Ogonowski, Rosebery, & Hudicourt-Barnes, 2001). They then observe firsthand that
not every plant in the fertilized group responds the same and that the effect of the
fertilizer becomes evident, if at all, only when comparing the two groups. With
observational data, students must reason backwards from observed differences to
possible explanations for those differences, and their tendency in explaining the data
is to offer different causal accounts for each individual value. With the experimental
setup, students first see the process and then the data resulting from it, a difference
in perspective that may help them focus on the class of causes that apply uniformly
at the group, as opposed to the individual, level.