The Discussion Around Mental Health: An Analysis of the Generational Divide as a Result of Confirmation Bias
Advisor: Mr. Kenneth Fleming
Internship Location: CityArts
Onsite Mentor: Tsipi Ben-Haim
The negative impact of social media on teen mental health is problematic; its impact on the conversations teens have about mental health however may present a larger issue. The main goal of my project is to demonstrate the negative impact of social media on Gen Z’s ability to converse productively about mental health. I hope to prove the differences in the ways in which Gen Z get their information on mental health changes how they converse about it. The research for this topic will take place in New York City, where there is a multitude of Gen Z students to survey. The effect of confirmation bias on social media and news apps is well researched. The effect of social media on mental health is well researched. But how does confirmation bias affect our understanding of mental health? Gen Z may talk more openly about mental health than previous generations, but that in no way means it is being discussed in an accurate, informed, and productive way. Older generations consume news usually from television or newspaper, while younger people tend to rely on social media for their information. This may be the reason Gen Z is having more frequent, yet less meaningful, discussions around mental health. I hypothesize that those in Gen Z who spend more time online, and get the majority of their news/info on it will have stronger opinions on mental health, and be less open to opposing opinions.
Project Blog
Week 0: Introductorion
Hello! My name is Fia Galperin. I’ve been at BASIS Independent, Brooklyn, since the 5th grade. My Senior Project aims to demonstrate social media’s negative impact on Gen Z’s ability to converse productively about mental health. I have fallen victim to “doom scrolling” for hours on Instagram and have noticed its impact on my daily life. In recent years, I have stepped away from social media for the most part. Gen Z may talk more openly about mental health, but that in no way means it is being discussed in an accurate, informed, and productive way. I hope to work on-site with someone in the news and media world to gain valuable insights about how people get their information and how social media has changed the way people interact with each other when discussing news. At the end of my project, I hope to produce a pitch deck for a less toxic social media platform that would cut out the divisions created by social media’s use of confirmation bias in their algorithms.
Week 1: Getting Started
It’s Week One! Much of this week was spent adjusting to off-campus life. I started creating a more detailed schedule for the upcoming week. I re-downloaded instagram to try and get a better understanding of my topic, and to get some more ideas for possible survey questions. I had to delete it again by the end of the week because it had once again succeeded in consuming my days with infinite scrolling. It was through instagram that I found out about the Tik-Tok banning bill, I thought it was interesting that I hadn’t heard about it outside the app, I'll start looking more at the bill next week and see how it may be relevant to my project.
On Monday, I continued the search for placement and found a few more leads that will hopefully work out. I also created a form email to reach out to those people in a consistent way. Finding a placement has proven challenging due to the unique scope of my project, but as I said, I am hoping to work with someone in the media and sociology fields.
On Thursday I went back on-campus and met with my in-school advisor, Mr. Fleming. We discussed ways to stay organized and productive despite being out of school, and solidified plans for next week. Mr. Fleming suggested I reach out to Zeynep Tufekci, a sociologist with an expertise in online activism. We listened to some of her talks in our Social Movements class.
Week 2: Diving Deeper Into Prior Research
Since it’s spring break, this week I took things a bit slower and looked back on my Senior Project Proposal, specifically the Prior Research section. I wanted to dive deeper into my sources, review methodologies, and find new sources to add to my bibliography.
Last week, I said I would investigate the TikTok ban bill, and I did just that. From my current understanding, if the company that currently owns TikTok, ByteDance, sells the company, it will not be banned in the U.S., if they do not, however, then the app will be banned. TikTok isn’t facing a possible ban because of its algorithms or its effect on young minds, it’s facing the ban because American Legislators fear it is controlled by the Chinese Communist government. I mention the reason it may be banned because it tells me that the ban is not so pertinent to my specific project. Additionally, the ban would not affect Instagram, Snapchat, Facebook, or X meaning that the algorithmic issues I am looking at will not be impacted if the ban occurs. Even if the ban impacts my research, the app would not be officially banned until six months after the bill’s passing, which is long after the timeframe of this project ends.
