📄 social_science

The Uneven Benefits of Hiding Social Cues: Demetrication and Well-Being on Instagram

This preregistered field experiment reveals that while hiding Instagram engagement metrics reduces passive content consumption and increases perceived social support, its impact on overall well-being is moderated by neuroticism, benefiting highly neurotic users while offering little to no advantage—or even a slight cost—to less neurotic users.

Original authors: Kokil Jaidka, Harshit Aneja

Published 2026-07-03

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Kokil Jaidka, Harshit Aneja

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Idea: Hiding the Scoreboard

Imagine you are at a party where everyone is wearing a digital scoreboard on their forehead. These scoreboards show exactly how many people are talking to them, how many friends they have, and how many people are looking at them.

For some people at the party, this scoreboard is fun. For others, it's a nightmare. They constantly compare their low score to the high scores of others, feeling inadequate and anxious.

This study asked a simple question: What happens if we turn off all those scoreboards? Does the party become happier for everyone, or does it change things differently for different people?

The researchers tested this on Instagram by creating a special version of the app (called "Smolgram") that hid all the "likes," follower counts, and view numbers for some users, while others saw the normal app.

What They Found: The "One Size Does Not Fit All" Result

The study had three main discoveries:

1. People scrolled less, period.
When the scoreboards were hidden, people didn't just scroll slower; they scrolled less overall.

The Analogy: Think of scrolling through Instagram like eating popcorn at a movie. Usually, the "likes" and "views" are like a movie theater that keeps telling you, "This scene is popular! Watch it!" When you hide those numbers, it's like the theater goes silent. You still eat the popcorn, but you lose the urge to keep eating just because the theater is cheering you on. People viewed fewer photos and read fewer comments because the "social proof" (the numbers telling them what's popular) was gone.

2. The average person didn't feel "happier," but they felt more supported.
Surprisingly, hiding the numbers didn't make the average user feel significantly more joyful or "flourishing." However, it did make them feel like their friends cared about them more.

The Analogy: Imagine you are in a room where everyone is shouting their achievements. If you stop shouting, you might not suddenly feel like a superhero, but you might feel less like you're being judged. The study found that without the numbers, people felt their social connections were stronger, perhaps because they stopped comparing their "score" to everyone else's.

3. The "Neuroticism" Split: The Real Story
This is the most important part. The effect of hiding the numbers depended entirely on the person's personality, specifically a trait the researchers called "Instagram-specific neuroticism." This isn't just general anxiety; it's the specific worry someone feels only when they are on Instagram.

The "High-Neurotic" Group (The Anxious Scrollers):
- Who they are: People who get nervous, self-conscious, or anxious specifically when they are on Instagram. They are the ones constantly checking their scoreboards.
- The Result: When the scoreboards were hidden, they felt better.
- The Analogy: For these people, the scoreboard was a heavy backpack filled with rocks. Taking the backpack off (hiding the numbers) made them feel lighter and happier. They stopped comparing themselves to others and felt relief.
The "Low-Neurotic" Group (The Chill Scrollers):
- Who they are: People who use Instagram casually and don't get anxious about their numbers.
- The Result: When the scoreboards were hidden, they felt slightly worse.
- The Analogy: For these people, the scoreboard wasn't a burden; it was a friendly high-five. It told them, "Hey, people are looking at your photos!" When the high-fives disappeared, they felt a little disconnected or confused, as if the app had lost some of its fun.

Why This Matters

The study proves that you can't just say "hiding likes is good" or "hiding likes is bad." It's like saying "turning off the music at a party is good."

For the person trying to relax, turning off the music is great.
For the person who loves to dance, turning off the music ruins the vibe.

The researchers found that the "anxious" users were the ones who needed the scoreboard removed to feel better, while the "chill" users actually enjoyed seeing the numbers.

What the Study Did NOT Say

It did not say that hiding likes will cure depression or anxiety disorders in the real world.
It did not say that Instagram should permanently delete all numbers for everyone.
It did not say that this works for all social media apps (like TikTok or Facebook), only that it worked for this specific group of Instagram users.

The Bottom Line

Hiding social media numbers changes how we use the app (we scroll less) and how we feel. But it doesn't help everyone equally. It acts like a relief valve for people who are stressed by social comparison, but it can feel like a loss of connection for people who aren't stressed by it. The best solution, according to the authors, is to let users choose for themselves whether they want to see the scoreboard or not.

Technical Summary: The Uneven Benefits of Hiding Social Cues

Problem Statement
Social media platforms, particularly image-sharing services like Instagram, are frequently linked to declines in psychological well-being, especially among young adults. Existing research often attributes these negative outcomes to individual behaviors (e.g., passive scrolling) or trait-level vulnerabilities (e.g., upward social comparison). However, this "bottom-up" focus may overlook structural features of platform design, specifically the visibility of engagement metrics (like counts, follower tallies, view numbers). These metrics render others' popularity legible and one's own relative standing instantly visible, potentially affording chronic upward social comparisons. Conversely, metrics may also function as social rewards that sustain self-disclosure loops. While policy interest in "demetrication" (hiding metrics) has grown, causal evidence regarding its effects on well-being remains thin. Prior observational studies cannot isolate metric exposure from other features of heavy use, and existing experiments have been limited to single posts or short durations. Crucially, there is a lack of understanding regarding who benefits from demetrication and who does not.

