Meta/Facebook | Matthew Bracaglia

Research Assistant III at Meta/Facebook Via Vertisystem

In my last position as a Research Assistant for Meta/Facebook, I conducted Data Collection for product development. I first joined a team of 3 researcher assistants to help build the groundwork of a widespread data collection for a device Meta was developing.

For my first year, I helped develop and run the research study protocols to run data collection on 30 participants a day with two other research assistants.

During my second year, I was promoted to Research Assistant level III where I oversaw 10 other Research Assistants where I was in charge of the daily running of one of three different research facilities.

Types of Projects and Responsibilities

Basic band device data collection

Verify NDA and Consent forms, collect demographic info.
Explain study goals, rules, and expectations.
Assist with wrist device setup.
Ensure device configuration and functionality.
Begin collecting hand gesture data.
Monitor accurate data collection.
Adjust sessions as needed (e.g., 4 sessions or combine 1-3 from different studies).

Optitrack System

Calibrated and set up the Optitrac System for the study environment.
Prepared the rig with multiple cameras to capture hand movements.
Applied reflective dots to the participant's hand to enhance data accuracy.
Utilized Optitrack system to collect detailed hand movement data.
Enhanced device training accuracy through precise hand movement data capture by the cameras.

Daily Running of Research Facility

Started managing a facility in February 2023.
Responsible for daily setup of morning and afternoon data collection sessions.
Supervise and coordinate 7 other Research Assistants (RAs) in conducting research studies.
Ensure timely setup of studies to collect required data.
Provide guidance and assistance to RAs when addressing study errors or participant issues.
Collaborate with the project manager to meet new research facility requirements.

The following video is Mark Zuckerberg talking about the device that I spent two years doing data collection. The device uses EMG to read people's hand gestures to control a VR device. My job was to collect data from large populations of people with different backgrounds to train the device to be as accurate as it could be with anyone who uses it.

Main Job Duties

Lead Research Assistant: running the facility

As the lead Research Assistant, I managed a facility and led a team of 8 research assistants, guiding them through complex studies and ensuring the collection of valuable data for product development.

I oversaw the research process, from protocol development to execution, mentoring my team to enhance their skills and contribute effectively to our projects' success.

Working with Engineers

Collaborating closely with engineers, I played a key role in designing and updating study protocols, focusing on refining participant instructions for clarity and comprehension. Together, we iterated on study designs, integrating feedback to ensure that participants could easily understand and follow the research processes.

This partnership led to the development of user-friendly study environments, enhancing participant engagement and the overall quality of data collected.

Working with Product Manager

I collaborated with the product manager to coordinate new studies, aligning our research objectives with product development goals. Together, we regularly assessed the workload of research assistants, ensuring they were supported and efficient in their tasks.

Additionally, I reported the facility's needs and progress, facilitating resource allocation and strategic planning to optimize data collection.

Working with Research Assistants

Working alongside other research assistants, I coordinated the setup of various studies, ensuring that each aspect of the research environment was prepared for participant engagement.

We collaborated effectively to facilitate study sessions, assisting participants throughout the process to complete the studies successfully. Our teamwork ensured smooth operation and optimal participant experience, contributing to the acquisition of high-quality research data.

Projects at Meta

Due to the product still in development and the nature of the work, I do not have the ability to present specific data of my projects or any documentation of my work.

I worked on a number of different projects, some spanning months, some spanning several days. So below I have written descriptions of UX Research methods I used during the different types of projects I worked on.

A/B Testing

Some studies had A/B testing for the ability for the device to recognize gestures with participants, presenting two different versions of the computer interface used in hand gesture studies. Some varied between speed of gestures needed, speed of the text that was shown on screen after handwriting, or other incorporated design improvements based on past user feedback.

These variations helped both the device read the different variations of gestures and users seeing their gestures control what was on the screen. This allowed for a comparison of user preferences and performance during making gestures or performing gestures or playing games.

