Research + Design + Strategy
ABOUT PROJECT
This project was undertaken as part of our Understanding Virtual Social Behavior module wherein we studied the virtual behaviors which are exhibited on YouTube with specific regarding learning channels.
We also looked into the behaviors exhibited by different stakeholders, the drivers responsible and other occurring phenomenons.
​
APPROACH
METHODOLOGY
The methodology adopted helped us map out the gaps to eventually tap into insights and hypothesis those were further validated via experiments and simulations. Based on these, design directions/ recommendations were proposed.
1. AREA
2. RESEARCH
3. ANALYSIS
5. IDEATE
Secondary research
Online
Learning
Validation
Literature review
Case studies
Netnography
Auto-netnography
Word/ picture
association
Survey
Insight
Design recommendations
Hypothesis
Primary
research
4. OPPORT UNITIES
SECONDARY
RESEARCH
-
Intended purpose: Video sharing platform.
-
Started in February 2005 by three former employees of PayPal- Chad Hurley, Steve Chen and Jawed Karim. In November 2006 it was bought by Google and now operates as one of its subsidiaries.
-
Currently, it is the largest search engine after google.
1 billion
average number of mobile
YouTube views in a day.
70% of the Indian users are below the age of 35
300 hours of video uploaded per minute.
70% increase in the search of “ How to” videos year on year
ONLINE EDUCATION India
AFFORDABLE
Online education is much cheaper (almost 53%)
GOVERNMENT
Plans like digital India and other policies are enabling.
MOBILE PENETRATION
Mobile/ Internet penetration: New mobile users by 2012 expected to be 180 M and 735 M Internet users.
EMPLOYABLITY
Industry or job relevant training can be acquired.
AVAILABILITY
It is easily accessible and available. (even where offline education is low)
DISPOSABLE INCOME
Disposable income: Expected to grow by 55% by 2021.
Virtual social platforms have the power to amplify attitudes, behaviours and motivations, compared to that which exist in the real world.
​
s
Quick and easy access
Anytime- anywhere usage
In synch with the ‘digital age’
Geographic liberation
Equity and equality
Abundance of content
Cost effective (free)
Boundless
2nd largest platform in the world
Device/ gadget friendly
Enables sharing
Flexibility and ease of choosing time
STRENGTH
O
Increased internet penetration and Infrastructure
Policies in favour
Future of education
Cost effective
No limitations
No limit of no. of students
Spread education beyond borders
Discrimination free
Large community
Schools without classroom
Existing ecosystem of online learning
OPPORTUNITY
W
Lack of feedback
Lack of collaboration and peer learning
Lack of authenticity and credibility
No assessment of learning
Prone to bullying
Intrusive and distracting ads
Low moderation
Lacks transparency
Traps/ click baits
Conformity bias/ discrimination
Problem of plenty
Low retention
Distracting
WEAKNESS
T
Consistency
Uniqueness
Students tend to get laid back
Procrastination
Alternative learning platforms
Classroom education
Distance learning programme
THREAT
UNDERSTANDING VIRTUAL BEHAVIOR
.jpg)
netnography
​
Netnography is an online research method originating in ethnography which is applied to understanding social interaction in contemporary digital communications contexts.
content uploader
.jpg)
Kurzgesagt: In a nutshell
Subscribers: 9,209,503
Areas of education: General science learning
Content uploaded: Graphic Animation / Gamification
-
Explain the process to the audience to gain their trust, provide source information
-
Comments were more relevant and discussion-oriented.
-
The comments section was interactive and deviations were less.
-
The quality of the videos was high and engaging.
More learning-oriented.
How do content uploader gain the trust of the viewer
-
Making viewers aware of the benefits/ outcome
-
Sharing the channel’s success stories
-
Providing genuinely rich content
-
Sharing their sources to validate their content
-
Collaborating with trusted sources
-
Open source/ providing further reference
-
Based on the no of likes and viewership
-
Clickbait/ Taglines
-
​Tapping into people’s greed (ex: “crack UPSC in 2 days.”)
Discrimination on online platforms
FEMALE
Comments are oriented towards the educator’s physical appearance and looks.
They are lued, demeaning, and sexist.
​
Most of the comments are highly deviant and irrelevant.
​
The viewers also try to flirt and ask inappropriate questions.
MALE
The comments very relatively more relevant than they were for the women.
​
The comments were de-motivating and the sarcastic.
INANIMATE
The comments very the closest to being relevant.
​
The discrimination and distraction were seen to low.
​
There were more relevant discussions.
​
The videos were more engaging and had consistency of delivery.
