Socratease
Wisdom at your finger tips
Mobile application that curates news and simplifies the process of exploring multiple perspectives.
If one were to attempt to identify a single problematic aspect of human reasoning that deserves attention above all others, the confirmation bias would have to be among the candidates for consideration.
- Nickerson, R. S. (1998)
What is confirmation bias?
Confirmation bias is the tendency to search for, interpret, favor, and recall information in a way that confirms or supports one's prior beliefs or values.
Literature Review
A probing literature review was done to understand different types of biases and they are mapped against different contexts to understand their affect on decision making.
Insights from literature review:
Confirmation bias is the most prevalent cognitive bias.
Problem Framing
Further research was carried out to narrow down the problem area and arrive at below problem statement.
Problem Statement:
Social media is showing the end user what they are likely to agree with and excluding the alternative viewpoint. This is resulting in the formation of filter bubbles, a state of mental isolation.
HMW:
How might we provide alternative perspectives to social media users so that they can inculcate better reading habits while browsing social media?
Action Research
An action research focuses on producing actionable input, rather than theoretical input. For this project, a set of 10 cards, along with 1 CTA (Call to Action) card, were created as shown below.
5 out of the 10 cards (shown on the left) have questions that will lead to answers proving the hypothesis on CTA card to be true. A circle and a QR code were printed on the back of these cards. It says "If you have more circles, please scan below QR to know how India occupied sikkim."
The other 5 (shown on the right) have questions that will lead to answers proving the hypothesis on CTA card to be false. A square and a QR code were printed on the back of these cards. It says "If you have more circles, please scan below QR to know how India taught non-violence to the world."
Participants were asked to pick 6 out of 10 questions to verify their opinion about the hypothesis on CTA card. Once they selected their cards, they were asked to flip them. This is supposed to break their tendency to look for confirming information. No futher prompts were given. After a brief observation, they were interviewed.
Observations:
-
1 participant showed extreme confirmation bias
-
2 participants showed moderate confirmation bias
-
2 participants showed no confirmation bias
-
1 participant broke the confirmation bias
-
They are not excited to scan the QR
Insights:
-
Participants get bored by limited narratives (1 in this case)
-
Participants experienced a high cognitive load
Number of participants: 6
Note: The sample size is very small and statistics may vary with a larger and diverse pool of participants
Pre-Prototyping
Building on insights from action research and brainstorming solutions to reduce cognitive load, I conceptualized smart widgets synchronized with users' mobile devices to monitor and manage social media and news consumption. These widgets provide real-time, cross-platform insights into users' digital habits. By tracking and analyzing the type of content consumed, the widgets encourage users to explore varied sources, thus mitigating confirmation bias and breaking filter bubbles. Customizable settings allow users to set goals and receive reminders to engage with a broader range of information, promoting a more balanced and informed digital experience.
A desktop widget inspired from the metaphor "Break the Bubble". The size of the bubble increases when user consumes a topic from same perspective. It will break when another perspective is chosen.
Smart goggles whose glasses keeps on increasing their opacity when user consumes a topic from same perspective. The glasses will become clear when another perspective is chosen.
Smart widgets face a significant limitation in terms of portability. Users are unlikely to carry them consistently, which restricts their utility in diverse situations where news consumption occurs. To address this challenge, I decided to develop a more accessible and versatile digital solution. A non-linear narrative was constructed around a topic (Increasing hate speech on Twitter) using Twine. This approach also addresses the issue of limited narratives seen in the card activity, allowing users to explore multiple perspectives on a chosen topic based on the choices they make. The style of conversation was inspired from Socratic method, a form of argumentative dialogue between individuals, based on asking and answering questions.
Click here to test the narrative
Feedback & Observations:
-
Participants seemed to be more engaged with non-linear narrative.
-
They reported a positive attitude towards the narrative style.
Working Prototype
The narrative was further refined for the working prototype of SOCRATEASE. Screens were designed in Figma and ProtoPie was used to create an interactive prototype which includes the following features.
Feature 01
Users can start a conversation with SOCRATEASE just like the way they talk to someone when they hear about something
Feature 02
They can choose a topic so that Socratease will pick the trending issue from the topic of their interest
Feature 03
Users can save any conversation to continue it again whenever a new perspective on the topic is observed
Feature 04
Widget to improve media literacy which pops up when user clicks on a biased headline.
The name SOCRATEASE is a word play of "Socrates" and "Tease" referring to the way the philosopher used to interact with his disciples. It has a playful and light-hearted tone that can resonate with users.
User Comments
"Question and answer section is little tedious"
"It conveniently takes me through a conversation and making it easy to understand the other side"
"Just like it's name, the app is intellectually pleasing"
"It would be better if conversation looks like a WhatsApp chat"
"It would be great if it starts a conversation on the topics I read on other apps and provide other points of view"
"Colors are very soothing to the eye"
Product Guidelines
Based on the feedback gathered throughout all stages of the product development process, the following guidelines have been formulated.
Integration of Large Language Model (LLM):
-
Philosophical Conversations: Incorporate an LLM like ChatGPT trained to emulate Socratic dialogue, encouraging critical thinking and self-reflection through questioning and conversation.
-
Contextual Understanding: Ensure the model can understand and respond to a wide range of topics, providing insightful and philosophical perspectives.
Information Processing Algorithms:
-
Concise Summaries: Develop algorithms to process and summarize information effectively, presenting users with concise and relevant content.
-
Bias Detection: Include mechanisms to identify and highlight potential biases in the consumed information, encouraging balanced viewpoints.
-
Behavior Monitoring: Implement real-time content analysis to monitor user behavior and interactions, identifying patterns and preferences.
Chat-Based Interaction:
-
Social Media Familiarity: Design conversations to mimic chat interfaces of popular social media platforms, making interactions intuitive and engaging.
Soothing Color Theme:
-
Sage Archetype: Utilize a soothing, neutral color palette that reflects the "Sage" archetype, promoting a calm and contemplative atmosphere.
Conversation Saving:
-
Save and Archive: Provide options for users to save and archive their conversations for future reference and reflection.
-
Export Options: Allow users to export conversations in various formats (e.g., PDF, text) for easy sharing and offline access.
Guided by:
Dr. Bhaumik Nagar | Shobhan Shah | Bhaskar Bhatt | Mayank Loonker