In research and policy, we often rely on what people say they will do. We ask patients if they would take a new vaccine, ask doctors if they would perform a risky surgery, and ask investors how much risk they can tolerate. However, a recurring theme across decades of behavioral research is that stated preferences rarely align perfectly with revealed behavior.
This discrepancy, often called the "intention-behavior gap" or "hypothetical bias," suggests that human decision-making in a survey is fundamentally different from decision-making in the wild.
The Anatomy of the Gap
Why do people fail to follow through on their stated intentions? The research suggests three primary drivers:
- Social Desirability Bias: In surveys, participants often provide the "correct" or "moral" answer to appear responsible to the researcher. Kim Lavoie and Mingtao Huang note that this is particularly prevalent in public health, where individuals overstate their willingness to comply with recommendations like vaccination or biobank participation to align with social norms.
- The Absence of Real-World Pressure: A survey is a "cool" cognitive environment. Real life is "hot." Kazuyoshi Okada found that while nephrologists might theoretically support withdrawing dialysis for terminal patients, the emotional and legal pressures of a real clinical setting create significant hesitation. Similarly, Emily D. Sherman argues that hypothetical financial rewards in studies cannot replicate the physiological stress of losing actual money.
- Situational Constraints: Intentions are often formed in a vacuum, ignoring the "friction" of reality. Janeta Nikolovski and Alexis Larson highlight that even when a patient is 100% willing to receive care, logistical hurdles—like insurance coverage, travel time, and scheduling—can prevent that intent from becoming an action.
Measuring Preferences in Healthcare
To get closer to the truth, researchers use Discrete Choice Experiments (DCEs). Instead of asking "Do you want this drug?", they force participants to make trade-offs, mimicking real-world constraints.
Jingyi Meng and Jinsong Geng have used this method extensively in China to evaluate preferences for cancer and diabetes treatments. Their work reveals that while patients say they value survival most, their actual choices are heavily dictated by out-of-pocket costs and quality-of-life improvements. Even in these sophisticated models, however, Xiaolan Yu warns that a gap remains: a patient’s "best" choice in a simulation may still differ from their choice when facing an active disease.
Bridging the Gap with Better Proxies
If surveys are unreliable, how can we predict behavior? Researchers are moving toward more immersive and physiological evaluation methods:
- Health Concept Surveying (HCS): Alex Mariakakis proposes using scenario-based "vignettes" that combine psychological models with low-fidelity prototypes to untangle why a user might abandon a health technology before it is even built.
- Virtual Reality (VR): Clara Vandeweerdt demonstrated that letting people "experience" the benefits of herd immunity in a gamified VR environment was significantly more effective at increasing vaccination intentions than traditional text-based information.
- Rational Inattention: Rahul Bhui suggests that "impatience" in decision-making might actually be a rational response to the mental effort required to simulate the future. By understanding the "noise" in how we think about future rewards, we can better predict when people will choose immediate gratification over long-term health.
Why It Matters
The intention-behavior gap is not just a statistical curiosity; it is a major hurdle for public health and economic stability. As M. D. Hagen and Chuanyi Ning argue, if we evaluate the success of a program based only on "reported intent," we may be funding interventions that look great on paper but fail to save lives in practice. Moving the field forward requires shifting focus from what people promise to do toward the situational and psychological triggers that actually drive them to act.