Building Plots

A backpack problem: Some doctors believe that heavy backpacks are responsible for an increase in back problems among adolescents. To pursue related questions, 7th graders in our TinkerPlots test site explored data collected from 79 students in grades 1, 3, 5, and 7.

Building a graph with TinkerPlots: Data icons first appear in a haphazard arrangement in the plot window, as shown below. The icons are colored along a gradient for the attribute selected on the Data Card, in this case "PackWeight." The bluer the icon, the heavier the backpack. Clearly, the heaviest backpack belongs to Gayle, because it is the darkest blue.



You can see Anthony's Data Card (above, upper left). Dragging the data icons separates them into groups. Students use other controls including stack and order to organize the case icons into a display that they can use to answer their question.

Here are the routes taken by three pairs of 7th grade students in exploring the relation between backpack weight and grade:

Pair 1:

One pair of students ordered the data vertically by "PackWeight" and then stacked them:	Then they changed the icon color to display "Grade" and saw that the heavier backpacks at the top of the stack belonged to students in the higher grades (darker shades of red):

Pair 2:

Another pair of students first separated values of "PackWeight" horizontally and then stacked them vertically:


Then they changed the icon color to "Grade" and observed that students in grades one (lighter colors) clustered at the lower end of the graph and thus tended to have lighter backpacks:



Finally, they separated "Grade" into four groups and ordered from smallest to largest, which they thought made the differences more obvious:

Pair 3:
A third pair of students displayed "PackWeight" as value bars, and then ordered them:



They then changed the icon color to "Grade" and observed that students in grades 1 and 3 (lighter colors) clustered at the lower end of the graph and thus tended to have light backpacks. Finally, they separated the data into the 4 different grades: