Measuring cognitive load with subjective rating scales during problem solving: differences between immediate and delayed ratings

Subjective cognitive load (CL) rating scales are widely used in educational research. However, there are still some open questions regarding the point of time at which such scales should be applied. Whereas some studies apply rating scales directly after each step or task and use an average of these ratings, others assess CL only once after the whole learning or problem-solving phase. To investigate if these two approaches are comparable indicators of experienced CL, two experiments were conducted, in which 168 and 107 teacher education university students, respectively, worked through a sequence of six problems. CL was assessed by means of subjective ratings of mental effort and perceived task difficulty after each problem and after the whole process. Results showed that the delayed ratings of both effort and difficulty were significantly higher than the average of the six ratings made during problem solving. In addition, the problems we assumed to be of higher complexity seemed to be the best predictors for the delayed ratings. Interestingly, for ratings of affective variables, such as interest and motivation, the delayed rating did not differ from the average of immediate ratings.