Data available through API may be limited IRBs may not fully understand Twitter research Need to ﬁlter data for desired users (e.g., random, by language, type of user) Many studies violate the Twitter TOS
should we report measures taken to account for limitations in data from the Twitter API? 2. What are best practices for sampling and ﬁltering data and for reporting these processes? 3. How can we account for violations of the Twitter TOS when conducting our research? 4. What are best practices for Twitter privacy research when considering the ethics of our research?