Identifying Random Samples of TikTok Users

Date Issued

Journal Title

Journal ISSN

Volume Title

Publisher

Northern Kentucky University

Abstract

"As the TikTok social media platform rapidly grows in popularity, the question was asked on how to obtain a random sample of TikTok users indicative of the general population of users. This research tests two different methods for obtaining a random sample of users. An analysis determined the uniqueness of the samples retrieved using a comparison of the fan counts for each user obtained. The first method utilizes usernames as seeds for producing suggested users while the second method retrieves a common datapoint, such as a hashtag, and locates users of the common datapoint. The findings revealed that suggested user crawlers, when starting with trending TikTok usernames, only revealed users with fan counts in the millions. Even iterating the suggested user list as the new seeds maintained a nearly unchanged list of users in later iterations. The hashtag discovery method revealed a diverse range of users from single digit fan counts and up to millions of fans. It was determined that, based on the diversity of fan counts achieved, the discovering hashtag method is the best method for obtaining a random sample of users."

Description

2021 Celebration of Student Research and Creativity presentation

Keywords

Sampling (Statistics), Social media, Hashtags (Metatdata)

Citation

DOI