digitalwellbeing.orgHow to thrive in our hyper-connected world

The ‘Facebook Files’ on Instagram harms – all leaked slides on a single page

Here, on a single, convenient, scrollable page, are Facebook’s internal research presentation slides from the 2021 ‘Facebook Files‘ leak. The slides reveals Facebook’s own research on the potential effects of Instagram on teen wellbeing – ranging from suicide and self-injury (SSI), to sleep, anxiety, unhealthy social comparison. body image issues, and an inability to limit time spent on social media. Just click on any slide to see a downloadable high resolution version.

The original leaked files (mostly PDFs of screen grabs) can also be downloaded here.

As of October 2021, it’s unclear whether this much publicised leak will precipitate any change, as advocated by former Facebook employee and ‘whistleblower’ Frances Haugen. In addition to providing enhanced digital wellbeing tools and resources, changes could involve

  1. Better age-gating on Instagram and other youth-oriented platforms
  2. No algorithmically-promoted content or ads for young users
  3. Making Facebook legally responsible for algorithmically-promoted content
  4. Verified identities for users
  5. Opening up anonymised Facebook data for independent peer-reviewed effects research
  6. Putting more Facebook tax dollars to the service of improving digital literacy in schools.

In China, TikTok (Douyin) has limited access for young people to 40 minutes per day, whilst national mandates in China limit digital gaming time to three hours a week — from 8 PM to 9 PM on Friday, Saturday and Sunday. The leaked Facebook research reports propose their own alternative actions.

However, with attention-seeking media attention now turning to the potential harmful effects of algorithmically-promoted content on TikTok, the further pathologisation of online-gaming/e-gaming as an addictive “Gaming Disorder” in the latest international disease manual ICD 11 that comes into effect on Jan 1, 2022, and potential harmful media effects associated with Netflix’s Squid Game, the media panic over Facebook may be over – for now.

The message is the algorithm

Nevertheless, the research reports leaked in “The Facebook Files: A Wall Street Journal Investigation” are worth re-publishing here for convenient access – for at least two reasons. Firstly, the leak may mark a watershed moment in upgrading how we deal with the impact on algorithmically-driven social media on our lives. As such, for posterity, and for future reference, it is useful to have a convenient access to the files. Just click on any slide for a high resolution version. Secondly, widespread media mis-reporting of what is in the files makes it it important to show what the data actually shows. In sum, these leaked studies are opinion polls (and followup interviews) with Instagram users designed to capture subjective user opinions about how Instagram might influence mental health. The research does not provide evidence about the impact of Instagram on teen mental health.

Of course, this does not mean that algorithmically-promoted content on Instagram, or other digital media, has no harmful or beneficial effects. Decades of media effects research, including my own research, suggest that it may. However, peer reviewed research is tentative and suggests that any effects are likely to be small – and sometimes positive, whilst at other times negative. But these Facebook studies are not the ‘smoking gun’ as presented in media reports, and they do not show that “Facebook Inc. knows, in acute detail, that its platforms are riddled with flaws that cause harm“. Neither do they provide evidence to support any direction or presence in causality between observed correlation rising levels of social media use and rising numbers of teens reporting mental health issues.

Anyway, here are the slides from the two research presentations on the potential harms of Instagram on content on mental health. (Facebook itself, released annotated versions of these reports in PDF form, which can be downloaded here and here).

Instagram Leaked Study 1: Hard life moments – mental health deep dive

[2019 internal document reporting findings from a 6-country user survey (n=22,410*) across Brazil, India, Indonesia, Japan, Turkey, USA]

* not all respondents answered all questions, Mexico appears in footnote on slide 8, ‘SSI’ stands for suicide and self-injury.

Slide 1: Hard Life Moments – Mental Health Deep Dive
Slide 2: LT;DR: Life moments on Instagram – Mental Health
Slide 3: Objectives
Slide 4: What We Did
Slide 5: Hard Life Moments
Slide 6: Mental Health Findings
Slide 7: Work/School Stress Has The Highest Reach
Slide 8: Most Issues Had Higher Reach Among Teens
Slide 9: SSI [Suicide and Self-Injury] Was The Most Intense Experience
Slide 10: SSI [Suicide and Self-Injury] Was The Most Negative Experience Across Teens And Adults
Slide 11: People Expect Support From Instagram
Slide 12: Teeens And Adults Had Slightly Different Expectations
Slide 13: Instagram Is More Likely To Make Things Better Than Worse
Slide 14: But, We Make Body Image Issues Worse for ! in 3 Teen Girls
Slide 15: Most Users Used Instagram To Take Their Mind Off Experience
Slide 16: Most Wished Instagram Had Given Them Better Control Over What They Saw (Except SSI [Suicide and Self-Injury])
Slide 17: Prioritization
Slide 18: Social Comparison Has High Reach, Moderate Intensity And We Are Making It Worse For 1 In 4 People
Slide 19: Body Image (A Related Issue) Also Stands Out For Teens
Slide 20: Social Comparison Ranks Highest Among Issues Instagram Should Care About
Slide 21: For Teens, It [Social Comparison] Is Followed By Body Image Issues
Slide 22: Instagram
Slide 23: Teens Had Similar Expectations of Support As Adults For Instagram
Slide 24: Relative Importance Of Issues
Slide 25: Issues by Reach, Instensity, Responsibility
Slide 26: Relative Importance of Issues
Slide 27: Relative Importance of Issues

