README
------
The files in this module represents all data anlyzed to produce the statistics and figures for *"An experimental study of team size and performance on a complex task"*.
These are captured in three files:
- **results_nominal.csv** The aggregated statistics for each group that participated in the crisis mapping experiment of team size. *(Fig. 5, 6, 7, 8)*
- **results_0.75.csv** The same statistics for each group, calculated 3/4 of the way through the session
- **results_0.5.csv** The same statistics for each group, calculated 1/2 of the way through the session
- **results_0.25.csv** The same statistics for each group, calculated 1/4 of the way through the session
- **synthetic.csv** The simulated data generated for the synthetic groups composed of (effectively) independent workers. *(Fig. 9)*
----------
**results_nominal.csv** and **results_0.75.csv**
- **age**: participant's age
- **gender**: participant's gender
- **instanceId**: identifier for participant-session
- **dropped**: whether the participant left the session
- **tutorialMins**: how long the participant spent on the tutorial
- **tutorialWords**: how many words the participant wrote in the tutorial session
- **exitSurveyWords**: how many words the participant wrote in the exit survey
- **effort**: the total effort by the participant
- **entropy**: the entropy over actions types for the participant
- **time**: total time spent by the participant
- **normalizedEffort**: effort normalized by time spent
- **chatFrac**: fraction of time spent chatting
- **chatWeight**: weighted proportion of time spent chatting
- **classifyFrac**: fraction of time spent classifying content
- **classifyWeight**: weighted proportion of time spent classifying content
- **filterFrac**: fraction of time spent filtering content
- **filterWeight**: weighted proportion of time spent filtering content
- **verifyFrac**: fraction of time spent verifying content
- **verifyWeight**: weighted proportion of time spent filtering content
- **groupEffortFrac**: the proportion of the entire group's effort contributed by this individual
- **groupChatFrac**: the proportion of the entire group's chatting contributed by this individual
- **groupClassifyFrac**: the proportion of the entire group's classification contributed by this individual
- **groupFilterFrac**: the proportion of the entire group's filtering contributed by this individual
- **groupVerifyFrac**: the proportion of the entire group's verification contributed by this individual
*The following are group-level variables (i.e., constant for all group members)*
- **g_nominalSize**: the intended size of the group
- **g_wallTime**: the total wall time spent by the group on the task
- **g_fracFemale**: the proportion of group members who are female
- **g_personTime**: the total person-hours spent by the group on the task
- **g_totalEffort**: the total effort of the group
- **g_effortPerPerson**: the effort per person in the group
- **g_fractionalScore**: the score of the group compared to the gold standard
- **g_binaryScore**: whether the group succeeded or failed at the task
- **g_precision**: the precision of the group's events compared to the gold standard
- **g_recall**: the recall of the group's events compared to the gold standard
- **g_f1**: the F1-measure of the group's events compared to the gold standard
- **g_avgIndivEntropy**: the average of the group members' entropy over task types
- **g_effortEntropy**: the entropy over all team members' effort
- **g_groupEntropy**: the entire group's entropy over task types
- **g_eventCollaboration**: average proportion of group members working on each event
- **g_eventCollaborationExVoting**: average proportion of group members working on each event, excluding voting to verify the event
- **g_chatFrac**: fraction of time the group spent chatting
- **g_chatWeight**: weighted proportion of time the group spent chatting
- **g_classifyFrac**: fraction of time the group spent classifying content
- **g_classifyWeight**: weighted proportion of time the group spent classifying content
- **g_filterFrac**: fraction of time the group spent filtering content
- **g_filterWeight**: weighted proportion of time the group spent filtering content
- **g_verifyFrac**: fraction of time the group spent verifying content
- **g_verifyWeight**: weighted proportion of time the group spent verifying content
----------
**synthetic.csv**
- **personTime**: the total person-hours spent by the group on the task
- **totalEffort**: the total effort of the group
- **fractionalScore**: the score of the group compared to the gold standard
- **binaryScore**: whether the group succeeded or failed at the task
- **precision**: the precision of the group's events compared to the gold standard
- **recall**: the recall of the group's events compared to the gold standard