This study compared physical and digital models of a scientific topic in order to determine representational competence, short- and long-term cognition, and the extent to which physical or digital interfaces enhance spatial ability. The DNA molecule was used as a representative spatial topic to investigate conceptual and spatial understanding. Findings suggest a significant effect among four conditions: F(3) = 3.47, p = 0.02. A Tukey-HSD post hoc analysis identified a significant difference between the paper instruction/paper assessment dyad and the digital instruction/digital assessment dyad (d = 0.78). The results of an independent-samples t test analyzing the difference between the methods of instruction regardless of method of assessment demonstrated a significant difference in the scores for paper instruction and digital instruction conditions; t(130) = 2.56, p = 0.006. These outcomes demonstrated greater 3-D representational and conceptual understanding of the DNA molecule when paper models were constructed.