We're nearing the end of data collection for round 2. Once the test is complete please rerun the analysis performed in round 1, but with the topical level breakdown in place of the "evergreen/controversial" breakdown.

The full scope of research questions is covered in previous task, but core questions:

• Any significant difference between the three exposure groups (no test, just the inline and full modal) in their survey responses or event logging engagement?
• Any notable differences between the groups when topics are taken into account?
• Any notable differences between topics independent of the exposure group (ie. all sports articles are more trusted than history, no matter what variant you saw)?
• Any major differences in survey or event logging results between round 1 (evergreen/controversial most read articles) and round 2 (topical most read articles)?