Week 3: Placement and Further Research
This week I received an internship at an organization called CITYarts. I will be working with them on connecting youth across the world through art. The organization seeks to bring people together through murals and community activities. This is perfect for my project because I am looking into how social media disconnects young people from each other. Working with CITYarts will give me important insight into successful ways of connecting young people not just with each other but with issues across the world. I might also help them with their social media account which will give me an interesting new perspective on apps like Instagram considering the goal would be to increase engagement and use Instagram for good and connection!
My initial research question sought to compare Gen Z with Gen X, as I continued along that path I realized that far too much generalization about the different generations would be involved in the research. This week I decided to really explore filter bubbles and the presence of algorithms in the various news sources of Gen Z such as social media and some news apps such as The New Yorker, Apple News, The New York Times etc. My initial mistake was assuming Gen X doesn’t feel the disconnecting effects of social media because they read more long-form news articles. After discussing this with Mr. Fleming (an older millennial) and other teachers, I realized that while some may be skeptical of what they read, plenty of Gen X adults trust the information they see online. It was these important discussions that led me to revise my research question to: “To what extent does the presence of algorithmic bias in Gen Z New Yorkers’ news sources affect their perspectives on and discussions surrounding mental health?”
After looking into some prior research, I learned that the presence of algorithms is not the only issue. The issue is actually their increasing extremity. For example, if I were to read one story on the New York Times app about plants, I might see more similar articles pop up in the future., But there will still be a multitude of other stories that I can easily access by looking through the app. On Instagram if a viewer shows the slightest sign of interest in a video, the algorithm throws more of it, and not much of anything else, at them. If I watch a video about the symptoms of ADHD, that might grab my attention, so I would watch it the whole way through. Soon my feed would be full of videos about mental health posted by non-professionals. This hidden bias is a huge problem, and its why youth are increasingly isolated in the info they are seeing.
The algorithmic bias in filter bubbles on social media is especially dangerous because, unlike picking up a left—or right-wing newspaper, we are not as aware of the biases.
Some questions I’m left with this week:
What are the best ways to get my survey out to my target demographic (Gen Z)?
How can Instagram and other social media platforms be used for good?
How can I take full advantage of the resources and experience at my placement?
What I hope to address in Week 4:
I hope to answer some of this week's questions and finish up a solid draft of my survey. I also look forward to updating my abstract now that my project has really gotten started!
Week 4
Week 4: Survey & Looking Into the Algorithm
This week I focused mostly on my survey. Last week I hoped to come up with some ways to get my survey out there. I think the best way to get kids to take the survey is to use the ‘devil of which I speak’, Instagram. I plan on posting a link to the survey on my social media, and as friends to do the same. One student, last year I believe, put up QR codes across NY in places like the subway. If my survey doesn’t get enough responses this may be the way to go.
As I was looking into how to answer last week's question about how Instagram can be used for good, I found out Instagram changed its algorithm in 2024. Every section of Instagram has its own algorithm, meaning that what you interact with or like on stories doesn’t really affect the content you see on your explore page or reels. Additionally, Instagram has put greater restrictions on certain content surrounding self-harm, violence, eating disorders, and content that contains restricted goods such as drugs. All these changes make no effort to restrict content on mental health. The restrictions Instagram has placed are definitely for the better, but they can be doing more. Instagram has started to prioritize original content which is great! They have started taking down accounts that simply repost other creators' content as their own, and started mixing smaller creators into the explore page, rather than only showing already popular creators. This gives more people an opportunity to use social media for its intended purpose to share your life and creativity with others.
When I started my placement I was mostly doing research on highschools around the country, but I may start to help with the social media accounts. This will definitely provide insight into the positive impact of social media, as it helps get the word out about all the amazing work CITYarts does.
Questions I’m left with this week:
How might the differences in algorithms of different apps impact the content Gen Z is seeing?
