Thank guys, yours advice helps me a lot (especially Victor_G which emphasizes his steps, since I am not an expert in JMP).
Regarding bee complete the task, I will have to give more details.
The bee enters to room with 8 artificial flowers, I defined visit in all the flowers as success at task.
Finding the first flower occurs almost immediately, finding the last flower is the most difficult because there is a high chance the bee will reach to flower that she already visited (The amount of reward in each flower is very small so the bee eventually will go through all the flowers before returning to the hive).
All the events are recorded so I can know when the bee reach the first flower, second etc.
Because reaching the last flower is the most difficult, there is a lot of "noise" in the measurement.
The data I provided at the beginning shows the time it took to reach flower number 7, now I also added the time to reach flower 6.
In general it seems that there is learning from trial to trial.
I'm interested to search if there is a difference between the treatments, the difference can be the time that required to complete the task or difference in the variance between groups (maybe your suggestion on what to look for).
Thank.