UC Santa Barbara

The study of self-effects—the effects that sending messages have on the message senders themselves—has become a prominent research topic within a social media context, partly because social media are relatively unique, compared to other types of media, in allowing individuals to send messages to large numbers of people. This dissertation adopts a distributed cognition perspective to explore how digital technology complements human cognition to engender self-effects. More specifically, it proposes a digital episodic memory framework to analyze the influence of mobile phone photography and social media during the encoding, storage, and retrieval of personal experiences within a self-effects paradigm. Factors pertaining to each memory stage are identified, including self-relevance and engagement during encoding; publicness, ephemerality, valence, volition, and approval motivation during storage; and perceived social approval and presumed audience knowledge during retrieval. Through the hypothesized processes of commitment (desire to be consistent with a particular image of oneself) and capitalization (sharing positive events with others), these factors are predicted to influence message senders’ subjective memory of personally-experienced events in addition to their recalled enjoyment of and emotions during those events. In the three studies described herein, encoding factors and retrieval factors are measured, while storage factors are either manipulated (in experiments) or randomly chosen to be further probed (in a survey). The studies require participants to take photographs during certain personal experiences (some experiences are researcher-assigned, others are of participants’ own choosing) and, if instructed, to share the photos on their actual social media accounts. Study 1 manipulates publicness (share photos publicly on social media/keep them private) and ephemerality (photos are persistent/photos eventually disappear), focusing on whether the public sharing (vs. private storage) of past experiences can lead to stronger self-rated subjective memory, and by extension permit an inference whether people show greater commitment to what they share about themselves. Study 2 manipulates valence (write positive/neutral captions to accompany photos) and volition (participants write their own captions/receive instructions to write specific captions), focusing on whether experiences that are publicly portrayed as positive (vs. neutral) can foster stronger subjective memory, greater enjoyment, more positive emotion, less negative emotion, and better self-esteem. In both Study 1 and Study 2, the self-relevance of and engagement during the photographed personal experience are tested as predictors of the aforementioned outcomes. Both studies also have a longitudinal component: Participants retrieve (i.e., look at) their photos after at least three days, having given their social media contacts some time to respond to the photos they share. Perceived social approval and presumed audience knowledge (the extent to which one believes that other people have seen and therefore know about one’s publicly shared photos) are measured, and are then used to predict the same set of outcomes. Study 3 is designed to test several important mediators that bridge the relationships between publicness and subjective memory, enjoyment, or emotion. Commitment is measured (instead of being inferred in Study 1) and tested as a mediator, alongside perceived social approval and presumed audience knowledge by close others. As a whole, the findings show that the factors associated with the encoding and retrieval stages of digital episodic memory are better predictors of the various cognitive and affective outcomes than the factors associated with the storage stage. Results demonstrate that the digital episodic memory framework illuminates the psychological and social dynamics pertaining to self-effects. Segmenting the process of creating/sending messages into the encoding, storage, and retrieval stages of memory highlights important factors that are typically not considered in self-effects research.