How can I account for this difference in my survey?
What I hope to address in Week 5:
I hope to answer last week's questions and edit my survey accordingly.
Week 5: Possible Interviews and Focus Group
Welcome to Week 5!!! This week I added questions to the survey that ask participants about what social media apps they use regularly. My concern when starting this project was that because there are so many social media apps, each with their own algorithms and restrictions, there would be far too much variation in the survey results. In addition to adding questions to the survey I am also going to do a few interviews with Gen Z students who regularly use more than one social media platform, students who only use one on occasion, and those who have no social media at all. I will ask the students about the differences they notice across platforms. In these interviews I hope to get a more personal look at how social media is impacting young people’s conversations around mental health. The survey will be a great tool for gathering the information I need across a large population, but the interviews will allow me to ask more specific questions, especially for those who don’t spend a lot of time online. Another option to consider would be a focus group of students. I would ask the group roughly the same questions as in the interviews, hopefully the discussion setting would allow participants to open up more about their experiences with social media’s impact on their lives.
As I mentioned last week, Instagram has recently changed its algorithm. One of the major changes was an increase in restriction on content surrounding suicide and eating disorders. While I was looking through my explore page this week, however, I saw plenty of content surrounding other mental health issues. I have found that as I try to research how mental health issues are discussed online, that my explore page has become pretty focused on mental health, because at the moment that is what I am looking for. It shouldn’t be this easy for a 17 year old to get tons of misinformation from other people who have no clue how to diagnose people.
When I was giving my Senior Project Proposal presentation, I read out the captions of a “put a finger down” TikTok that ended with a diagnosis of General Anxiety Disorder if the viewer had answered yes (put a finger down) to 5 or more questions about their habits and mental state. The video then prompted viewers to share the video with an anxious friend to “spread awareness”. The video was shared thousands of times, implying that younger audiences share the content, may self-diagnose, and further spread misinformation.
This issue is a huge concern for me, and that’s why I chose to pursue this project. I hope that the results of my survey and project overall will shine some serious light on this issue and give those who see it a better understanding of the extent of the issue.
Week 6: Survey Complete!
As I completed my final round of edits on my survey with Mr. Fleming, I felt the question format wasn’t ideal for answering my research question, so we decided to shift towards questions that used the Likert Scale that would allow me to look at the correlation between top answers and relational answers. Google Forms does not have a Likert Scale question format, so I used a scale of 1 to 5 format.
Mr. Fleming also suggested adding a question about identification with the “Gen Z label” and participants’ actual birth years. The addition of this question will provide insight into the difference in social media’s effect on members of Gen Z’s conversations, as well as those who may be Gen Z, but were born earlier on in the generation and identify more culturally with millennials.
I aim to reach over 150 responses to keep individual responses from causing large shifts in the data. I faced this issue last year in AP Research when I was researching the impact of choral training on reading comprehension skills in high schoolers. I had a very small group of participants–roughly 20 in each test group–and when testing the results, I found that if just two students had scored 100% on the test, the entire group’s average went up significantly. This leads to results that aren’t statistically significant.
As I mentioned in previous weeks, the survey is going out on my own social media, I have asked friends outside of BASIS to spread the survey around to their schools, I have put up a few QR codes in my neighborhood, and will hopefully get the QR codes up at BASIS after APs. I also messaged my brother, a freshman in college, asking him to repost the link to my survey on his social media in order to reach a wider audience. The additional questions and changes I made this week will hopefully improve data collection and get me a clear answer to my research question.
Week 7: Survey Responses So Far
In my Survey’s first week I only received 30 responses. This is a challenge I expected to face, as I faced it during my AP Research project last year. When working with a younger age group, especially during the end of the year when students are focused on finals, getting sufficient responses can be extremely difficult. I have continued to post on my own social media about the survey, and have asked others to repost the link. Out of the responses I have received so far, over 50% have identified social media as their main source of information. Although this result was expected, I still found it somewhat striking, I am eager to see how this number may change as I continue to receive responses in the coming week. In addition, all respondents use Instagram, with only 42% on TikTok. So far my data is aligning with all of my prior research, but is it aligning with my hypothesis?
