an emerging area of research but it should be conducted within the ICC/ESOMAR Code that governs other forms of research. “If you read the Code and the passive observation guidelines it is pretty clear what researchers should or should not do. Nevertheless, because SMR is different from other kinds of research, we felt it would be helpful to explain how researchers can stay within the requirements of the Code”, commented Adam Phillips, who chairs ESOMAR’s SMR project team and its Professional Standards Committee. “

“Companies are still finding their way around how best to use the social online space. In addition, some of the new players that are exclusively engaged in SMR have been less closely involved with established codes and guidelines so it was felt that some guidance would be beneficial” says Pete Comley of Virtual Surveys and a member of the project team. “Not everything in the social media sphere is done by market researchers and not everybody feels the need to work within specific guidelines. Market researchers who are still experimenting with SMR may be tempted to take a more laissez-faire approach”.

Upcoming legislation
Another reason to develop guidelines is to avoid inappropriate privacy legislation and to defend the freedom to conduct research activities. Even in the US, the traditional free-for-all philosophy of self-regulation is coming under pressure. Peter Milla, who advises both ESOMAR and CASRO, lists the recent Federal Trade Commission Privacy Report and the Best Practices Bill by Congressman Bobby Rush as examples of impending legislation. “This could lead to a comprehensive privacy framework, which, at least in part, is modelled on the EU Data Privacy Directive. The potential outcome is indeed more regulation in the US and if large numbers of individual citizens support the idea of a ‘do-not-track list’ even the most pro-business politician is unlikely to take a hard stand against it”.

The market research industry has a vested interest to show law makers and consumers that it is already addressing these sensitive issues. “We want to draw attention to potential problems, to get a discussion going and agreement about what is and is not permissible to encourage researchers to move in a sensible direction. Importantly, we want to avoid upsetting people and to keep the public as potential respondents on our side so we can continue with this kind of research, which can be extremely useful to our clients” says Phillips.

ESOMAR’s guidelines cover a range of topics, including protection of identifiable data and conforming to existing regulations on data collection from children and young people. The focus is on defining the social media space and matching participants’ expectations with recommendations for best research practices in each space.

The social media space
The forthcoming guidelines distinguish between four different spaces in the online world

In the public space content is contributed with the expectation that it will be read by anyone in the public and where contributors are likely to be happy that it is linked to, copied or cited, such as public blogs and comments left on news websites. Research can be carried out subject to the site’s terms of use; users can be identified and quoted unless this might cause harm, in which case quotes must be ‘cloaked’, that is edited so they can’t be retraced via a search engine but without losing the essence of their meaning.

In semi-public spaces people contribute content where, although open to all to read, many would not expect it to be read or used by people not interested in that topic or conversation. Research can be carried out subject to the site’s terms of use but more cloaking may be required. The boundary between public and semi-public space may often be blurred and if in doubt researchers are encouraged to treat sites as ‘semi-public’. Examples include Facebook (homepages and many niche (but open) forums or communities), open chat rooms and Twitter.

Private space is a place where users would expect their comments to be private and available only to genuine community members. Often called ‘walled gardens’, they require registration and passwords and include private forums, communities, chat rooms and instant messaging. Researchers require permission from the site owner to carry out work and users cannot be identified without their prior consent. Cloaking is essential unless users’ permission for verbatim has been obtained.

Market research spaces have been created for market, social and opinion research purposes where users have been specifically informed of its function and the use to which their comments might be put. Typically (but not always) these are also private spaces and include Market Research Online Communities (MROC’s) and online ethnographic and co-creational techniques which utilise social media platforms. Due to the nature of these sites researchers face relatively few restrictions but some degree of cloaking should still be used if verbatim quotes could cause harm to respondents.

Fundamental principles
Fundamental principles of all market, social and opinion research are the ‘protection of identifiable data’ (ensuring that personal data is only used for research purposes); the protection of ‘respondent privacy’ (ensuring respondents are not adversely affected as a result of research); and that respondents give their ‘informed consent’ to their participation. Depending on the space in which a research project takes place, the guidelines recommend different specific steps to comply with these principles.

Protecting online users’ identity is not always easy, because search engines can be used to trace an author of verbatim quotes and services such as PeakYou can identify individuals, even when using an assumed online username. In addition, at least in the US, PeakYou links into public data bases where for a few dollars addresses, credit ratings and criminal records can be obtained. Identifying the specific site, forum or blog from which a comment is quoted in the survey report can also help identify individual users.

“Most consumers are aware that their online conversations could be monitored but a small percentage may not understand that this is the case. A small percent of two billion online users is a lot of people who would be surprised, and possibly embarrassed or offended, that their information is being shared in an arena outside of what they originally intended”, says Annie Pettit, chief research officer at Toronto-based Conversition Strategies and member of the guideline project team. “Whilst ultimately consumers should always protect themselves this is not the ethical standard that market researchers can align themselves with and we must always work diligently on the contributors’ behalf.”

Adam Phillips chairs ESOMAR’s SMR project team, Matt Anchin is Senior VP of Global Communications at Nielsen, Pete Comley is the chairman of Virtual Surveys and member of the ESOMAR’s SMR project team, Peter Milla, who advises both ESOMAR and CASRO on privacy issues, Annie Pettit is chief research officer at Conversition Strategies.

Manfred Mareck is an independent media and marketing consultant

ESOMAR Guideline on Social Media Research
The guidance is intended to ensure that researchers uphold the highest ethical and professional standards but not to cover methodological standards which are is still developing. It is ESOMAR’s aim to cooperate with associations around the world to reach consensus on internationally agreed best practice and ESOMAR is already working closely with CASRO in this regard. A consultation has been launched on which all members and associations will be invited to comment. For more details go to http://www.esomar.org/index.php/professional-standards.html