5 The future of ipseology
5.1 The demise and legacy of the Twitter 1% stream
The main source of ipeological data had been Twitter. Specifically, most of my datasets were derived from the stream of a 1% random sample of all tweets. That has disappeared. But all is not lost:
- There is still plenty of science to be done with a decade-plus worth of data from 450 million global users.
- In 2024 and beyond, I am still observing hundreds of thousands of Twitter bios and their edits for users in the US and elsewhere.
- With modest funding, we could collect self-authored self-descriptions through representative-sample surveys.
With students, I am currently conducting further analysis pertaining to each of the following sections.
5.2 What is a bio revision event?
In ipseology, a bio revision event occurs when one observes a change over time in the set of signifiers an individual chooses to include in their self-description. Much work remains to be done examining Add and Delete events - i.e. which words do individuals add or remove, and when do they do so? Which words predict the addition or removal of which others, and how has that changed over time?
5.3 What is an identity alloy?
In ipseology, an identity alloy is the mixture of two elements of identity. For example, in the bio Father of two and prototypical Scorpio, father and scorpio form an identity alloy. Every pair of signifiers is an identity alloy. Some alloys are observed more frequently than others.
5.4 What is an identity transmutation?
In ipseology, an identity transmutation occurs when an an individual stops using one signifier to describe themselves and starts using a different signifier. For example, if I edit my bio from Scientist who studies identity to Ipseologist who studies identity, then I have transmuted from scientist to ipseologist.
5.5 New directions for ipseology
As computational social scientists and ipseologists, we need identity data at scale. I suggest two paths to new data streams. The first is traditional surveys. Scales that elicit personally expressed identity text are sometimes referred to as Who am I? instruments. Instruments like these can be delivered as short surveys. For a low investment of research expenditure ($1000) one could administer an instrument such as the Twenty Statements Test to a representative sample of a few hundred respondents.
Second, I urge the development of web and phone apps to collect self-authored, self-descriptive text from longitudinal panels. An app might simply collect periodic bios from volunteer citizen scientists. A more ambitious approach would be to build an app that provides value for users – perhaps feedback or accountability on goals for personal growth – in exchange for use of personally expressed identity text in research.