In regards to mental health content, 24% of respondents said they see content surrounding mental health weekly, 18% said they see mental health content often, and 10% said they saw it daily. In addition, only 24% of respondents said they rarely speak to friends about mental health. This combination of data aligns with my hypothesis that Gen Z is speaking pretty openly about mental health. When asked if they believe that discussions about mental health on social media can lead to meaningful support and connections with others, the majority of respondents ranked it 3 or above on a scale of 1-5.
Hopefully as I receive more responses, my results will become more accurate and I will truly be able to answer my research question. So far the survey is working well, more updates to come in Week 8!
Week 8: Survey Responses are Rolling In!
This week I got up to 100 responses! By asking peers to repost my survey, getting BASIS teachers to encourage their students to take it, and posting the survey repeatedly on my own social media, I reached my goal! I plan on closing the survey at the end of this week so I can start finalizing my findings. The data has for the most part stayed the same, I noticed most respondents were born between 2005 and 2008, this is likely due to most of my responses coming through highschoolers. I noticed something interesting as I looked at individual responses this week. Some respondents born 2010 or later answered “no” when asked if they identified with Gen Z. I expected to see older respondents Identifying more with the “Millennial” label, but I did not expect members of GenZ to identify more with Gen Alpha. It makes sense that on the edges of a generation people may not know where they fit, the survey results of these few respondents were not significantly different, however, than those who did identify with Gen Z. I wonder how Gen Alpha will be affected by social media, seeing as they are growing up with it, unlike older members of Gen Z. I don’t expect to see much change in the results before I close it on Friday, but only time will tell. See you all in week 9!
Week 9: What about AI?
There was a lot of commotion online this week about AI’s role in the political discourse surrounding the conflict in the middle east. AI was used to depict Rafa, and many were upset about this. AI is becoming more ubiquitous, it’s everywhere, especially on social media. The constant stream of infographics one comes across on social media platforms like instagram is part of the reason I decided to look at social media for my senior project. In my proposal I discussed social media’s increasing role in political discourse, and how it is negatively shaping our conversations with each other. Through filter bubbles and social media algorithms, we as viewers, or “scrollers” usually only see one side of things. This issue is very clear at the moment in the political world, but it only makes me more worried and curious about the impact of social media on the conversations young people are having about their mental health. As I closed my survey this week the results largely confirmed prior research, and only made me more sure of the negative impact of social media on young people, and older generations. Next week I hope to get a fuller view of social media's impact on the way we discuss mental health. I am very much looking forward to presenting my finding to you all in person on the 14th. See you in Week 10!
Week 10: Blog Before Presentation
Welcome to my tenth blog post! As upsetting as some of the results of my survey have proven to be, it has been an absolute pleasure working on this project these past months and sharing all I have been doing and learning with you all. I spent most of this week finalizing my presentation and final deliverable, but I wanted to share something interesting I found as I finished analyzing my survey. As I will mention in my presentation, many respondents were actually not Gen Z, but I used their responses as a point of comparison. I assumed older generations would not be as impacted by confirmation bias on social media, this is why my original research question compared Gen Z and Gen X. After looking at these individual responses from older respondents, I found that they were spending almost the same amount of time online as Gen Z respondents. Even my mom, who only used facebook to check on politics up until 2018, now finds herself doom scrolling instagram after dinner. I cannot wait to share my work with you all on the 14th, thank you for following my progress, it’s been a blast!
Post 12: Final Blog
Hello all! This will be my last blog of the senior project. I have had such a great time working with my advisors and participants throughout this past trimester. With graduation coming up, it seems appropriate to reflect on my high school experience. After my presentation, Mr. Goldman said how amazing it was to have the students become the teachers, I loved being able to share some of what I learned with you all. Overall it was an amazing experience researching and presenting. Thank you for going on this journey with me!