This in given limitations well-designed Dutch study by van de Kamp et al that I found via Dylan William may well be the opposite of what you would expect if I ask you how to promote creativity. Instead of discovery learning and brainstorming, the pupils received explicit instruction on metacognition.

And what are the conclusions?

Albeit statistically medium effects, we think we have found important effects on the three indicators of divergent thinking, taken into account the short span of the intervention of the experimental condition of just 50 min and the complexity of the process of enhancing original ideas. The intervention focused on students’ metacognitive knowledge and metacognitive skills in creative processes and the divergent thinking activities and strategies used in these processes. Due to the molar characteristic of the intervention—an explicit metacognitive strategy instruction on divergent thinking—it is impossible to single out one specific element as the element that might have caused the effect.

In the experimental condition, students were stimulated through explicit instruction to focus on a better representation of their mental model on creativity and divergent thinking—i.e., the meta-level—and were explicitly stimulated through activities to focus on monitoring and control of their divergent thinking surpassing students’ idiosyncratic perceptions of creativity that could hamper students’ divergent thinking. Divergent thinking activities and the generation of original ideas were explained, discussed and modeled, by using and illustrating the two dimensions of the matrix (Table 1). In this intervention lesson, students also learned about metacognitive skills for regulation, related to their own metacognitive monitoring and control processes of the specific and rather complex divergent thinking activities and strategies. Students from the experimental condition were stimulated to deliberately use divergent thinking strategies in real world exercises (Table 3, lesson phase 8) and to monitor and control their generation processes (Table 3, lesson phase 8 and 9).

Students from the comparison condition were stimulated to apply brainstorming activities related to their own visual art products (Table 4, lesson phase 5) without teachers explaining metacognitive knowledge about divergent thinking and without explicit stimulation to monitor their generation processes. In the intervention lesson of the comparison condition, students were brainstorming and reflecting on the content of the theme of ‘Time-grasping’ and they were able to learn by exercising, evaluation and feedback, to focus on activities at object-level: creating ideas for their individual photography series.

In sum, in the redesigned intervention lesson of the experimental condition, three aspects were added: (1) knowledge about twelve specific activities in divergent thinking (2) knowledge about strategies of divergent thinking aimed at generation through remoteness and through abstractness (3) an instructional design focusing on enhancement of students’ metacognitive knowledge and skills to regulate—monitor and control—these complex divergent thinking processes aimed at generating original ideas.

Still we do need to bear in mind that we know that brainstorming as such is often one of the worst ways to get creative.

Abstract of the study:

Visual arts education focuses on creating original visual art products. A means to improve originality is enhancement of divergent thinking, indicated by fluency, flexibility and originality of ideas. In regular arts lessons, divergent thinking is mostly promoted through brainstorming. In a previous study, we found positive effects of an explicit instruction of metacognition on fluency and flexibility in terms of the generation of ideas, but not on the originality of ideas. Therefore, we redesigned the instruction with a focus on building up knowledge about creative generation strategies by adding more complex types of association, and adding generation through combination and abstraction. In the present study, we examined the effects of this intervention by comparing it with regular brainstorming instruction. In a pretest–posttest control group design, secondary school students in the comparison condition received the brainstorm lesson and students in the experimental condition received the newly developed instruction lesson. To validate the effects, we replicated this study with a second cohort. The results showed that in both cohorts the strategy instruction of 50 min had positive effects on students’ fluency, flexibility and originality. This study implies that instructional support in building up knowledge about creative generation strategies may improve students’ creative processes in visual arts education.