Some studies implemented A/B testing for modified instructions provided to participants in research studies, where Version A followed the standard instructions, and Version B incorporated refinements based on past participants' user experiences. The comparison helped identify the version that led to clearer understanding and improved compliance in gathering clear data for device development.

User Interviews

Conducted one-on-one studies with participants to set them up on the Optitract system, setting them up and explaining the instructions. As they went through the study, I talked with them about their experiences with the research study, exploring the challenges they faced making small gestures, and their impressions of the hand gesture activities.

Engaged in handwriting activities to understand participants' perspectives on the effectiveness of the computer in accurately capturing their handwriting, probing for feedback on ease of use, and soliciting detailed information about any strain experienced during the activities.

Surveys

Most of the research studies had surveys that the participants filled out each research session, I helped them answer the questions and clear up any confusion they had.

We were running anywhere between 1-5 different studies each day, each with different types of questions they had to answer. Some surveys were connected to the studies, some were given to them afterward in a different room.

The topics of the survey questions ranged from the perceived difficulty of specific hand gestures, gauging their level of discomfort or strain, and obtaining structured feedback on the computer's accuracy in recognizing gestures.

With each new study, we also talked to the engineers to give them feedback on any confusion participants had with the questions, to have them reword questions with confusing wording.

Usability Testing

Many of the studies had Integrated usability testing into the research sessions, where we had to observe participants in real-time as they engaged in hand gestures and activities on the computer, identifying any usability issues they had in understanding the study instructions, and refining the studies based on immediate feedback.

With some studies, we ran short runs of one or two sessions, then the data was analyzed to see if there were ways to refine the study protocol to enhance user comprehension and engagement.

Focus Groups

Some studies had us ask participants to discuss their experiences, where participants talked with each other about their experiences with the research studies.

There was group discussion with participants who engaged in hand gesture studies, where we encouraged them to share their collective experiences, perceptions, and challenges faced during the research sessions. The group dynamic allowed for the exploration of shared insights and diverse perspectives.

Tree Testing

In some of our research studies, they were designed to evaluate the navigational structure of specific activities, particularly focusing on the device's capability to recognize gestures. Participants were assigned tasks to complete, where they were asked questions on how easy the tasks were.

Participants would navigate the tasks, backtracking if needed, until they either successfully completed a task or decided to discontinue. This iterative process was repeated for multiple tasks, with participants completing different gestures and activities.

Following the completion of tests by several participants, the collected data was thoroughly analyzed to derive insights by both the engineers and the research assistants to discuss the optimal setup for future studies, specifically aimed at gathering the precise gestures required for the development of the device.

Five Second Testing

A couple studies we conducted had participants prepare to do specific gestures, and then see a quick screen or program to do a gesture before the screen disappeared.

This type of five-second testing (the actual time varied from shorter to a little longer) to assess the immediate visual impact of the computer interface during hand gesture studies, gauging participants' initial impressions and ability to perform gestures.

We used that data to capture their attention within the brief timeframe and determine how well users could perform gestures quickly and how the band responded to quick gestures.

Work Experience

August 2021-April 2022

May 2022 - December 2022

Collaborated with a team of 2 other Research Assistants and a manager.
Established the first major data collection lab for a specific device.
Successfully recruited 30-40 participants daily for data collection.
Developed research protocols for various studies.
Organized and set up study stations in the lab.
Formulated training protocols for onboarding new Research Assistants.

Part of a small team that conducted data collection studies in Arizona and Florida.
Successfully recruited 300 participants during each 5-hour study.
Worked long 11-hour days to ensure study success.
Conducted studies to test the device's performance in varying climates and controlled for humidity and heat.
Promoted from Research Assistant II to Research Assistant III, taking on the role of a lead Research Assistant.

Anchor 1

January 2023 - August 2023

Led the expansion of the team from 3 Research Assistants in one facility to more than 20 Research Assistants across three research facilities.
Assigned as the manager responsible for daily operations in one of the research facilities.
Supervised a team of 8 Research Assistants overseeing 4-5 distinct research studies.
Managed a high volume of daily participants, totaling around 60-70 individuals.