March 19, 2019

Sharing our Findings from the Girls Who Code Activities

Blog post written by Christine Tabila and Robear Gillis

Our team shared activities with the Girls Who Code to understand how teens might interact with a social robot meant to help with their stress, as reported in last week’s blog. Girls Who Code is a free after school program where girls meet each week at the RAIN Incubator to code and interact with each other. We collected data and analyzed the results from these activities. We will use our findings to help us learn about the features teens want when interacting with a social robot that would live in their own high school.

User-created personas

Our team had the girls participate in an activity where they created personas of what stress looks like for teens. We asked them to create a collage illustrating what stress looked like on themselves and on their friends using images they found in magazines.

One participant created a collage of her stress using a photo of a boy who was wearing clothes that were too small for him. She also included pictures of an apathetic, inexpressive woman staring blanking forward, and a baby with their face squished against a window to show how she expresses what stress looks like.

On the other side, the participant used pictures of two people covering their faces with their hoodies, a man with long hair and an unbuttoned shirt, and a woman with her arms and legs pulled close to her chest. These images represented what she thought stress looked like on her friends, where themes of anti-social behaviors and unkempt appearance were common.

Similar themes appeared in the personas created by the other participants, highlighting themes of sadness, crying, and sleep changes caused by stress.

Storyboarding

We also had the girls participate in an activity where they drew storyboard sketches of EMAR interacting with teens who were experiencing stress. One of the girls drew a storyboard of EMAR being able to hug and play videos for stressed teens. Another girl drew a picture of herself stressed at an student government meeting, and EMAR helped her by displaying cute images of animals.

Exploring EMAR’s speech and display functions

Our team also had the Girls Who Code club interact with a prototype of EMAR. The participants  began by exploring how EMAR spoke and were able to type in responses that EMAR could use to respond to a stressed teen.

The participants  came up with responses from EMAR including motivational quotes and jokes. They expressed that they wanted EMAR to have an option for teens to click on its belly tablet or tell the robot that they want to hear a motivational quote. The girls also shared that they wanted EMAR to have a feature where it would display calming images like sunsets when they experienced stress.

Designing eyes

Our team also asked the participants  to design different types of eyes for EMAR. We asked them  to create different eyes that would look like EMAR was listening, responding to a teen that was stressed, or responding to a teen that was not stressed. Most of the girls designed eyes for EMAR that looked like it was listening. When the girls were designing the eyes for EMAR, they commonly described the preset designs of EMAR’s eyes as being “cute.” The girls also shared that it was important that EMAR’s eyes did not look “creepy” so that students would be willing to talk to it when they are stressed.

Huggable EMAR

At the end of our activities with the Girls Who Code, we asked the girls to share how they wanted EMAR to evolve in the future. One of the suggestions that came up during the discussion was the idea that EMAR should be a huggable robot, as many of the girls believed that hugging a social robot would help with their stress. The idea of a huggable robot is a theme we’ve heard repeatedly when working with teens.

However, one of the girls shared that the dirt and germs EMAR would accumulate from being hugged by students would be a major turnoff for her wanting to hug EMAR to relieve her stress. This brought up an interesting issue that our team hadn’t yet explored. While the idea that a huggable social robot will eventually incur physical damage by being a part of the classroom over a long period of time, not much has been said specifically about the decline of cleanliness the robot would incur over that period of time.

Some of our questions moving forward include:

• If dirtiness is a turnoff for teens when it comes to hugging a robot, then how will the robot be kept clean?
• Who will be tasked with cleaning the social robot?
• How will maintaining the social robot affect the relationship between the students and the robot?
• If the robot deteriorates enough that it needs to be replaced, how will that affect the relationship?
• Would students have the same relationship with the new version of the robot as they did with the old one?

Next steps

While these questions are beyond the current scope of our research group, our team is looking forward to exploring the idea of EMAR being huggable in the future.

Sharing these activities with the Girls Who Code helped our team better understand how teens envision themselves interacting with a social robot that would measure stress. We are looking forward to sharing the activities we created with more teens in the future.