Behavioral Archetype
real
Fear
Guilt
Fraud
Shame
Embarrassment
Altruism
Gratitude
Aggression
Discrimination
Bullying
Transaction
Self
presentation
BEHAVIOR
PLOTTING
Human behavior from the real world often manifests in the virtual one. A comparative plotting of certain reoccurring and common social behaviors was done between these two
virtual
mapping
behavior
Aggression
Youtube is available for everyone without any restrictions. People can keep their accounts private/ unknown and share their feelings without any inhibition of being judged. Moral values and the line of social ethics also blurs. When people’s beliefs are challenged, it triggers them and amplifies their anger leading to aggressive comments.
Transactional
Like, share, subscribe! Probably the first three words that come to anyone’s mind on hearing, ‘YouTube’. Even the educational platforms are highly driven by this. The aim becomes to gain likes and views and monetize on the content. The intent of teaching and educating dilutes and gets overshadowed.
Cyber-bullying
Cyberbullying and trolling are a highly common occurring in the comment section. The feedback is highly deviant. Multiple reasons can be attributed to the same. Some being the anonymity, no defined consequences, unaccountability, social conformation and normalcy, they feel this is acceptable because everyone is doing it.
Polarisation
One’s views or behavior tends to get extreme or polarized on the platform. One of the key reason for this is confirmation bias. People likely seek out and agree with views that align with their pre- existing beliefs. Curating the content we want to see potentially makes it easier for us to listen to speakers/ educators who validate our worldview
Discrimination
Discrimination is another fairly recurring phenomenon. This happens against the peer who comment as well as the ones who are uploading the video. The most common one being that based on the gender and then ethnicity. Women educators receive more irrelevant and inappropriate comments compared to men.
Exploration
With 300 hours worth of video being uploaded on YouTube, there is absolute abundance of data one can access, due to youtube’s anyone- everyone approach. With the recommendation algorithm of YouTube is extremely engaging and results in amplification of usage.
mapping
drives
It allows the viewers to express without fear and gives a feeling of security.
It encourages bullying, aggression, nasty and inappropriate behaviour and comments.
Anonymity
Gives people a chance to control or shape how others see them.
This leads to manipulation, inauthentic content, deceitfulness. The uploaders trick the viewers into believing what they desire and mostly mislead.
Self presentation
It provides a sense of self surety and confidence.
It makes the viewer irrational, blindsided and often leads to mob behaviour like bullying and also harassment.
Conformity
One does not feel restrained while communicating.
​
​
​
​
This also turns toxic and leads to bullying. This enables viewers to say mean and inappropriate things, having no consequences to face outside the internet.
Online Dis-inhibition
There is a sense of belonging and being a part of the group.
It makes one highly susceptible to being influenced, promotes herd mentality, loss of authenticity and suppresses new ideas and opinions.
Group thinking
It increases authenticity as people express themselves better and put honest opinions. They engage better when they trust the uploader too.
It can be misleading, people try to assume maximum information with minimum information. It makes the viewers baised and polarised.
Trust
CORE
drives
Drive to acquire likes, views, fame, recognition, influence and eventually money is the key driver. Their purpose of educating or teaching is highly transactional. This is followed by drive to feel. This deals with the feeling of acceptance, thrill and emotional experiences.
The Uploader
The most common drive that was observed was drive to bond. Viewers mostly want their opinion be valued. People incline towards individuals with similar worldviews and demographics. This follows drive to defend, which triggers them to become active when they feel threatened by another opinion or a person.
The Viewer
Drive to learn: encourage curiosity and provide means for exploration to improve understanding. This should appropriately have been the key drive for both of these stakeholders as it the very essence of education and learning. Due to overpowering of other drives, this takes a backseat.
Both
SURVEY
KEY FINDINGS
-
The majority of the participants use YouTube for learning or education.
-
The majority ends up spending more time than planned.
-
Features like subscribe, like, playlist are the most commonly used while, comment section and reporting lags.
22.7%
77.3%
Do you use youtube for educational / learning video?
On the scale of 1-5, rating how much they trust the authenticity of the video the most common was 3. Implying that the majority does not trust the content completely.
13.6%
72.7%
13.6%
Do you end up spending more time than you planned when you use Youtube?
KEY
OBSERVATIONS
-
The feedback on YouTube is delayed and weak.
-
The current comment section discourages people from participating and is also misused.
-
There is a ripple effect (One negative comment/ video leads to more)
-
Ratio of the no. of views to comments is extremely skewed.
-
The relationship on YouTube is highly transactional (likes, shares, etc)
-
The uploaders gain users trust by tapping onto their greed.
-
People tend to open up to people who are perceived as anonymous.
-
There is discrimination based on gender, caste, region, etc.
-
The viewers are unable to navigate themselves through the ample amount of data available hence, confusion.
-
Drive to learn takes a backseat on learning channels.
-
Recommendations are distracting.
Mild to wild
YouTube’s recommendation algorithm polarises and amplifies behavior to extremes.
Perceived lack of accountability
Disinhibition leads to more irresponsible behavior
Authenticity
The anyone- everyone approach is diluting the authenticity.