Instagram Leaked Study 2: Teen Mental Health Deep Dive

[2019 internal document reporting findings from a 2-country (UK and US) qualitative research study (n = 40 in-person interviews, with follow-up video calls (n = 8) with young Instagram users (aged 13-17), supplemented by online survey (n = 2,503)]
Cover Slide: Teen Mental Health Deep Dive
Slide 2: Objectives
Slide 3: Methodology
Slide 4: Seven Key Takeaways
Slide 5: Teens generally agree on what “mental health” is
Slide 7: Teens generally agree on the definition of mental health and what issues are related to it
Slide 7: There is a large consensus on what issues constitutes mental health among teens
Slide 8: Mental unwellness was depicted in sex themes
Slide 9: The phrases “down” and “a bit depressed” describe teens when they’re feeling bad, but descriptions are gendered
Slide 10: Teens say the phrases “down” and “a bit depressed” best describe them when they’re feeling bad
Slide 11: Language around mental health is gendered
Slide 12: Most teens report feeling a mental health issue
Slide 13: About half of teens report having positive well-being, but boys have higher levels of well-being than girls
Slide 14: 82% report feeling a wide range of emotional issues
Slide 15: The perfect image, feeling attractive, and having enough money are the most likely to have started on Instagram
Slide 16: Depression, low self-esteem, and loneliness are both experienced and identified as mental health problems
Slide 16: Friedns and family have the most positive impact on mental health
Slide 18: More teens who are satisfied with life say that various attributes have a positive effect than those not satisfied
Slide 19: The effects of Instagram depends’ on teen’ subjective well-being
Slide 20: Teens are quick to point out the positive role Instagram plays in their well-being
Slide 21: One in five teens say that Instagram makes them feel worse about themselves, with UK girls the most negative
Slide 22: Meme accounts, comments, and conversations on social media makes teens feel the best
Slide 23: While the overall effects of Instagram are positive, the effects are determined in the moment
Slide 24: Teens blame Instagram for increases in the rates of anxiety and depression among teens
Slide 25: Teens who struggle with mental health say Instagram makes it worse
Slide 26: Teens not satisfied with their lives are more likely to say Instagram makes then feel worse than those who are satisfied
Slide 27: The perceived effects of Instagram differed by market
Slide 28: Three categories of harm on Instagram
Slide 29: Harm on Instagram falls into three categories
Slide 30: Social comparison creates a negative feedback loop
Slide 31: Social comparison exacerbates problems teens are dealing with
Slide 3oys and girls report comparing themselves to others with similar frequency but about different topics
Slide 33: Social pressure leads to teens setting unrealistic standards for themselves
Slide 34: Social pressure, especially around body image, makes teens feel bad
Slide 35: Social pressure leads to negative downstream effects
Slide 36: Bullying and friendship drama leas lasting wounds
Slide 37: Bullying is intense but time-limited
Slide 38: How mental health plays out on Instagram
Slide 39: About 70% say they’ve seen Instagram posts or Stories related to mental health issues
Slide 40: Teens who are unsatisfied with their lives are more Lilly to see content related to mental health on Instagram
Slide 41: Content about SS makes teens feel the works, followed by people feeling they aren’t good enough
Slide 42: Feeling not good enough, questioning friendship, and feeling alone have high reach and intensity for teens
Slide 43: Teens feel they have to cope alone
Slide 44: Teens feel they have to cope by themselves, but they don’t want to be alone
Slide 45: Most teens think friends and parents should help them with the challenges they face
Slide 46: At the same time, parents can’t understand and don’t know how to help
Slide 47: Product suggestions – Personal and custom
Slide 48: Teens want help controlling the time they spend on the app
Slide 49: Teens in both markets want Instagram to enhance their control of both Feed and Explore
Slide 50: Teens in the UK, especially, wanted Instagram to punish bad actors
Slide 51: Outreach suggestions – Make it easier to talk
Slide 52: About 3 in 4 teens, and half of girls, would be likely to look at a page around feeling good about yourself
Slide 53: Teens want to hear from people like them to be their voice and make them feel less alone
Slide 54: On average, teens say that about 25% of the content should be practical tips for those in need
Slide 55: Based on qualitative feedback, we tested seven statements
Slide 56: Positive accounts, content to close friends, and make me laugh were the most resonate territories
Slide 57: Positive accounts, content to close friends, and make me laugh were the most resonate territories
Slide 58: Make me laugh is the most preferred in both markets
Slide 59: Make me laugh and communicate with close friends have the best fit with Instagram
Slide 60: Make me laugh and communicate with close friends are the most realistic
Slide 61: Positive accounts (in the US), communicate with close friends, and make me laugh are the most believableS
Slide 62: Positive is the most appealing in the US, and cares about mental health in the UK
Slide 63: Teens respond best to positive language and uplifting actions
Slide 64: Explore best matches positive accounts, laugh, and escape; DMs with communication
Slide 65: Many thanks to the team that made this possible
Slide 66: Have ideas or thoughts for next steps? Click here.
Written by
Dr Paul Marsden
Join the discussion

2 comments
  • Indirect reports indicate that Facebook is fully aware that its product is commercially successful, but often has bad consequences. Meanwhile, internal documents show that Facebook has made little effort to address these issues or make them public.

  • When you are looking to find the best Asian wife, your first action might be to travel to a country known for its beautiful female population. For example, in Vietnam, and start looking for a woman there. However, this is not the best way to look for an Asian bride.Women can be very low-key. Although they will be very polite when they meet, they are unlikely to agree to a date or marry you after you have met a few times. Therefore, the most effective way to find an Asian wife bride is through dating sites.

digitalwellbeing.org

digitalwellbeing.org

Digital wellbeing covers the latest scientific research on the impact of digital technology on human wellbeing. Curated by psychologist Dr. Paul Marsden (@marsattacks). Sponsored by WPP agency SYZYGY.