A New Look at Infant Problem-Solving: Using DeepLabCut to Investigate Exploratory Problem-Solving Approaches

Abstract

When confronted with novel problems, problem-solvers must decide whether to copy a modeled solution or to explore their own unique solutions. While past work has established that infants can learn to solve problems both through their own exploration and through imitation, little work has explored the factors that influence which of these approaches infants select to solve a given problem. Moreover, past work has treated imitation and exploration as qualitatively distinct, although these two possibilities may exist along a continuum. Here, we apply a program novel to developmental psychology (DeepLabCut) to archival data (Lucca et al., 2020) to investigate the influence of the effort and success of an adult's modeled solution, and infants' firsthand experience with failure, on infants' imitative versus exploratory problem-solving approaches. Our results reveal that tendencies toward exploration are relatively immune to the information from the adult model, but that exploration generally increased in response to firsthand experience with failure. In addition, we found that increases in maximum force and decreases in trying time were associated with greater exploration, and that exploration subsequently predicted problem-solving success on a new iteration of the task. Thus, our results demonstrate that infants increase exploration in response to failure and that exploration may operate in a larger motivational framework with force, trying time, and expectations of task success.

Keywords: DeepLabCut; automated behavioral analysis; cognitive development; exploration; infant development; motion capture technology; problem solving.

FIGURE 1

(A) A still from a participant video showing markers generated by DLC after training the neural network. The three parts of the rope labeled and tracked were: the base of the rope handle (blue), the middle of the rope handle (green), and the end of the rope handle (red). (B) A graphical representation of the coordinates extracted from DLC. Additionally, we have uploaded a video of one of this participant’s test trials with computed marker overlay. Video playback is in real time and can be found here: https://osf.io/5z74k/.

FIGURE 2

Imitative similarity plotted as (A) an additive score combining variation from imitation in the x- and y-axes and (B) a difference score subtracting variation from imitation in the y-axis from the x-axis. Each graph is plotted by time, divided by condition and trial number. In the case of the additive score, greater values represent greater imitation, while in the case of the difference score, greater values represent greater variation in the x-axis relative to the y-axis.

FIGURE 3

Variability plotted (A) using a per-second score to represent variability in real-time and (B) using an overall score to represent the variability created across the entire trial, divided by condition and trial number. For each score, greater values represent greater variability and exploration.

FIGURE 4

Relationship between performance measures and average imitative similarity: (A) maximum pulling force, (B) trying time, (C) negative affect, (D) help-seeking, and (E) hints during recovery. The shaded region along the line of best fit represents standard error.

FIGURE 5

Relationship between performance measures and overall spatial variability: (A) maximum pulling force, (B) trying time, (C) negative affect, (D) help-seeking, and (E) hints during recovery. The shaded region along the line of best fit represents standard error.