Vyo Anxiety Study

Vyo Home Robot: Robotic Intervention for Anxiety

Background

University students face frequent stressors, including exams. Such stressors are known to increase state anxiety, or the level of stress at a particular moment. Test Anxiety affects about 16-20% of the student population. Aside from stress, adverse effects of test anxiety in postsecondary school students include decreased performance, GPA, and expectation of success. Relaxation training is one form of behavioral treatment which can reduce test anxiety. Breathing exercises such as pranayama breathing are a common form of relaxation training, and can decrease perceived anxiety and heart rate.

Recently, physically embodied robots and user interfaces have been developed which guide the user's breath in order to reduce anxiety. Netherlands startup Somnox and the Media Innovation Lab (miLab) from the Interdisciplinary Center in Hertsliya, Israel have both created robotic pillows designed to breathe against the user to help them fall sleep Both robots breathe at a certain rate to encourage the user to sync their breath with the pillow, and in turn feel more relaxed as they try to sleep. Meanwhile, researchers at Stanford University have developed a technique that uses peripheral visual cues on a computer screen to slow students' breath as they work. While effects of these cues did not last throughout the duration of the students' work, they were able to significantly reduce the students' breathing rates.

This study design utilizes the relaxing effects of deep breathing along with previous HRI and Human-Computer Interaction designs to examine how physically embodied home robots can be used to reduce state anxiety. As home robots become more integrated into users' lives, there is a greater potential to introduce anxiety-reducing techniques into the robots' behavior.

Research Question

This study will investigate whether and to what extent a home robot can reduce anxiety in teenagers through explicit breathing exercises with the user and continued breathing during a stressful task.

Hypotheses

H1: The increase in participants' heart rate (HR) during a stressful task will be smaller in the case where Vyo breathes continuously during the TSST than in the case where Vyo is still.

H2: The decrease in participants' galvanic skin response (GSR) during a stressful task will be smaller in the case where Vyo breathes continuously than in the case where Vyo is still.

H3: The increase in participants' reported state anxiety after a stressful task will be smaller in the case where Vyo breathes continuously than in the case where Vyo is still.

H4: Participants' HR and GSR will return to normal more quickly after a stressful task in the case where Vyo breathes continuously than in the case where Vyo is still.

Materials

Robotics: Vyo Robot Sensor System: Arduino, Grove Shield, Grove GSR sensor, Grove heart rate sensor, laptop Surveys for Assessing Anxiety: Spielberger State-Trait Anxiety Inventory (STAI) Methods for Provoking Anxiety: Trier Social Stress Test (TSST) Recruiting Participants: Cornell SONA System via Department of Communication Wizard of Oz: Conference Camera, Google Hangouts, Google Firebase, online WOz tool

Procedure

At the beginning of each experiment, the researcher welcomes the participant into a closed study room, with the Vyo Robot sitting still on the table. After the participant signs a consent form, the researcher attaches a heart rate (HR) sensor and a galvanic skin response (GSR) sensor to the participant's non-dominant hand. The researcher leaves the room, and the participant has five minutes to relax so that their HR and GSR can settle.

After five minutes, the researcher returns and administers a twenty question survey representing the state-anxiety section of the State-Trait Anxiety Inventory. Then, the researcher leaves again as the participant engages in a guided breathing exercise with Vyo. The breathing exercise is played as an audio clip and the robot participates in the exercise with the user as a peer. The breathing exercise takes about three minutes, and instructs the student to inhale and exhale at various points. During the instructions, Vyo turns and nods its head, establishing itself as a peer. Similar to a type of slow pranayama breathing, the inhale is longer than the exhale. The inhale lasts about four seconds, while the exhale lasts about eight. In addition to relaxing the participant, this exercise establishes Vyo's periodic movements as breathing. For the control group, the robot halts its breathing motion when the exercise has ended. For the experimental group, the robot continues to breathe for the duration of the experiment.

Then, the researcher leads the participant through an abridged version of the TSST. The test includes ten minutes of preparation for a presentation, a five-minute presentation on why the participant would be a good fit for their dream job, and a five minute math section where the participant is asked to count down from 1022 in steps of thirteen. A camera is placed on the table and directed at the participant throughout the TSST. At the end of the TSST, the participant takes a second STAI survey and then has 15 minutes to relax.

At the end of the experiment, the researcher asks a few interview questions and then removes the sensors, ending the experiment. The start and end times of each activity are tracked throughout the experiment, so that the HR and GSR data can be analyzed correctly.

DIRECTIONS

The scripts, consent form, and other documents related to this study can be found in the "Vyo_Anxiety_Study_Documents" folder and zipped folder at the following link: Vyo Anxiety Study Documents An instructions page is available in this folder with links to relevant documents.

REPOSITORY

The repository for this project can be found here: repository The code for this project was adapted from previous code created by the HRCC lab.