This week’s topic was about how we approach research ethics and care for our participants. We also learned how we ranked the risk level based on the Falmouth University Integrity and Ethics policy.
The articles and the references provided in this week’s bibliography include one about Facebook’s emotional contagion study; this led me to think about the risk level of my day-to-day research works.
Indeed’s researchers and data scientists conduct researches in many ways, from market research and data analysis to usability tests. I occasionally work with them to run several types of research to understand the behavior of our users. Hence, to understand their pain points using our products to improve.
I wanted to use this week’s reflective journal to assess the risk level of my researches at Indeed.
I usually run two types of researches at Indeed – in-person research with our researchers and usability research through usertesting.com.
The in-person research involves recruiting our existing employers from our database and contacting them to participate. We often give them credit to use on our website for sponsoring their jobs. We will then invite them to our research lab, where we set up a camera to record the user’s facial expression and desktop screen recording or a camera over the mobile device to capture both the screens and their hand gestures. The devices the participants operate on are provided by us. The researcher will moderate the sessions and ask them to complete different tasks based on our planned topics.
I would rate this type of research as medium risk:
- Research involving individuals
- Research involving access to records or personal or confidential information concerning identifiable individuals
- Research involving interaction with individuals where different cultural perceptions of ethics might result in misunderstandings
To minimize the exposure of private information, we will then anonymize the profile of the participants (real name, company name, gender, and age). However, we would occasionally release video clips of their facial expressions or quotes in video format that may be identifiable in our internal reports. For public releases, we would transcript the facial expressions and quote instead.
The second type of research is usability tests on usertesting.com. Compared to in-person research, we collect less information about the individual. We have access to the screen recordings and anonymized identifiers with their age, gender, and incomes unless their obligately told us (usually their company name and job title). However, I design the study plan with questions that do not lead them to disclose such information.
I would rate this type of research as medium risk as well:
- Research involving individuals
- Research involving access to records of personal or confidential information concerning identifiable individuals
- Research involving interaction with individuals where different cultural perceptions of ethics might result in misunderstandings
One tricky thing about this type of research is that they often use their personal or work devices. They might reveal some info through other applications like emails and photos during the screen recording. That information may disclose their or other person’s information. When sharing the video internally, I would often trim the recordings so the clips would not contain their private data.
In conclusion, conducting research can help us understand our users and provide valuable insights to design our product. I also agreed that we should consider research ethics and minimize the risk of collecting unnecessary information.