Materializing Data Bias(7)| FMP


🤝Teammates: Tiana Robison, Jinsong (Sylvester) Liu,Luchen Peng.

According to the previous workshop, we realize that having participants make objects individually will be limited to their past experience and cognition, and lacking deas and assessments from different knowledge backgrounds in this process. It’s easier for the output to contain the unrecognized bias integrated from individuals, and lack of inclusiveness and debate space.

We decided to iterate the workshop according to the previous findings, letting participants make critical objects together with the changes in assigning three roles (designer, data scientist, and general public) in one group.

Participants in each group. ( Image: Sylvester)

Workshop flow with details:

Workshop flow with details ( Image: made by group)

Material iteration:

During the previous workshop, participants often need to make shapes to express the function. However, participants who are not major in design-related often rely on our help to make their idea physical. The use of handicraft materials, such as pipe cleaner and color paper, will hinder participants to make. because they need to spend a lot of time thinking about how to make a 3D shape, and they are often dissatisfied with their works, thinking that what they make can’t convey their imagined functions. Therefore, it is important to iterate materials.

Pondland gives me lots of inspiration. Some materials, such as clothespin, are ambiguous in form, which could easily be deconstructed, repurposed, or reconfigured from their original use. Some materials, such as wigs and transparent catheters, show the biological properties of medical field. I like the bule bathroom floor mat, its texture and color show the sense of future.The practice proved that they were attracted by various materials and used the materials creatively.

Some materials bought from pondland, ( Image: me)

Groups’ Outcome:

The first group’s work called CV Checker.
(Credit: Eda, Esther Zeng and Siqi Shu)

The first group builds a CV checker according to their past job-seeking experience. They express that the CV filter will filter resumes according to keywords. However, it is unfair to filter out those job seekers who have outstanding abilities if their resumes lack some keywords.

The second group’s work called Walkaware. (Credit: Pranjal,Svaney Shen and Damul Yang)

The second group made a wearable device. They express predictive policing algorithms are more and more popular. The people whose faces are like Muslims captured by CCTV will have high opportunity predicted as a potential criminal offender. Therefore, this device could identify the hidden cameras all around users. When they detect the camera, they will say “ watch out!” and will trigger a facial converter to generate a face with a low crime risk score to trick algorithms.

Workshop Feedback:

Workshop Flow and Feedbacks in each step ( Image: Tiana)


Our workshop is successful but also has disadvantages

The successful part is that we let more participants understand what is data bias and how it affects society. Many participants indicated we provide a good experience to participate this topic discussion.

The failure part is that we didn’t let this topic go further. It is universally acknowledged and indisputable that we all need to resist data bias. However, there are more ethical topics that the public can participate in under this topic, like how to balance the promotion of fairness with the right to privacy when collecting diversified data. Compared with throwing a huge topic “data bias” to the public, this kind of topic has more space for discussion.

What’s more, in the group critical making process , we still lack right and further guidance, such as how to stimulate them to be equal contributors in the design process. One of the participants said that “as a general public, I felt helpless and stressful, because I obviously felt lack of knowledge and experience on this topic, maybe it is better to add more general publcs in one group”

UX MA Student in LCC