Research:MoodBar/First month of activity
Appearance
From Meta, a Wikimedia project coordination wiki
This page in a nutshell: This report describes an analysis of the first month of operation of MoodBar.
Research questions
[edit ]MoodBar posts
[edit ]- What kind of users signal their mood?
- How many users signal their mood?
- At what point after registering an account do users post their mood?
- What proportion of mood messages is tied to an edit transaction?
- What proportion of mood posts contain a message?
- Does mood type change as a function of user experience at the time when it was posted?
- Can messages sent by MoodBar be easily categorized by type?
- To what extent expressing mood positively affects the retention of new users?
Response to MoodBar posts
[edit ]- How long does it take for a new message to be seen and acted upon by other editors?
- How many new users ever see a response to their MoodBar message?
- How long does it take for a response to a new message to be seen by the sender?
- Does a response to a MoodBar message affect the retention of the user who posted it?
- To what extent the timing of a response affects the chance that the new user will read it?
- Does the timing of a response affect new user retention?
Summary of results
[edit ]Mood by category
[edit ]Data collected during the first month indicates that most feedback submitted by new registered editors is of the "happy" type (58.6%), followed by "confused" (30.9%) and "sad" (10.5%). New editors tend to submit "happy" mood earlier in their user experience than "sad" or "confused" mood.
-
Daily volume of Moodbar feedback by category
-
Proportion of Moodbar feedback by category
-
Distribution of edits of Moodbar senders by category (Log scale)
Mood with comments/transactional mood
[edit ]75% of Moodbar events submitted by users include a non-null text comment, with a higher proportion of comments submitted in the "confused" (84.2%) and "sad" (81.4%) categories. A relatively small proportion of Moodbar events (10-12%) were submitted while editing, suggesting that feedback is not immediately linked to the edit transaction.