The more structured, the better? Learning effects of online scaffolds with different structures for student question generation

Research generally substantiates the positive effects of student question generation (SQG) on students’ cognitive and affective development. Nonetheless, challenges affecting its adoption in classrooms have been reported (e.g. perceived difficulty and elevated concerns over performance at the task). To alleviate such hurdles, researchers have proposed scaffolding of various types to support SQG. Scaffolding types are different in terms of structures, which affect learners’ locus of attention, mental processing, and learning. Given that studies investigated their effects are limited, this study examined the impact of two well-adopted scaffolding types for SQG—what-if-not (i.e. the structured type) and the main idea (i.e. the semi-structured type) on academic achievement, learning strategy use, and attitudes toward the study subject. This study adopted a nonequivalent quasi-experimental research method. Students from two sixth-grade classes (n = 56) participated for five weeks. An online system afforded with customizability to support the two scaffolding types for weekly SQG science learning activities was adopted. Although the results showed that both groups had comparable academic achievement, the participants in the what-if-not scaffolding group significantly outperformed those in the main idea scaffolding group in the use of cognitive and metacognitive strategies and attitudes toward science. Content analyses of student-generated questions further revealed that the what-if-not group generated significantly more proportions of questions containing interlinks among related key concepts both within the current study topic and to prior topics than the main idea group.