"This Line Needs to Remain Available for People Who Are in Crisis": Inappropriate or Ambiguous Messages to a Web-Based Sexual Assault Hotline

Background and Purpose: Sexual assault crisis hotlines, whether telephone or web/text based, provide crucial support for survivors. Practitioners acknowledge that some portion of hotline users engage in offensive or inappropriate conduct, yet little scholarly attention has been paid to examining the content of these hotline interactions and skills that hotline workers use to respond. This paper uses transcripts of hotline chats to explore patterns of inappropriate or ambiguous chats and how hotline responders navigate ambiguous or graphic content.

Methods: The research team was provided 233 de-identified transcripts of chats by a university-based program that oversees a web-based sexual assault hotline. From the original 233 transcripts, a sub-sample of 41 transcripts thought by program staff (n=27) or the research team (n=14) to be potentially inappropriate or ambiguous were further analyzed. Transcripts were coded by three researchers using Dedoose and guided by the principles of thematic analysis.

Findings: Exploring the range of inappropriate or ambiguous chats led to the generation of three themes—graphic transcripts, repeat inappropriate chatters, and ambiguous chats. Graphic chats included graphic sexual or violent content, often evoking disturbing images in ways that clearly distinguished them from survivor narratives. Repeat inappropriate chatters were characterized by numerous chats that featured nearly identical topics and phrasing that diverged from typical help-seeking patterns. Ambiguous chatters used slang (i.e., “dick”) or referred to specific sex acts (i.e., “edging”) in ways that made them particularly subject to interpretation and misinterpretation.

The team then explored the process through which responders and chatters navigated chats. Responder replies seemed to depend on their interpretation of the conversation and the severity of the chatter’s boundary violation. Ambiguous and less egregious boundary violations were navigated through soft redirects, though the extent to which soft redirects invited further conversation differed. For example, discussions of mandatory reporting without noting potential impacts on the chatter often shut down further conversation. Responders sometimes set firm, direct boundaries when facing more egregious boundary violations by reiterating the chat’s purpose, calling chatter behavior unacceptable, or informing chatters of potential consequences for their actions. Chatters responses to boundary-setting provided insight into boundary-setting efficacy. Chatters ending conversations after boundary-setting suggested responder success in deterring further inappropriate conversation. Other times chatters prolonged conversations by attempting to re-engage responders, arouse sympathy, or escalate their inappropriate behavior.

Conclusions and Implications: Findings highlight the ambiguities and challenges faced by responders to web/text based sexual assault helplines. Responses provide examples of effective and ineffective techniques that could be implemented by other providers of crisis services. Analyzed conversations suggest the need for more support, training, preparation, and debriefing protocols for responders to web/text based sexual assault hotlines. Additional work to understand potentially inappropriate or ambiguous chats could help providers improve services offered to survivors and reduce emotional challenges for responders while optimizing chat efficiency and effectiveness.