Author(s):

Intille, Stephen, S.; Tapia, Emmanuel, M.; Rondoni, John; Beaudin, Jennifer; Kukla, Chuck; Agarwal, Sitij; Bao, Ling; Larson, Kent

Title:

Tools for Studying Behavior and Technology in Natural Settings

Publication:

Ubicomp 2003: Ubiquitous Computing 5 th International Conference

Keywords:

anthropology; anthropology-visual; context-aware computing; diary studies; diary studies-photographic; ethnographic study; ethnography-video; experience sampling method; interviews; field studies; naturalistic observation; psychology; qualitative research methods; quantitative research methods; self-reporting; social sciences; user evaluation; user questionnaire; user research

Paper Summary:

The purpose of this paper is to discuss three tools for learning about users, their needs and their tasks. The authors believe that it is important to understand users and their behaviors and then develop technology suited to users' needs (not the other way around, as frequently seen in product development lifecycles). The primary motivation for the development of these tools it to study people using the technology in realistic, non-laboratory settings for long periods of time in order to measure learning and behavior change.

The authors discuss five traditional methods of naturalistic observation:

1. Interviews. Users explain tasks. Difficulty recalling tasks, impact of context, etc.
2. Direct observation . Provides accurate data. Costly, time-consuming, disruptive.
3. Self-report: recall surveys . Suffer from recall and selective reporting biases.
4. Self-report: time diaries . User log items in journal as event occurs. Provide less biased data than recall surveys but are cumbersome for user and can impact activity.
5. ESM/ESA. Experience sampling method (ESM) users a timing device to trigger self-reporting. Can be disruptive and irritating to users.

Seeking to improve user assessment techniques, the authors created tools. They authors believe that these tools (or a combination of tools) must:

• Measure a great deal of activity, over long periods of time, without being invasive
• Be easy to install and low in cost
• Minimize the burden on users (especially self-reporting)

The three tools include:

1. Context-Aware Experience Sampling (CAES). The tool uses sensors to allow researchers to acquire feedback from users only in particular situations.
2. Ubiquitous Environment State-Change Sensor System. This tool uses sensors installed in a user's environment to passively collect data by measuring user behaviors. This tool compliments the self-report data from CAES.
3. Image-Based Experience Sampling. This tool uses sensing applications to trigger audio / video recording to observe user behaviors.

A - Z List    |    By Author    |    By Keyword    |    By Evaluation Technique    |    By Application