HabitLab : in-the-wild behavior change experiments at scale
Abstract/Contents
- Abstract
- Behavior change systems help people manage their time online. However, existing productivity systems have tended to assume a one-size-fits all solution, whereas there are many factors - novelty effects, attrition, influences from other apps and devices, and differences in individual motivation - that we must take into account. That said, these effects have been researched mostly in small-scale labs studies in domains other than online behavior change, so there is a large space of opportunities for studying how these effects manifest in real-world online behavior change contexts, and how to design better behavior change systems using these insights. In this thesis we present HabitLab, an in-the-wild experimentation platform we developed for conducting behavior change experiments, as well as a set of studies we ran on the platform. HabitLab is a browser extension and mobile phone app with over 12,000 daily active, voluntary users, that users install to help them reduce time online and on their phones. It works by displaying one of 20+ interventions whenever they open an app or site they wish to spend less time on. We use HabitLab as a large-scale experiment platform to understand behavior change. In our first set of studies, we investigate novelty effects of interventions, finding that compared to always showing the same intervention, a strategy of rotating between different interventions improves intervention effectiveness, but at the cost of increased attrition. This attrition is partly due to users being unfamiliar with rotating interventions, and improving users' mental models with a notice shown whenever a new intervention is shown is able to reduce this attrition. In our second set of studies, we investigate whether reducing time on one site or app by intensifying interventions influences time on other sites, apps, and devices. We find that on the browser, reducing time on one site reduces time spent elsewhere, but we do not observe the effect on mobile devices, and do not observe cross-device effects. In our third set of studies, we investigate users' motivation levels over time as indicated by the difficulty of interventions they select. We find that users initially overestimate how difficult of interventions they want, and their choices of difficulty progressively decline over time. Thus, we have found that online behavior change is a domain where incentives for users and researchers line up such that researchers can run large-scale in-the-wild experiments gaining ecologically valid insights about how behavioral psychology and economics theories play out in the real world, while users benefit from the more effective, scientifically informed behavior change systems that we can develop using these experiments and data.
Description
| Type of resource | text |
|---|---|
| Form | electronic resource; remote; computer; online resource |
| Extent | 1 online resource. |
| Place | California |
| Place | [Stanford, California] |
| Publisher | [Stanford University] |
| Copyright date | 2019; ©2019 |
| Publication date | 2019; 2019 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Author | Kovács, Géza |
|---|---|
| Degree supervisor | Bernstein, Michael S, 1984- |
| Thesis advisor | Bernstein, Michael S, 1984- |
| Thesis advisor | Fogg, Brian J |
| Thesis advisor | Landay, James A, 1967- |
| Degree committee member | Fogg, Brian J |
| Degree committee member | Landay, James A, 1967- |
| Associated with | Stanford University, Computer Science Department. |
Subjects
| Genre | Theses |
|---|---|
| Genre | Text |
Bibliographic information
| Statement of responsibility | Geza Kovacs. |
|---|---|
| Note | Submitted to the Computer Science Department. |
| Thesis | Thesis Ph.D. Stanford University 2019. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2019 by Geza Kovacs
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...