Eliciting desirable difficulty through font manipulation: effects on metacognitive judgments, cognitive load and learning outcomes

Abstract Recent multidisciplinary research has investigated whether incorporating desirable difficulty into instructional materials can enhance learning outcomes through minimally demanding interventions. One notable assertion in this literature is that font manipulation, particularly disfluent fonts, may induce desirable difficulty and support learning. In a computer-assisted learning environment, the current study investigated this claim using a between-subjects experimental design. The primary dependent variables were achievement, retention, cognitive load, and judgments of learning, with prior knowledge and working memory capacity serving as covariates. A total of 165 participants were randomly assigned to one of three treatment groups, with instructional materials manipulated with Arial, Helvetica, or Sans Forgetica fonts. The results revealed no significant differences across treatment groups on any outcome measure after controlling for individual differences. Accordingly, all three hypotheses were rejected. These findings indicate that font-based manipulations are insufficient to reliably generate desirable difficulty effects in computer-assisted instructional materials. Overall, the findings reveal that font-based manipulations are insufficient to reliably generate desirable difficulty effects in computer-assisted instructional materials and underscore the need for more cognitively substantive approaches to instructional design in digital learning environments.