In:
ACM Computing Surveys, Association for Computing Machinery (ACM)
Abstract:
Smartphones can unobtrusively capture human behavior and contextual data such as user interaction and mobility. Thus far, smartphone sensor data have primarily been used to gain behavioral insights through correlation analysis. This paper provides a tutorial on the causal analysis of human behavior using smartphone sensor data by reviewing well-known matching methods. The key steps of the causal inference pipeline employing matching methods are illustrated using a concrete scenario involving the identification of a causal relationship between phone usage and physical activity. Several practical considerations for conducting causal inferences about human behaviors using smartphone sensor data are also discussed.
Type of Medium:
Online Resource
ISSN:
0360-0300
,
1557-7341
Language:
English
Publisher:
Association for Computing Machinery (ACM)
Publication Date:
2024
detail.hit.zdb_id:
215909-0
detail.hit.zdb_id:
1495309-2
detail.hit.zdb_id:
626472-4
Permalink
