Abstract
Constructive behaviors, such as drawing and writing, have long been theorized to be a potent driver of learning (Chi & Wylie, 2014). Consistent with this view, prior work has found that creating drawings can enhance the ability to recall previously studied verbal concepts relative to other, non-constructive encoding tasks (Wammes et al., 2016). By what route do such constructive behaviors improve memory and to what degree do such benefits generalize to other targets for memory, such as visual objects? Here we explored these questions by conducting an experiment to measure the impact of constructing block towers on the ability to later discriminate those towers from other highly similar, but never studied towers. Participants (N=50) were presented with a series of 2D block towers in an initial encoding phase: on half of these trials, they were asked to build the tower from blocks in a virtual building environment; on the other half of trials, they were asked to simply view the tower for several seconds. In a subsequent test phase, participants then had to discriminate previously studied towers (whether built or viewed) from novel foils by providing old/new judgments. We found that participants achieved above-chance discrimination performance for both built and viewed towers (d’=1.38, p<0.001), showing they had successfully engaged with the task. However, participants performed reliably worse in recognizing the towers they had built compared to those they had only viewed (p=0.00960). One possible explanation for this initially unexpected finding is that this constructive encoding task — during which participants focused on individual parts of the tower — led to a weaker representation of the whole tower, at least by comparison with a more holistic encoding task. More broadly, these findings raise important questions concerning the interaction between perception and action for supporting visual memory.