Blog Update #6a – Pilot Test:

In our first pilot, we found improvements to our protocol and medium-fidelity prototype. We discovered the running time per participant was only half an hour, which eased recruitment. Verbal instructions were clear. The participant did not require additional help in the onboarding experience. However, the participant relied heavily on the prompts we provided, and felt they had to use those words specifically, rather than being able to use variations. We realized Facebook timestamps were not specific enough for conversation length, and used a stopwatch after the first conversation. We established that the webcam was not intrusive, and participants forgot about it.

For the second pilot, we made minor changes to the instructions in the onboarding experience. We added labels to the social burnout images and the friends page to facilitate user understanding of the images. We renamed the Social Burnout page after receiving feedback that it was abrupt. We finessed the experimenter roles on our end. The process was more smooth in the second pilot, and the team has a better understanding of their roles and responsibilities. We had forgotten to update the screenshots in the survey for Pilot 2, so updated them before commencing the experiment.

Blog Update #6b – Experiment Abstract:

We introduce a chatbot to Facebook Messenger, called Social Butterfly (SB), that improves socializing by reducing planning time in online chats. SB is designed to assist event planning by parsing chat prompts from conversations, and making informed suggestions based on user interests and schedules. We were particularly interested in the benefits of SB for different personality types (introvert and extrovert). To demonstrate the benefit of SB we conducted one experiment with a total of 10 users to show: (1) that SB improved performance in planning one-on-one events, (2) that trends emerged that SB can broadly improve performance in a variety of event-planning scenarios and (3) that introverts may be more satisfied using SB. The trends observed with SB are encouraging, and suggest that with further research and refinement, a social-assistant chatbot could greatly improve socializing for both introverts and extroverts.

Blog Update #6c – Revised Supplementary Experiment Materials:

Blog Update #6c – Revised Supplementary Experiment Materials:

There were no changes required in terms of any documents such consent forms and call for participation. During our lab, we conducted a demo on our prototype and from the feedback we received, changes were made in our prototype and user satisfaction survey accordingly.

Some of the alterations added to our prototype were as follows:

1. A back button was added to our onboarding experience screens. This was to ensure the fact that while adding their preferences, users can go back and forth between different pages to alter their preferences.
2. For one of the onboarding experience screens, we decided to rename the title from “Social burnout” to “How much me-time”. The reason for this alteration was considered due to the fact that it was hard for the users to grasp the concept of social burnout. Additionally, on the same screen, captions were added to each image so the users can comprehend the meaning of each image.   
3. Additionally, changed instructions on Social Butterfly prompts page (last screen in onboarding experience) to indicate that they are suggestions or a special set of Strings that will catched by the bot in order to plan an event within that chat.  
4. Lastly, the fake profiles created for the experiments were given names that creates a scenario where actual people are interacting since we did not have profile pictures set up for each individual taking part in a conversation with the participant.

Our user satisfaction survey conducted consisted of some screens from onboarding experience, were also according to the changes mentioned above.