Is it Really About Me? Message Content in Social Awareness Streams
Citation
Mor Naaman, Jeffrey Boase, and Chih-Hui Lai. Is it Really About Me? Message Content in Social Awareness Streams.
Online version
Summary
This paper uses quantitative (statistical) methods to analyze the Twitter use postings in a qualitative manner. The authors used human coders to categorize the Twitter user feeds into 9 categories, and answered the research questions such as: what are the most common message types of Twitter user? What these type of messages say about the user him/herself? How are these differences in users' content practices related to other user characteristics?
Method used
Each tweet is examined by exactly two human coder. The coder can assign multiple categories to a tweet, and the union of the categories from each coder is considered as the final category for the tweet. The categories are:
- Information Sharing (IS) - Self Promotion (SP) - Opinions/Complaints (OC) - Statements and Random Thoughts (RT) - Me now (ME): talking about the user's own feelings - Question to followers (QF) - Presence Maintenance (PM) - Anecdote (me) (AM) - Anecdote (other) (AO)
Over-coding was not a problem as messages had 1.3 categories assigned on average.
Experiment results
The top-four most frequent categories for the tweets are: Information Sharing, Opinions/Complaints, Statements, and "ME now".
Most people engage in at least some activities that talks about him/herself, while only a few undertake information sharing as a major activity; and female users are more likely to talk about themselves than male users (which follows the common sense).
By using Ward's linkage cluster analysis, the authors categorize users based on the types of messages they typically post. The resulting clusters are: "Informers" (20% of the user) and "Meformers" (80% of the user).
On average, "Informers" have more friends and followers than "Meformers", and "Informers" also have a higher proportion of mentions of other users in their messages.
Data set
The authors did their analysis from the Twitter data crawled in three weeks.
