The data set contains counts of ants, seen on specific activities in various Arizona ant colonies. This data set has led to the detection of the daily round of harvester ants: the ants start their daily round by patrolling, followed by nest maintenace workeres cleaning up their site. Only later the forager ants are sent out to collect new food. Unfortunately, a competing species is earlier up and ends its day by messing up the site of our harvester ants. The consequence is that the harvester ants spend the better part of the day by cleaning up, and have too little time to havest....
The daily round of ants: activity count of ants, drawn as box&whisker plots, against time of day.
Each box&whisker plot represents data from several colonies, by activity group.
The different activity patterns of the groups is evident: first activity of patrollers, with maximum at 6:00 (pat6), then main nest maintenance activity with peak at 7:00 (nst7), then foraging with peak at 8:00 (for8).
Still to be improved: the shaded bands provide a quick test for pairwise comparison of medians.
What would be adequate bands for a profile curve ?
For more on ants, visit the Gordon Lab.
The data analysis has been done using DataDesk.
$Source: /u/math/sa3/cvswww/www/www.statlab.uni-heidelberg.de/data/ants/index.html,v $
$Revision: 1.3 $
$Date: 2013/11/02 16:41:40 $