Machine learning reveals the diverse nature of habit formation, as a recent study conducted by Caltech highlights the varying timelines required for different habits to take root. While establishing a gym habit may take an average of six months, healthcare workers can develop a hand-washing habit within a few weeks. This groundbreaking research underscores the immense potential of machine learning in exploring human behavior outside the controlled environment of a laboratory.
A recent machine learning study uncovers the time it takes for different habits to become ingrained.
Initially, putting on your workout attire and making your way to the gym may feel like a daunting task. However, over time, you may effortlessly find yourself attending your Zumba class or going for a run on the treadmill, as going to the gym becomes second nature. Social scientists at Caltech have now shed light on the duration required to establish a gym habit, with an average of approximately six months.
The same study also examined the time it takes for healthcare workers to develop a habit of hand-washing, revealing an average timeline of just a few weeks.
Anastasia Buyalskaya (PhD ’21), currently an assistant professor of marketing at HEC Paris, emphasizes, “There is no definitive time frame for habit formation.” The study, published in the Proceedings of the National Academy of Sciences, also includes contributions from Caltech’s Colin Camerer, Robert Kirby Professor of Behavioral Economics and director and leadership chair of the T&C Chen Center for Social and Decision Neuroscience, as well as researchers from the University of Chicago and the University of Pennsylvania. Another author is Xiaomin Li (MS ’17, PhD ’21), formerly a graduate student and postdoctoral scholar at Caltech.
“While you may have heard that it takes around 21 days to establish a habit, that estimation lacks scientific basis,” explains Camerer. “Our research supports the idea that the speed of habit formation varies depending on the specific behavior and several other factors.”
This study stands as the first to employ machine learning tools to investigate habit formation. The researchers utilized machine learning techniques to analyze extensive datasets comprising tens of thousands of individuals who either swiped their badges to access gyms or engaged in hand-washing during hospital shifts. Collaborating with 24 Hour Fitness for the gym research and a company employing radio frequency identification (RFID) technology to monitor hand-washing in hospitals, the datasets tracked the behaviors of over 30,000 gym-goers over four years and more than 3,000 hospital workers during nearly 100 shifts.
“Machine learning enables us to observe numerous contextual variables that may predict behavioral patterns,” elucidates Buyalskaya. “Rather than starting with a specific variable hypothesis, machine learning does the work of identifying the relevant ones for us.”
Moreover, machine learning allowed the researchers to study individuals in their natural environments over an extended period, surpassing the limitations of previous studies that relied on participant surveys.
The study revealed that certain variables, such as the time of day, had no impact on gym habit formation. However, factors like an individual’s previous behavior did play a role. For instance, 76 percent of gym-goers exhibited a correlation between the time elapsed since their last gym visit and their likelihood of returning. In other words, the longer the duration since their last visit, the less likely they were to develop a habit of going to the gym. Additionally, 69 percent of gym-goers were more inclined to attend the gym on specific days of the week, with Mondays and Tuesdays being the most popular.
Regarding the hand-washing segment of the study, the researchers analyzed data from healthcare workers who were introduced to new requirements involving RFID badges to record their hand-washing activities. “While it is possible that some healthcare workers already possessed the habit before our observation, we consider the introduction of RFID technology as a ‘shock,’ assuming they might need to rebuild their habit from the moment they start using the technology,” Buyalskaya notes.
“In summary, our findings demonstrate the formidable potential of machine learning in studying human habits beyond the confines of a laboratory,” concludes Buyalskaya.
Reference: “What can machine learning teach us about habit formation? Evidence from exercise and hygiene” by Anastasia Buyalskaya, Hung Ho, Katherine L. Milkman, Xiaomin Li, Angela L. Duckworth, and Colin Camerer, 17 April 2023, Proceedings of the National Academy of Sciences.
DOI: 10.1073/pnas.2216115120
The Behavior Change for Good Initiative, the Ronald and Maxine Linde Institute of Economics and Management Sciences at Caltech, and the Tianqiao and Chrissy Chen Institute for Neuroscience at Caltech provided funding for this study.
Table of Contents
Frequently Asked Questions (FAQs) about habit formation
How long does it take to form a habit according to the Caltech study?
The Caltech study on habit formation revealed that the time required to form a habit varies depending on the behavior. On average, it takes about six months to establish a gym habit, while healthcare workers were found to develop a hand-washing habit in just a few weeks.
Is the belief that it takes 21 days to form a habit supported by scientific evidence?
No, the belief that it takes 21 days to form a habit is not supported by scientific evidence. The Caltech study debunks this myth and highlights that the speed of habit formation depends on various factors, including the specific behavior being studied.
How did the researchers use machine learning in the study?
The researchers employed machine learning techniques to analyze large datasets of gym-goers and healthcare workers. They utilized data from badge-swiping at gyms and radio frequency identification (RFID) technology for monitoring hand-washing in hospitals. Machine learning helped identify relevant variables and patterns related to habit formation in these natural environments.
What factors influence habit formation?
The study found that certain factors, such as past behavior, played a role in habit formation. For example, the time elapsed since the last gym visit was a significant predictor of future visits for many participants. Additionally, the day of the week also influenced gym attendance. However, some variables, like the time of day, did not have a significant impact on habit formation.
How significant is machine learning in studying human habits outside of the laboratory?
Machine learning proves to be a powerful tool in studying human habits beyond the controlled environment of a laboratory. It allows researchers to observe numerous contextual variables and analyze large datasets, providing valuable insights into habit formation in real-world settings. This study highlights the potential of machine learning in advancing our understanding of human behavior.
More about habit formation
- Caltech study on habit formation: Read more
- Proceedings of the National Academy of Sciences: Read more
- Behavior Change for Good Initiative: Read more
- Ronald and Maxine Linde Institute of Economics and Management Sciences at Caltech: Read more
- Tianqiao and Chrissy Chen Institute for Neuroscience at Caltech: Read more
3 comments
Omg this study is so cool! It shows how long it takes to form habits like goin to da gym or washin hands. No more 21-day myth, guys! Machine learning rocks!
As a healthcare worker, I’m not surprised that hand-washing habits form in just a few weeks. We’re all about hygiene! Great study by Caltech, it’s about time they used machine learning for this stuff.
This Caltech study is mind-blowing! Machine learning is changing the game when it comes to understanding human behavior. Who knew gym habits and hand-washing could be so fascinating? Props to the researchers!