Here’s the latest thinking on the goal for the first experiment on reactions after lengthy discussion with @sydney. We want to figure out if there’s a significant effect of different kinds of reactions—symbolic vs. text-based—on people’s perception of a post’s content.
Within the experiment, there are four groups, each with a set of screenshots of conversation spaces with posts in these contexts:
- Just the post (control)
- The post with text reactions in response (simple responses such as lol, hah, yes, sure, etc.)
- The post with emoji reactions in response (the most basic emoji counterparts to the text reactions)
- The post with emoji reactions attached to the post (like Discord/Slack do it)
In each category the task is to rate different dimensions of a specific posts’ qualities on a scale. Each category provides a different kind of reaction style context that may or may not have an effect.
Some things to sort out:
- What are the ramifications of using these different approaches to reactions? It might be good to hypothesize outcomes and the resulting potential interface/system designs.
- What are the different dimensions being rated? In other words, what are the judgements we are asking people to make about the posts?
- How do different kinds of emotes have an effect? Does the 👍 have a significantly different effect than the ✅? This is probably best expanded into its own experiment.