Element of play
Lack of element of play leads to less engagement/retention.
Excess of data
The abundance of data leads to confusion.
INSIGHTS
HYPOTHESIS
VALIDATION
Recommendation Algorithm
On the scale of 1-5, rating how much they trust the authenticity of the video the most common was 3. Implying that the majority does not trust the content completely.
Result 1: It helped us to establish that the existing algorithm pushes viewers towards engaging and controversial videos.
​
Result 2: It helped us to establish that the existing recommendation algorithm polarises people and pushes them to extremes.
Feedback
The comment section biases viewers opinions and currently is irrelevant and delayed.
The simulation helped us to establish that comment section create biases amongst the viewers and sometimes result in a stalemate condition where the uploaders efforts and intent get countered by the negative, irrelevant and discouraging comments.
Retention
Retention of a learner is lesser on YouTube learning channels.
The average score of participant group 2 was more than that of 1 as they retained more information. The delayed feedback also affects the learning.
DESIGN DIRECTIONS
The most crucial, primary and the broader recommendation is to create a subcategory of YouTube, making a more relevant platform for learning, suggestively calling it YouTube ed.
Recommendation two and three are feedback framework and recommendation algorithm respectively. These two are essentially a part of the broader suggestion, the new proposed platform.
SOLUTION 1
YouTube Ed
TRAITS:
Conscientiousness (Reliable, prompt, organized thorough)
Participative
Rewarding
Protective
Ethical
Screen 1: Sign up/ sign in
WHAT:
-
YouTube ed would require a separate sign up.
-
The learner would have to provide their email ID and Phone number to which they would receive an OTP to verify the user.
WHY:
-
This would help address anyone- everyone approach.
-
Bring learning for the forefront.
-
More relevant and learning oriented content (for both the uploader and the viewer)
-
Keeping the anonymity intact, the verification process via OTP would make the viewer feel more credible, accountable andless invincible.
Screen 2: Subject preferences
WHAT:
-
Giving viewers an option to choose their preferred subjects that they want to learn.
WHY:
-
Planned and focused learning.
-
Viewers in authority to choose what content is shown or he/she wants to see.
-
Aid them learn WHAT THEY WANT to learn.
-
Negate confusion in the abundance.
-
Relevant and customised content.
Screen 3: Main screen (video playing)
Feedback is an extremely crucial part of communication. It is a window to the future. Yet, the feedback on Educational channels of YouTube is delayed, skewed, deviant, unstructured, unengaging, monotonous and extremely irrelevant. Also, it creates a negative comment/ discussion atmosphere due to elevated discrimination, aggression and other sentiments.
SOLUTION 2
FEEDBACK
Inclusive: Most people do not feel involved in the feedback/ comment section. It is a one-way communication as the uploaders don’t involve in the comment section.
Engaging: The current feedback section has no engaging factor and there is no perceived gain or value for their time they contribute to giving feedback. Based on the social exchange theory they weigh their loss of time higher over the return they get for the same hence, there is a lack of feedback.
Prompt: The feedback on YouTube is delayed. This causes the negative sentiment to sore further. Rapid replies can help counter this and therefore response time matters.
Relevant: The feedback should be designed such that it should minimise deviations.
WHAT:
-
Replacing the comment section with the discussion forum as a method of feedback.
-
Adding a quick and an alternative way of giving feedback via emoticons.
-
Removing the dislike options while only keeping the like or the appreciation button.
-
Gamifying the feedback to add the element of play and introducing incentives
-
Via gamified levels and badges to show the competency of the commenter in the discussion forum.
WHY:
-
Comment is perceived to be more casual whereas, discussion forum is perceived to be more crucial. This will remove unnecessary and random comments, making it more relevant.
-
Viewers can provide instant feedback.
-
Removing lingual barriers hence, inclusive and prompt.
-
Makes the viewing environment less toxic.
-
Engage people.
-
Through gamification, feedback to the viewer about their learning progress.
-
Credibility of the commenter
SOLUTION 3
ALGORITHM
​Existing Algorithm
​
​
-
Focuses on watch time rather than the quality and content of the video.
-
Mild to wild.
-
It pushes viewers to extremes.
-
It creates polarity.
-
Amplifies the behavior.
-
It becomes a trap to engage people and amplify their sensation.
​Desired Algorithm
​
​.
Proposed Recommendation Algorithm
Ranking algorithm
Contextual - Demography and language
​
Setting the right context by understanding the basic and preliminary requirements of the user
Users Skill level and requirements
​
User skill level will segregate the content according to the users current understanding hence making is relevant
Users history and Peers view list
​
Creating an environment where user's history is used for self-analysis and retention. Peers view list will further make the learning more contextual
Video quality and presentation
​
Through our research, we found that audio and video quality of the content affects the level of trust.
Trending
​
Through our research, we found that audio and video quality of the content affects the level of trust.