░░░░░░░ WEEK 2 ░░░░░░░
First Consult
During the initial consultation with Andreas, I made a concerted effort to provide him with comprehensive
updates on the modifications I had implemented in my project. The primary focus of this session was to
gain insights into potential avenues for presenting my work effectively.
While I had initially brainstormed some options during the first week, upon reflection, I realised that
these options did not fully capture the depth of my project's. Andreas helped to delineate the four
primary elements central to my project: design, software, narrative, and experience. His suggestion to
document these elements in a publication resonated with me, as I had already contemplated this approach.
However, the practicality of a publication in the context of a graduation show or open studio setting
posed a challenge. Having attended similar events, I observed the pace at which attendees move and the
often casual nature of their interactions with the showcased works. Therefore, if I opt for a publication,
it must not only encapsulate the essence of my work but also be engaging enough for viewers to absorb its
content fully.
Personally, I find publications with fewer than 20 pages less appealing, coincidentally the probable page
count for my publication. As a result, I am exploring alternative options, such as a folder containing
loose sheets of documentation or devising strategies to increase the page count. Despite these
considerations, I believe that having a physically printed outcome would contribute an additional layer to
my work, serving as a tangible representation that showcases the intricate layers of thought embedded in
the project.
Andreas' Notes
Thought Experiment Planning
As I’m planning to conduct the first round of my thought experiment in Week 4, my focus for this week will
be planning the experiment flow and preparing the collaterals needed for it. The thought experiment (alpha
test) would make up the main contents in my dissertation discussion section, as well as provide me a
conclusion to my research objective.
The purpose of the thought experiment is to:
1. Enhance algorithm literacy through interaction with the unethical software prototype
2. Evaluate the efficacy of speculative design methods in gaining understanding of complex topics such as
social media algorithms
Enhancing Algorithm Literacy
Within the thought experiment, various components intertwine, each contributing distinctly to the
cultivation of algorithm literacy. These components are intricately designed with a particular emphasis on
addressing the three algorithmic dilemmas spotlighted in my dissertation—namely, algorithm bias, filter
bubbles, and algorithm transparency. This deliberate focus ensures a comprehensive exploration of these
critical aspects within the context of the thought experiment.
1. Narrative
The narrative woven into the thought experiment centers around Inclusivision, a storyline carefully
constructed to unfold during the experiment. As the experiment kicks off, I'll reiterate the software's
purpose and its pivotal role in addressing the issue of filter bubbles. This initial narrative touchpoint
serves to underscore the hazards associated with filter bubbles and accentuates the significance of
inclusivity and diverse content within the realm of social media.
In a strategic move to stimulate critical thinking, an additional section will be integrated, revealing
Inclusivision's future aspirations. This forward-looking segment details the software's plans to
collaborate with major social media platforms. This foresight is strategically positioned to prompt
participants to contemplate the effects of unethical algorithms in the real world as they engage with the
software later in the experiment.
2. Software Front End
The front end of the software gives the participants an idea of how the software works. They get to
interact and experience the software firsthand. Within this phase, participants are encouraged to discern
the nuances, spotting potential inaccuracies and inconsistencies in the facial detection algorithm. Any
participant observations regarding inaccuracies prompt an interactive exchange. I explain the underlying
factors contributing to these discrepancies, emphasizing the deliberate trade-off between processing speed
and accuracy. Subsequently, participant reactions are meticulously documented, setting the stage for later
evaluation.
I’m still in the process of adding more elements into the front end of the software. The front end is
important as this is where I need to retain the suspension of disbelief. I’m planning to add a few more
features, such as simplifying how media can be uploaded into the software, and also transitions between
imported media and screen recording modes. These augmentations contribute to a more immersive and
believable participant experience.
3. Software Back End
Delving into the intricacies of the software, the back end reveals the algorithm's decision-making
processes, particularly in determining diversity through a biased and unethical measurement of pixel
values in detected faces.
With a clear understanding of the software's purpose, participants are prompted to engage in critical
thinking. Anticipating diverse reactions, I look forward to discussions on potential algorithm abuse,
skepticism regarding the software's efficacy in promoting inclusivity, and considerations of its broader
impact on the social media landscape upon going live. These discussions serve as valuable indicators of
participants' algorithm literacy levels, setting the stage for comprehensive evaluation at the conclusion
of the experiment.
4. Question Prompts
Before the thought experiment concludes, a group feedback sharing will be conducted. This session serves
as a fallback if the desired reactions were not obtained in the previous phases, or contained a lack of
indication of algorithm literacy.
