Tweets for mood tracking? Research participants sought

Status updates have been used to examine population mood trends and how they correlate with historical events and economic trends. Can such analysis support emotional-awareness of individuals? AffectCheck, a prototype app for Twitter (that uses Firefox add-on), color-codes Tweets by affective tone. The point of this post is to invite research participants(https://affectcheck.org) and to discuss the potential for mood tracking and feedback based on analysis of status updates and other writing.

Interesting, Margie! Is it only for Firefox, or do you have a version for Chrome? Couldn’t you also just analyze people’s tweets without it being an add-on, by looking at their publicly available, historical tweet streams?

Thank you Alexandra for your question. I’m one of the developers of AffectCheck.

Currently, AffectCheck is only for Firefox. We do not have a Chrome version yet.
The tool is designed to analyze and color-code tweets as they are written, including tweets that you write, but decide NOT to publish. So we can’t rely on historical tweet streams. Since we need to color-code words in tweets as they are typed on Twitter.com, we needed to embed code onto the Twitter.com webpage. To do this, we decided to use a Firefox add-on.

Other tools, like HappyTweets, attempt to analyze the history of your tweets and give you an overall measure of the happiness of your tweets. These tools do not provide any real-time feedback on the affective tone of the tweet you are currently writing. AffectCheck does provide real-time feedback.

AffectCheck also differs from other tools, in that it allows you to personalize the meaning of positivity and negativity. You can click on any word to change its positivity or negativity, and all future tweets will use your personal classification to color-code your tweets.

I see, so it measures the mood of your tweet before you publish it - interesting. I wonder if that would make people self-censor and change their language before posting. Do you measure changed tweets as well, like if I edit a tweet 5 times with different emotion words, are they all recorded? And I’m curious, what is the purpose of the study?

No answer to “what is the purpose of the study?” This sounds like it is not privacy sensitive a.k.a. big brother is watching your tweets.