1. Slime molds (genus Physarum) are single-celled (but multi-nucleate) organisms that grow via expanding tentacles to cover surfaces such as the undersides of rotting bark or simple agar plates. They display some fascinating traits and abilities that are very much like primitive learning (habituation). As they explore their environment by spreading tentacles out in random directions, they leave slime trails; if a given extension does not receive positive feedback (such as access to food), it is retracted and the slime trail is not crossed again. In this way, the organism eventually adopts the shortest path connecting itself to positively-reinforcing stimuli and learns to ignore non-productive routes. As it turns out, genetically-different strains from different parts of the world learn at different rates.
2.  A researcher is studying these genetic differences by recording the amount of time (in hours) that it takes a slime mold to achieve the optimal growth pattern; she looks at 10 replicate slime molds from each of two different genetic strains (total n = 20). She determines that, while the medians of the two strains are the same (suggesting no difference in learning), the means and standard deviations are quite different; strain 2 has a significantly greater mean time-to-optimal (suggesting slower learning), but also a much larger standard deviation and significant positive skewness.
3. 
   
4. How can you reconcile these two conclusions regarding differences in learning time generated by looking at the medians vs. means? Which is the most appropriate and why?