The multimodal main format of the TikTok channel @energiewende.erklaert, in which the communicating researcher answers questions about the energy transition, requires her to read and approve comments under her own videos. Verbal attacks in the form of polarising comments are not only a burden, but also influence the performance of the communicating researcher and her way of doing science communication. These attacks will be analysed and evaluated with the help of a qualitative data analysis of the comments. In order to analyse the comments comprehensively, to create comparable conditions and to give the researcher self-efficacy, a methodology is developed in which the participation and separation of the communicating researcher is weighted and balanced differently depending on the process step. It is shown that the participation of the communicating researcher in the research process offers advantages, for example in the case of access restrictions by the platform and through her specialist expertise. At the same time, the biases and distortions caused by self-research must and can be minimised through transparency and a clear division of responsibilities, especially in qualitative analysis and evaluation.