The simple physics task involved solving physics problems by analogy to a worked-out sample solution. Each problem comprised an introduction to the context of the problem, the names and values of relevant quantities, and a question indicating the unknown (desired) quantity. The sample problem included a solution with two parts: a variable equation for the answer using variables, and a value equation using the actual values in the problem (i.e. the instantiated version of the variable equation). Using a sample problem, subjects solved new problems by determining an appropriate answer and entering their response as an unsimplified expression. To analyze when subjects encoded the various facts on the screen, parts of the screen were covered with opaque blocks that could be uncovered by clicking on them with the mouse. Please refer to Figures 10.1 and 10.2 in the book chapter for sample screens with and without blocks.
The experiment was run in two conditions: the positive-correlation (PC) condition and the negative-correlation (NC) condition. In the PC condition, each variable in the variable equation was named using the first letter of the quantity it represented; for instance, "m" represented mass and "L" length. In the NC condition, each variable was named using the first letter of a quantity it did not represent; for instance, "L" represented mass and "m" length. The PC condition induced subjects to focus their attention on the (correct) variable equation, while the NC condition induced subjects to ignore the (misleading) variable equation and focus their attention on the value equation. The sample problems and solutions for the eight problem types in each condition are available here.
The stimuli screens were divided into various areas for analysis: the sample problem SP, the sample solution variable equation S1, the sample solution value equation S2, the test problem TP, and the test solution TS. Labels that indicate the location of these areas are included in Figure 10.1.