The table on the next page presents mean cone size (mass) of lodgepole pine in 16 study sites in three types of environments in western North America (Edelaar and Benkman 2006). The three environments were islands of lodgepole pines in which pine squirrels were absent (an “island” here refers to a patch of lodgepole pine surrounded by other habitat and separated from the large tracts of contiguous lodgepole pine forests), islands with squirrels present, and sites within the large areas of extensive lodgepole pines (“mainland”) that all have squirrels.

The main comparison of interest in this study, identified before the data were gathered, was the comparison between islands with and without squirrels, because this comparison controls for any effects of forest isolation on the mass of lodgepole pine cones.

a. What do we label this type of comparison?

b. Taking into account the type of comparison identified in part (a), calculate a 95% confidence interval for the difference in cone mass between islands with and without squirrels. Assume that sites were randomly sampled.

c. Using these data, carry out a test of the differences among the means of all three groups.