Developmental Trend of School-Age Students’ Divergent Thinking: A Meta-analysis

Background: Over the past decades, there has been a great deal of research on the development of school-age students’ divergent thinking. However, research findings regarding this issue have been inconsistent. Some studies have provided evidence for a continuous development of divergent thinking as grade level increases. Other studies have suggested a discontinuous developmental trajectory including one or more periods of significant drops. Torrance (1967) found in seven different cultures that a drop occurs in Grade 4, which has become widely known as the fourth grade slump in divergent thinking. The existence of the fourth grade slump has been reported in several subsequent studies. On the contrary, other studies have found no evidence of the fourth grade slump; some studies have found an increase or no decline in Grade 4, and other studies have found a slump but in other grades including Grades 1, 6, 7, and 9. In addition to the inconsistent results, most of the previous studies have been conducted on a small number of subjects and a limited grade range. Hence, the picture is less clear concerning whether divergent thinking slumps actually exist, how many there are, and when they occur. Objectives: The purpose of this study was to meta-analyze previous research results regarding the development of school-age students’ divergent thinking from Grades 1 to 12, with a particular focus on the fourth grade slump as it has sparked a major debate among researchers. We also examined whether the change in divergent thinking is affected by divergent thinking test, divergent thinking domain, intellectual ability, gender, and country of study. Research questions: This study attempts to answer the following questions: (1) How does school-age students’ divergent thinking change from Grades 1 to 12? (2) Does the fourth grade slump in divergent thinking exist? (3) Are there moderator variables that account for the observed variability across studies concerning the change in divergent thinking from Grade 3 to 4? Method: We calculated for each study a standardized mean per grade, and combined these standardized means in a meta-analysis. A meta-analytic three-level model was employed in order to account for dependence within studies. To examine divergent thinking changes from Grades 1 to 12, we included 11 (number of grades - 1) dummy variables as predictors in the meta-analytic model. The first dummy variable takes the value 0 in the case of Grade 1 and 1 otherwise, and the second dummy variable takes the value 0 in the case of Grades 1 and 2 and 1 otherwise. Other dummy variables were coded using the same procedure, until the 11th dummy variable which is equal to 1 in the case of Grade 12 and 0 for the previous grades. In this way, the coefficient of the first dummy variable captures divergent thinking change from Grade 1 to 2, the second coefficient captures divergent thinking change from Grade 2 to 3, and so on. To avoid an excessively complicated model, the effects of the moderator variables were investigated only for divergent thinking change from Grade 3 to 4. To study the influence of each of the moderator variables, we included an additional term in the model, which represents the interaction between the dummy variable capturing divergent thinking change from Grade 3 to 4 and the suggested moderator variable. Data sources: The present meta-analysis included divergent thinking literature published up to December 31st, 2017. The search process consisted of the following four steps: First, the following databases were searched: ERIC, Google Scholar, JSTOR, PsycARTICLES, Scopus, and Web of Science. Second, the reference lists of the papers identified in the first step were reviewed for other relevant references (i.e. “backward search”). Third, more recent references were retrieved by searching databases for papers that referred to the previously identified papers in steps 1 and 2 in their citations (i.e. “forward search”). Fourth, the relevant key journals were hand-searched. The papers identified using the search process were first screened for their relevance on the basis of their titles and abstracts. The remaining papers were included if they met the following two criteria: (1) reports on an original, empirical, and quantitative study, and (2) examines differences in divergent thinking between Grade 4 and other Grades (1-12). Moreover, we only included (1) journal articles, conference papers, or dissertations (2) that were written in English, and for which (3) the full text was available. Results: A total of 2,139 standardized means from 41 studies were analyzed. Overall, the results showed an upward trend of divergent thinking across grades; however, there were some discontinuities. Also, there was no evidence of the fourth grade slump; instead a seventh grade slump was noted at both the overall and subscale (i.e., fluency, flexibility, and originality) levels (see Figure 1). Task domain significantly moderated the change of the overall divergent thinking in Grade 4. At the subscale level, intellectual ability moderated the change of fluency, as well as country of study moderated the change of originality in Grade 4. Conclusions and implications: The results of this study inform the ongoing debate concerning the development of school-age students’ divergent thinking. Furthermore, these results suggest a slump in divergent thinking in Grade 7. This might have valuable implications for parents, teachers, and other professionals working with students and could be used to guide interventions and training programs to promote divergent thinking development. Finally, our study revealed different developmental trends of divergent thinking in terms of task domain, intellectual ability and country of study. Hence, these factors need to be considered carefully when investigating divergent thinking development.