Gesture-Speech Mismatch Predicts Who Will Learn to Solve an Organic Chemistry Problem

• In Event: Spatial Thinking in Chemistry

Abstract

When explaining answers to problems they have yet to master, some children spontaneously express a problem-solving strategy in gesture that is not expressed anywhere in the accompanying speech. Child learners who produce gesture-speech mismatches are more likely to learn from instruction than learners whose gesture does not convey additional strategies (Church & Goldin-Meadow, 1986). Do adult learners produce mismatches when they solve a more complex problem? If so, do adult mismatches predict learning in the same way that child mismatches do?

We gave adult novices (N=43) instruction in one strategy for solving a complex organic chemistry problem involving stereoisomers, which were defined to participants as “molecules with multiple non-superimposable spatial arrangements”. On each pretest trial, participants were asked to create a drawing of a stereoisomer of a given molecule and to explain why their drawing represented a stereoisomer. All participants then received the same instruction, which included only disambiguating pointing gestures. Posttest was the same as pretest, with different molecules.

We coded gesture and speech separately, and we found that adult learners did produce mismatches, on 36% of pretest trials. We grouped learners into four categories depending on the relationship between speech and gesture at pretest. Four learners were categorized as Matchers—their gesture did not add any information to speech on any pretest trial. Seven learners were categorized as Incorrect Mismatchers—they did express strategies in gesture that were not in their speech on at least one trial, but that information was incorrect. A majority of learners (N=18) were categorized as Correct Mismatchers—they expressed a correct strategy in gesture that was not in their speech on at least one pretest trial. Finally, 14 learners were categorized as Incorrect/Correct Mismatchers—they produced both types of mismatches at pretest.

Adult learners do produce gesture speech mismatches—and, unlike what has been reported with child learners, those mismatches sometimes include information that is wholly incorrect and therefore not useful for learning the task. Of the 11 adults who produced no mismatches or only incorrect mismatches, 0 of them displayed any mastery of the task at posttest. Only 9 learners did benefit from our very short instruction—of those, 5 were Correct Mismatchers and 4 were Correct/Incorrect Mismatchers. For statistical purposes, then, we grouped participants into two categories, and found that those who ever produced a correct mismatch were more likely to display mastery of the task than those who never produced a correct mismatch (9 of 32 vs. 0 of 11, Fisher Exact Test, p=0.05).

Adult learners whose gestures embodied correct problem-solving strategies that were not expressed in speech were more likely to learn from subsequent instruction on a complex organic chemistry task than were learners who did not express additional correct strategies in gesture. This suggests that instructing novices to produce correct representations in their gestures could be beneficial for learning.

Authors

• Raedy Ping, University of Chicago

• Mary-Anne Decatur, University of Chicago

• Samuel Larson, University of Chicago

• Elena Zinchenko, University of Chicago

• Susan Goldin-Meadow, University of Chicago