Self-determined Behavior Change Goals are Dynamic, Diverse, and Intrinsically-Motivated

Mina Khan, Pattie MaesBcss@persuasive 2021

Abstract

Traditional behavior change applications often impose rigid, top-down targets that fail to account for the fluid nature of human motivation. This study investigates "free-living" goal-setting, where users define their own objectives. Our findings reveal that self-determined goals are remarkably dynamic—frequently pivoting based on self-reflection—and span a wide variety of life domains beyond standard health metrics.

Introduction

The "quantified self" movement typically operates on a model where researchers or clinicians set specific targets, such as "10,000 steps a day." However, Self-Determination Theory (SDT) posits that long-term behavior maintenance requires autonomy and intrinsic motivation. This paper addresses the gap in understanding what goals individuals choose when not confined to a specific app's ecosystem, aiming to inform the design of "autogenous" systems that support evolving user motivations.

Methodology

We conducted a 4-week longitudinal study with 10 university students in a low-intervention, free-living environment. Each week, participants listed up to three behavior change goals via surveys, explaining the "what," "how," and "why" behind each. To test the influence of external nudges, we recommended a standard "stay hydrated" goal. Qualitative thematic analysis was used to categorize goals, while quantitative tracking measured goal "churn" (additions and abandonments).

Results

Goal Dynamics and Churn

Behavior change was found to be non-linear. Participants maintained an average of 1.725 goals per week, but these goals were "liquid," with a weekly churn rate of approximately 50%. On average, participants added 0.41 new goals and abandoned 0.47 old ones each week. This suggests that abandonment is often a form of recalibration based on new self-knowledge rather than simple failure.

Average total, added, and abandoned goals per participant per week over the study duration.

Diversity of Goal Categories

Participants largely rejected the default "hydration" recommendation (only 2/10 adopted it), preferring personalized goals. We identified seven distinct categories: Exercise, Diet, Mind, Sleep, Work, Chores, and Other. Even within categories, implementation was highly individualized, ranging from "stretching" to "gyming 30 minutes."

Distribution of goal categories and specific examples across the participant group over four weeks.

Discussion

The high rate of goal turnover highlights that self-reflection is a primary driver of behavior change. Participants frequently adjusted their goals as their self-knowledge evolved—for instance, moving from vague intentions to more achievable, specific plans. The holistic diversity of goals indicates that users view behavior change as a broad life-improvement process rather than a narrow health-tracking task.

Conclusion and Future Work

Future Behavior Change Support Systems (BCSS) should move away from rigid tracking of pre-set metrics. Instead, they should facilitate flexibility and periodic self-reflection, allowing users to pivot their goals as their circumstances and self-understanding change. Limitations of this study include the focus on intention over action, the small sample size, and potential bias in the student population.

Summary updated

BibTeX
@inproceedings{khan2021self,
  author = {Mina Khan and Pattie Maes},
  title = {{Self-determined Behavior Change Goals are Dynamic, Diverse, and Intrinsically-Motivated}},
  booktitle = {Bcss@persuasive},
  year = {2021},
}
Figure