Elementary school students struggle in interpreting the equal sign as a symbol denoting equivalence. Although many have advocated using a pan-balance scale to help students develop this understanding, less is known about what features associated with this model support learning. To attempt to control and examine these features, the investigators developed a digital, pan-balance computer applet. This allowed for experimentally manipulating three relatively grounded instructional conditions (involving the core principle of making two sides the same; a balancing analogy; or both, along with a dynamic demonstration), compared to a relatively idealized control condition. Results indicated that the relatively more grounded conditions promoted a relational understanding of mathematical equivalence among 148 second- and third-grade students and further suggest that providing dynamic, grounded support may not be as optimal as less-enriched supports to promote students' learning.