The purpose of this study is to investigate the effects of fill-in-the-blank problem structures (fully structured, partially structured, and unstructured) on the mathematical performance of third-grade elementary students, and to examine whether these...
The purpose of this study is to investigate the effects of fill-in-the-blank problem structures (fully structured, partially structured, and unstructured) on the mathematical performance of third-grade elementary students, and to examine whether these effects vary depending on the nature of the mathematical content. A total of 26 third-grade students participated in the study across three mathematics units: addition and subtraction, plane figures, and multiplication. For each unit, three problems with varying levels of structure were administered sequentially, and student performance was assessed using a rubric, paired-samples t-tests, and qualitative analysis of student work samples.
In the number and operations domain, results showed a common tendency for fill-in-the-blank structures to suppress students' mathematical performance. In the addition and subtraction unit, performance increased significantly as problem structure decreased (t = −3.389, p < .001), with only approximately 4% of students (1 out of 26) reaching the target level on the fully structured item, compared to approximately 73% (19 out of 26) on the unstructured item. In the multiplication unit, no statistically significant differences were found; however, 64% of students who rated the partially structured item as easiest actually scored at Level 0 or 1, suggesting that fill-in-the-blank formats can create an illusion of understanding.
In the geometry domain, no statistically significant differences were found across the three items in the plane figures unit. However, qualitative analysis revealed that the partially structured item using dot grid paper yielded the highest proportion of students reaching Level 3 (65%), while the unstructured item led to a reduction in directional diversity. It was also confirmed that the ability to verbally define a figure and the ability to draw it freely are independent competencies.
In conclusion, the effectiveness of fill-in-the-blank problems varies by mathematical domain. Unstructured problems were more effective in the number and operations domain, while appropriate structural cues enhanced diversity of expression in the geometry domain. Teachers should avoid using fill-in-the-blank items as standalone assessment tools and instead combine them with unstructured items and interviews to evaluate both accuracy and diversity of mathematical performance.