Methodology
The authors conducted a preregistered field experiment (January–February 2023) involving 140 enrolled participants (analytic $n=123$ ) recruited from Singapore.

Intervention: Participants used a custom Android client, "Smolgram," which routed data through the official Instagram API. This app programmatically manipulated the feed interface.
Conditions: Participants were randomly assigned to either:
1. Demetricated Condition: All engagement metrics (like counts, follower/following counts, view numbers) were hidden from every screen.
2. Control Condition: Metrics were displayed exactly as in the native Instagram interface.
Design: The design preserved follow networks, content selection, and the ordinary browsing experience, isolating metric visibility as the sole variable. The study lasted three weeks.
Data Collection:
- Surveys: Wave 1 (baseline) and Wave 3 (post-study) measured subjective well-being (Positive Affect, Affect Balance, Flourishing), perceived social support, self-disclosure, and self-discrepancy.
- Personality Measures: Both general trait neuroticism and "IG-context neuroticism" (a platform-specific measure of anxiety and self-consciousness on Instagram) were assessed.
- Behavioral Logs: The app logged objective event counts for seven behavioral types (e.g., photo views, comments viewed, profile visits) throughout the study.
Analysis: Primary analyses used Analysis of Covariance (ANCOVA) controlling for baseline scores and demographics. Moderation effects were tested using interaction terms with neuroticism. Robustness was assessed via a Specification Curve Analysis (SCA) across 120 analytical specifications.

Key Results

Behavioral Impact: Demetrication substantially reduced passive content consumption. Participants in the metric-free condition viewed significantly fewer photos ( $p=.032$ ) and marginally fewer comments ( $p=.073$ ) than controls. All seven logged behavioral categories trended in the same direction (reduced engagement), suggesting a broad reduction in passive scrolling rather than a redirection of engagement.
Main Effects on Well-Being: There were no significant main effects of demetrication on positive affect, affect balance, or flourishing. However, participants in the demetricated condition reported significantly higher perceived social support ( $p=.046$ ) compared to controls.
Moderation by Neuroticism: The effects on well-being were heterogeneous and moderated by IG-context neuroticism (but not general trait neuroticism). A crossover interaction pattern emerged:
- High IG-Neuroticism Users: Reported higher well-being (positive affect, affect balance, flourishing) under demetrication.
- Low IG-Neuroticism Users: Showed a small cost (lower well-being) under demetrication.
- This moderation was specific to the platform-context measure; substituting general neuroticism yielded null interactions.
Robustness: The Specification Curve Analysis confirmed that the positive interaction between demetrication and IG-context neuroticism was robust across 120 specifications (positive in 95%, significant in 27% of cases), whereas no specification using general neuroticism produced a significant interaction.

Key Contributions

Causal Isolation: The study provides causal evidence that metric visibility directly shapes both in-app behavior and psychological outcomes, isolating this variable from network structure and content algorithms.
Heterogeneity of Effects: It demonstrates that design interventions like demetrication are not uniformly beneficial. Instead, they restructure the psychological experience of the platform, benefiting users susceptible to platform-specific social comparison while potentially imposing a cost on those who rely on metrics as low-cost relevance signals.
Construct Specificity: The findings highlight that platform-specific self-evaluative processes (captured by IG-context neuroticism) are more predictive of responses to design changes than broad dispositional traits.
Behavioral-Psychological Dissociation: The study shows that behavioral suppression (reduced browsing) occurred across all users, but the psychological impact of that suppression varied by user type, indicating that behavioral and psychological interventions should be evaluated separately.

Significance and Claims
The paper claims that disabling metric visibility is particularly effective for users most susceptible to social comparison on Instagram, but it is not a universally positive design change. The authors argue that averaging treatment effects across a population can mask substantial heterogeneity, leading to misleading conclusions about whether an intervention is beneficial or harmful.

The findings suggest that platform-level design changes should be evaluated with an understanding of individual differences. Specifically, the authors propose that metric-hiding features should not be implemented as binary toggles but as user-configurable controls. This would allow users most susceptible to metric-driven social comparison to opt into relief, while preserving the informational value of metrics for users who experience them as benign. The study concludes that these results should be viewed as complementary to major life determinants of well-being rather than evidence of large-scale psychological impact, yet they offer theoretically meaningful insights for responsible platform design.

The Big Idea: Hiding the Scoreboard

What They Found: The "One Size Does Not Fit All" Result

Why This Matters

What the Study Did NOT Say

The Bottom Line

More like this