The effect of presenting worked examples for problem solving in a computer game

The development of problem solving abilities has been considered "the ultimate goal of education" (Hong & Liu, 2003, p. 245). An effective instructional method to teach problem solving is to use worked examples (e.g., Kalyuga, Chandler, Tuovinen, & Sweller, 2001). For reaching various learning goals, computer games have the potential to facilitate learning and motivation and can be adopted as a learning environment (O'Neil & Fisher, 2004). This study examined the effect of presenting worked examples differently (i.e., just-in-time vs. beforehand) for problem solving in the computer game SafeCracker®. A 3 × 2 mixed two-factor within-subjects experiment was conducted. The independent variable was the instructional intervention that included three levels - the control group (n = 38) that did not receive the worked examples, the worked example group (n = 39) that received the worked examples for studying beforehand and for referencing during the task, and the JIT worked example group (n = 40) that received the worked examples just-in-time during the task. The dependent variable was the pretest and posttest scores on the various measurements. The major hypotheses were that (1) the JIT worked example group would demonstrate greater problem solving and continuing motivation than the worked example group and the control group and (2) the worked example group would demonstrate greater problem solving and continuing motivation than the control group. Problem solving was measured in terms of content understanding, problem solving strategies (i.e., retention and transfer skills), and self-regulation. The JIT worked example group demonstrated significantly better content understanding than the other two groups. The worked example group demonstrated significantly better problem solving retention skill than the control group (by opening more worked-example-illustrated safes). There was no difference among groups on self-regulation while higher levels of self-regulation may or may not associate with better content understanding and problem solving strategies. Higher effort tended to be associated with more improvement on content understanding yet more planning tended to be associated with inferior problem solving transfer skill. Only 6 out of 117 participants exhibited continuing motivation. Various explanations (e.g., from a cognitive load perspective) were provided for the findings.