In a groundbreaking study, researchers have analyzed data generated by citizen scientists using a web-based motor test to gain insights into motor control errors. The study, published in Nature Human Behavior, explores how people correct for motor control errors, potentially paving the way for personalized physical therapy and tailored training routines for athletes.
While this approach does not replace lab-based studies, it complements them by investigating whether motor behavior generalizes to a larger population. According to Jonathan Tsay, the assistant professor from Carnegie Mellon University and the study’s first author, this large-scale approach democratizes motor learning research.
Traditional motor learning studies have typically been conducted in a lab setting, using expensive equipment to capture subtle changes in participants’ movements in response to motor control errors. However, these studies often involve a limited number of participants, leaving the generalizability of the findings uncertain.
Tsay aimed to explore motor skills from a different perspective by utilizing big data. To collect the data, he developed a simple motor-learning assessment that people could complete online from the comfort of their homes. The result was a comprehensive dataset of over 2,000 sessions from a diverse range of participants.
The study also examined the relative contribution of subconscious, implicit motor learning and conscious, explicit motor learning. With the collected data, Tsay was able to analyze how demographic variables affect the relative contribution of these two learning styles.
Compared to an 80-minute lab experiment, the short online test only took eight minutes to complete. Many participants contributed multiple sessions to the database, enabling the research team to efficiently track changes in motor learning over time.
By utilizing big data, researchers can gain a deeper understanding of variables such as gender, age, visual impairment, and even video game experience, which can impact motor adaptation.
Age serves as an example. Previous laboratory studies have produced mixed results regarding the effect of age on motor adaptation. However, due to the small sample size and the focus on extreme age groups (very young and very old), ambiguity remains. Through the use of big data, Tsay and his colleagues were able to examine age as a continuous variable. The results demonstrated how participants adjusted their strategies to correct for motor errors throughout their lifespan, with peak adaptation occurring between the ages of 35 and 45. These insights were overlooked in previous studies with limited sample sizes.
Moreover, utilizing machine learning and other techniques enabled the researchers to predict successful motor learning outcomes and identify key predictors of success, such as speed of movement and reaction time. Tsay notes that these exploratory findings can be further investigated using hypothesis-driven studies in a lab setting to uncover the mechanisms behind the observed online results.
It is important to acknowledge the limitations of the study. The simple motor learning task could only predict about 15% of the variance in the study, thus limiting the overall insights that can be derived from the findings. Additionally, the motor task was not conducted under the supervision of an experimenter or controlled for parameters such as the type of technology and internet speed, potentially resulting in increased noise in the data. Despite these limitations, Tsay believes that this large-scale approach allows for a detailed examination of variability, providing valuable insights to the motor research community.
Richard Ivry, a co-author of the study and distinguished professor in psychology at the University of California, Berkeley, affirms the significance of online studies in motor control research. He notes that while many questions in psychology can be explored through online testing, the field of motor control has lacked similar opportunities. The Nature Human Behavior study increases our confidence in the meaningfulness of online studies for studying motor control, with many labs around the world taking advantage of these tools.
The study, titled “Large-scale citizen science reveals predictors of sensorimotor adaptation,” involved collaboration between Jonathan Tsay, Hrach Asmerian, and Ken Nakayama from the University of California, Berkeley, Laura Germine from Harvard Medical School, and Jeremy Wilmer from Wellesley College.
*Note:
1. Source: Coherent Market Insights, Public sources, Desk research
2. We have leveraged AI tools to mine information and compile it
