People need to understand changing extreme weather events to effectively protect themselves and to adapt to further climatic changes. Extreme event attribution (EEA) science has been claimed to support this. It studies how climate change affects the likelihood, severity and impacts of specific extreme weather events. Drawing on the 'science of science communication', we study, first, the accuracy of peoples' estimates of the probability that a heatwave in 2019 in Germany had with and without climate change, using a correlational survey. We also assess whether these estimates vary inter-individually by climate change risk perceptions, subjective climate change knowledge, heatwave experience, political orientation and numeracy. Second, EEA results often come with different uncertainty types, such as probability ranges, conflicting outcomes or no robust findings at all. We therefore test in a follow-up survey experiment how different uncertainty types in EEA results affect people's estimates of how the probability of the 2019 heatwave changed due to climate change and their perceptions of such EEA results. We will present initial findings from both studies. These are relevant for climate scientists and climate change communicators who want to support people to better understand change in extreme events, while staying true to the underlying science.