User-centric approaches for collecting Facebook data in the ‘post-API age’: experiences from two studies and recommendations for future research

Figures & data

Figure 1. Exemplary data from the browser plugin.

Figure 2. Flowchart illustrating dropout stages in Study 1.

Figure 3. Fluctuations in the Facebook-plugin use.

Figure 4. Screenshot from the Facebook archive download page with the list of options for the DDPs.

Figure 5. Example of the anonymization process (with fake names).

Figure 6. Types of other Facebook activities.

Table 1. Comparison of key features of the two data donation approaches for collecting Facebook data used in our studies.

	BROWSER PLUGIN	DATA DOWNLOAD PACKAGE
Captured Data
Desktop use	✓	✓
Mobile use	X	✓
Exposure to content in news feed	✓	X
User interaction (e.g., likes, comments)	X	✓
Contacts/networks	X	✓
User (profile) information	X	✓
Time period covered	Current use (while the plugin is installed and used)	Past use (possible to capture the whole usage history)
Legal & Ethical Aspects
Compliant with platform ToS	X^1	✓
Possible to obtain informed consent	✓	✓
Transparency for users	Medium	High
Privacy
Control over what is collected/shared	High	Medium
Disclosure risk for participants (based on raw data)	Low	High
Risk of inclusion of data from other private users	Low	High
Deployment & Maintenance
Maintenance effort for researchers	High	Low
Respondent burden	Medium	High
Amount of user support required	Medium to High	Medium to High

Note. ¹ However, a recent court decision in the United States indicated that academic researchers should not incur criminal or civil liability when they violate a platform’s ToS (Sandvig v. Barr, 2020) for their research. Similarly, legal opinions from Germany suggest that non-commercial academic researchers can create databases for their research even if this is not in accordance with platform ToS as long as privacy rights are respected (see, e.g., RatSWD, 2020 for an expert opinion on the subject).

Sandvig v. Barr. (2020). Civil Action No. 16-1368 (JDB) (United States District Court for the District of Columbia March 27, 2020). https://www.aclu.org/sandvig-v-barr-memorandum-opinion.

 Google Scholar

RatSWD [German Data Forum]. (2020). Big data in social, behavioural, and economic sciences: Data access and research data management. RatSWD Output, 4(6). Berlin, German Data Forum (RatSWD). https://doi.org/10.17620/02671.52

 Google Scholar