To summarize, Study 2 examined whether readers of the textual self-description
portion of online dating profiles can detect profile deceptions, and which linguistic
cues influence their decisions. Based on people’s general inability to detect deception,
judges were expected to find it difficult to accurately assess daters’ trustworthiness
and to rely on faulty linguistic cues. These hypotheses were supported. First, judges
performed at chance when trying to classify daters on trustworthiness, consistentwith
previous research documenting humans’ poor deception detection ability. Second,
the only cue used by judges that is in fact related to deception (as established inStudy 1) was word count. The longer a self-description, the more trustworthy its
author was considered to be. In general, the pattern of results support theories
concerned with interpersonal perception, such as uncertainty reduction theory, and
they add to the growing literature about what people find trustworthy in online
contexts.
General discussion
The current studies set out to determine whether the language used by online
daters in their self-descriptions provides information about the deceptiveness of their
profiles. This question was addressed using both computerized linguistic analyses
(Study 1) and human coding (Study 2). Although computerized analyses identified
several linguistic correlates of profile deception, human judges were unable to detect
daters’ trustworthiness based on the written component of the profile, and they relied
mostly on linguistic cues unrelated to profile deception. These two studies advance
theory along several fronts, including the role of technological affordances in the
production of linguistic cues to deception, the relationship between deception and
self-presentation, and how individuals make judgments of trustworthiness in online
contexts.