This presentation will report results of the interviews. A structured, open-ended interview protocol was designed as a counterpart to the experiment. The interview included questions that mirror those from the experiment, which allows us to gain insight into the experimental results by revealing how and why people respond to the warning information in the ways that they do, including the reasons underlying the misunderstandings found in the experiment. The interview also included additional questions to reveal how people interpret and use warning information based on the real-world context of their lives and experiences. Interviews were conducted in-person in April 2019 with 36 members of the public who reside in or near Tuscaloosa, Alabama. This presentation will build on the experimental results reported in the first presentation (Part I) by discussing how and why interviewees discussed the perceived likelihood of a tornado hitting different locations, the perceived severity of damage if a tornado hits, and what behavioral responses they would engage in, each for the different warning visuals presented to them. For instance, the results suggest that people’s thinking about potential severity of tornado damage is complex and sophisticated, revealing systematic reasoning that is beyond a simple conflation with perceived likelihood. This presentation will also discuss people’s perceptions of time associated with the tornado threat and their worry about it. Finally, implications for future research will be discussed.