We have come up with a preliminary user study design. We have two groups: a control and experiment. Both groups will have to rate answers to various StackOverflow questions.
Control: In this group, all the answers to a particular question will be presented as they are on StackOverflow. These are usually ordered both by the number of votes and chronologically.
Experiment: In this group, only the answers that have a constructiveness score of three or higher are presented. We hypothesize that the user will still learn as much as from the control groups even though the number of posts has decreased, reducing navigation complexity.
Control: In this group, all the answers to a particular question will be presented as they are on StackOverflow. These are usually ordered both by the number of votes and chronologically.
Experiment: In this group, only the answers that have a constructiveness score of three or higher are presented. We hypothesize that the user will still learn as much as from the control groups even though the number of posts has decreased, reducing navigation complexity.