Using Big Data to Study Psychological Constructs: Nostalgia on Facebook

Abstract

Nostalgia, reflecting on the past, has many aspects. At times, we refer to it as a state of mind - as in, I’m nostalgic. At times, as a feeling – I’m feeling nostalgic. We sometimes use nostalgia as a way to deal with “in the present” emotions or feelings. For instance, we reminisce about past holidays to feel better about the present [1]. Nostalgia can be internally or externally activated. Internally, an individual can evoke feelings of nostalgia. On the other hand, external stimuli can evoke nostalgia, such as a song, an ad, or a brand [2-5]. Previous research has examined nostalgia. However, these studies have utilized small sample, been clinical in nature, and/or employed research techniques such as, surveys, experiments, focus groups and ethnography. For example, nostalgia studies have examined responses to specific stimuli that invoke nostalgia, such as pictures or music. Ideally, it would be xpedient to explore nostalgia in as natural a setting as possible. However, until the advent of online social media like Facebook, this was not practical. Now, Facebook users generate posts that provide rich data with all sorts of expression and sentiments. Social media posts encompass the entire range of human expression, including sentiment, information, emotion, encouragement, and opinion. The quantity of posts generated can number into the millions per hour. This provides large data sets of content that can be explored using machine learning techniques.