Social media data

RESEARCH ETHICS

Utilising social media as research data inevitably involves research ethics questions, challenges, and risks.

As a researcher, your task is to identify the ethical issues and challenges related to social media data and to minimise and manage risks.

Consider the following questions:

• Social media content is generally not intended for research purposes.
  • What risks or harm could arise for individuals from analysing such content?
• What risks or harm could result from processing personal data contained in social media material?
• Could harm occur to you or the individuals if their social media posts are analyzed and highlighted in a public thesis, thereby exposing the posts to a new audience?
• Can you include direct quotes from the material to support your analysis, knowing that the original post and its author may be easily found via search engines?
• Do the posts contain special categories of personal data or other highly sensitive information?
  • Examples of highly sensitive information include: criminal convictions, drug use, financial problems, mental health issues, controversial political opinions and activism.
  • If such information is present, what harm or damage could occur to individuals from being subjects of your research?
  • Avoiding social media material that contains sensitive topics is recommended, as such research may involve significant ethical challenges that cannot be resolved within the scope of a thesis.

These questions, challenges, and risks exist regardless of whether the material is publicly available online or not.

• Please note that researching closed groups involves extensive ethical challenges.
• Even in the case of open groups, it is important to verify whether the platform allows the use of its content for research purposes.
• You must also address other questions related to ethics, personal data, and informing research subjects.

Using social media data often requires preparing a Data Protection Impact Assessment (DPIA).

A DPIA is required especially when the research involves:

• Large-scale processing of data
• Combining or linking datasets
• Collecting special categories of personal data or highly personal information
• Situations where research subjects cannot be informed

More details on DPIA can be found in this educational resource section: Personal Data.

SOCIAL MEDIA CONTENT AND PERSONAL DATA

Social media content almost always includes personal data. In social media, personal data may include, for example, a username, a photograph showing places or people, or a comment containing identifiers.

In other words, when studying social media, you will usually be collecting and processing personal data. 

Informing

According to the current guidelines of the University of Jyväskylä, you must inform the individuals whose personal data you process, even when the data is collected from social media. If it is likely that the research involves commenters or content creators under the age of 15, you must also inform their guardians.

Inform

• For example, if you are studying an influencer’s social media channel, send them a privacy notice, research notification, and consent form.
• If you are studying discussion in a social media group that has an administrator, you should first contact the group’s administrator and ask whether you may post a research notification and a privacy notice in the group. Find out how you can ensure the group members' consent to participate in your research.
• If it is impossible or unreasonably burdensome to contact the individual and inform them personally, make the documents available by other means:
  • Publish the privacy notice.
    • Every JYU student has access to a personal JYU webpage which can be used for this purpose.
  • If you are studying comments on a public social media channel or messages on a discussion forum, link the published privacy notice and research notification in the comments section or message thread.
    • State that you are analysing the comments/messages for your thesis.
    • Indicate when you will start and stop monitoring the discussion.
    • Do not start monitoring immediately - give the participants time to delete their messages if they wish.
    • Include your contact details so that participants can reach you later on and request removal of their messages from the dataset.

Legal basis for personal data processing

A privacy notice must always state the legal basis for processing personal data.

• Information about legal basis can be found in this educational resource section: Personal data.
• You can also read about legal basis in the University of Jyväskylä’s Instructions for students, under the section Legal basis for personal data processing and consent.

Legal basis

• If your thesis is a scientific work, the legal basis is public interest.
• If it is a non-scientific work, the legal basis is usually consent.
  • Remember that consent to participate in research and consent as a legal basis for processing personal data are two different things.
  • If you collect data from social media and it is impossible to obtain consent from participants to take part in the research, then it is equally impossible to ensure their active consent for processing personal data.
• If neither public interest nor consent can be used as a legal basis, the remaining option is legitimate interest.
  • Using legitimate interest as a legal basis requires a Balance test.
  • However, legitimate interest cannot be used as a legal basis if social media data is collected directly from individuals, for example, by asking them questions in a discussion forum.

Tip!

Are you interested in studying online discussions but want to avoid handling personal data from a GDPR perspective? Here are some tips on Finnish language options:

• Posts from the Suomi24 discussion forum are available via Kielipankki (Language Bank of Finland).
  • Kielipankki is an archive that stores various language datasets.
  • Some datasets are publicly available, while others require academic login.
  • Certain dataset versions may contain personal data - this is clearly indicated in the license.
  • The license may include data protection conditions, and your project’s published privacy notice must be provided as a link to Kielipankki.

TERMS OF SERVICE AND COPYRIGHT

Review the social media platform's Terms of Service  / Terms of Use.

• The terms may change, so always check the latest version.
• Find out what the terms say about, for example:
  • Downloading and storing content
  • Publicity or public availability of content (not all material found online is public!)
  • Content sharing
  • Copyright
  • Automated data collection (scraping)

Exception for Text and Data Mining

• Section 13b of the Finnish Copyright Act allows for the reproduction and storage of copies of a work for the purpose of text and data mining. Under this exception, it may in certain cases be possible to deviate from the terms of use of a social media platform.
  • “Text and data mining refers --- to an automated analysis technique aimed at analysing text and data in digital form to generate information.” (Translated from: Tekijänoikeuden tiedotus- ja valvontakeskus ry)
  • Mining may be carried out unless the authors have reserved this right. If data mining is conducted for scientific research within a research or cultural heritage institution, rights holders cannot prohibit the mining.
  • A prerequisite for data mining is lawful access to the work.
• If you are considering not complying with the platform’s terms of use based on this exception, consult your supervisor to ensure that your research methods meet the criteria of the exception.

Texts written on social media may exceed the threshold of originality. In addition, copyright protects images uploaded to social media.

• According to Section 25 of the Finnish Copyright Act, it is permissible to cite an image that has been published, i.e., lawfully made available to the public. Thus, from a copyright perspective, image citation allows the use of a legally published work in a thesis.
  • For example, images from companies’ public social media accounts can generally be considered lawfully published.
  • Images from public social media accounts of public figures fall into a more ambiguous area. So far, there is no legal precedent addressing the right to cite such images in theses or scientific research.
  • Images from private individuals’ private social media accounts cannot be cited. In such cases, copyright issues intersect with ethical and data protection considerations. In principle, when requesting consent to participate in research—or when consent serves as the basis for processing personal data—permission to use images in a thesis may also be requested at the same time as consent.

You can find more information about terms of use and copyright in this educational resource section: Rights and copyright.