I’ve prepared a list of questions prompts to initiate discussions regarding algorithm dilemmas, which will
also give me more insight into participants’ algorithm literacy levels.
Thought Experiment Flow
Proof of Enhanced Literacy
To ensure there’s a change in algorithm literacy before and after the experiment, data has to be collected
before, during, and after the experiment for comparison. Since there is no existing scale to measure
algorithm literacy, in the following sections, I’ll walk you through my thought process in devising one
that will work in this particular experiment.
Indicators of Algorithm Literacy
Building on Leyla Dogruel’s paper ‘Development and Validation of an Algorithm Literacy Scale for Internet
Users’, I extracted the indicators of algorithm literacy in individuals.
Due to algorithm literacy being such a complex topic with a ton of underlying layers, these indicators
ensure that the key components in each layer is accounted for. This creates a fair system that prevents
discrimination against individuals who lack a certain understanding in certain components, but excel in
the rest.
Links:https://www.tandfonline.com/doi/full/10.1080/19312458.2021.1968361
Algorithm Literacy Evaluation Chart (ALEC)
The ALEC is a measurement chart that I created to assign scores to participants’ algorithm literacy levels
based on their responses. This allows a quantitative analysis of participants’ algorithm literacy before
and after the experiment, which informs the impact of the experiment. The ALEC is also designed to
streamline documentation and analysis further in my dissertation discussion.
Preliminary Survey
Before the thought experiment begins, a preliminary survey is issued to each participant. This survey,
disguised as a social media literacy survey, gauges participants’ algorithm literacy levels before the
experiment. The survey responses will then go through qualitative analysis for indicators of algorithm
literacy. Each indicator will be ranked 1-5 in the ALEC, with 1 being limited understanding and 5 being
advanced.
This survey is carefully designed to the context of the alpha test, carefully threaded to not expose the
thought experiment. It is essential to emphasize that the initial determination of algorithm literacy
levels relies exclusively on the inference drawn from survey responses, given the indirect approach to
addressing algorithm literacy in the questions.
Another crucial purpose of the survey is to help me assign participants to different experiment sessions.
This helps me to achieve a more diverse range of subjects with varying algorithm literacy levels in each
thought experiment session.
Process Reflection Tool
The Process Reflection Tool serves as a structured method for documenting and evaluating the thought
experiment. During the experiment, a video camera and mic will be set up to document participants’
reactions. The documentation will then be analyzed and recorded in the process reflection tool at the end
of the experiment.
The documentation in the process reflection tool will undergo another qualitative analysis, which will be
recorded in the ALEC. Both results from before and after the experiment will then be compared to
illuminate the impact of the thought experiment (how much algorithm literacy growth).
Indicators of Algorithm Literacy
Algorithm Literacy Evaluation Chart (ALEC)
Location Sourcing
Choosing the right location for the thought experiment is crucial for maintaining suspension of disbelief.
To ensure this, I cannot utilize school studios. An ideal setting would be an office space with ample room
for cameras and individual laptops, along with a presentation area. Controlled lighting is essential for
BTS photo documentation of physical outcomes. Just in case, I’ve asked for Andreas’ help to reserve a room
in Wilkie Edge if I can’t find a location by Week 3. Despite challenges in securing shared workspaces, a
friend has offered a suitable office space that aligns with the experiment's requirements.
I visited his work office to inspect the space, finding that its tech-industry aesthetics aligned with my
desired environment. The office boasts lighting equipment, spare computers, monitors, and adaptable
makeshift tables, providing ample resources for my experiment setup.
During the location recce, I shortlisted 2 possible spaces for me to conduct my thought experiment:
1. Recording Studio
The recording studio's grey wall complements Inclusivision's brand, and its modular table allows
flexibility based on participant numbers. The soundproof studio ensures clean audio recordings, and
abundant lighting equipment and power points enhance practicality. Despite a limited capacity for up to 3
participants, the minimalistic environment minimizes distractions, creating a focused atmosphere.
2. Showroom
The showroom, with pre-setup computers, alleviates the need for additional laptops and monitors for
participants. Its tech-centric ambiance also complements the fabricated company narrative, integrating the
experimental setup within the displayed technological environment. While the showroom offers more space,
its fixed furniture limits set design flexibility, posing constraints on capturing diverse camera angles
to document reactions.
After giving it some thought, and weighing the pros and cons, I decided that the studio would be more
feasible.
Dreamcore Studio
Dreamcore Showroom