For people who look for a partner, online dating largely increases the pool of potential mates. At the same time, users of online dating platforms have to cope with a large number of approaches and, therefore, need to choose selectively who they decide to engage in a conversation with. Especially, since the costs of rejection are low on online dating platforms, it is a common strategy to spam others with superficial approaches. With this in mind, and in the absence of nonverbal cues, targets base their decision of whether or not to respond to a message on (a) their impression of the sender’s pictures, and (b) cues which they extract from the content of the message. The purpose of this study is to hypothesize on which linguistic properties of a message in computer-mediated communication may signal various qualities of its sender, to predict how those properties determine a target’s decision of whether to respond or to ignore an initial message. Employing the Linguistic Inquiry and Word Count (LIWC) text analysis, relevant variables are operationalized from a corpus of 167,276 initial messages of an online dating platform. Regression analysis is performed in order to test the hypotheses. Results are discussed with respect to design implications for online dating platforms.


Computer-mediated Communication, Online Dating, Text Analysis


ISBN: [978-1-86435-644-1]; Full paper