"I thought the air quality in my house was better than it is outside, because we're next to the highway... But after doing this study and knowing that the air quality inside is worse than it is outside, it's kind of sickening honestly."
That's Jane, a Globeville resident, quoted in a published in Frontiers in Environmental Science (2024). She lives next to one of two interstate highways cutting through her neighborhood. She joined the study to understand how outdoor pollution affected her son's asthma. She finished it understanding something else: most of her exposure was happening at home.
Jane is one of 38 community scientists who carried an Atmotube PRO for two 30-day deployments in North Denver's Globeville, Elyria-Swansea, Cole, and Clayton (GESCC) neighborhoods — designated the most polluted residential ZIP code in the United States. The paper is led by Marisa Westbrook (Portland State University) with co-authors from CU Boulder and CU Denver, funded by the National Science Foundation as part of the Social Justice and Environmental Quality in Denver (SJEQ-D) project.
We've shared the SJEQ-D project before — most recently when it was featured on Rocky Mountain PBS's Colorado Voices. The newly published findings are the qualitative half of that work: 60 semi-structured interviews conducted immediately after each deployment, before participants received any researcher analysis. What people thought before, during, and after carrying the monitor — in their own words.
Participants gave three reasons for joining: curiosity about baseline exposure (45%), interest in conducting community science (37%), and existing health concerns (21%). The primary benefit didn't depend much on which reason brought them in.
Across 38 participants:
The most consistent shift, reported by participant after participant, was the reframing Jane described. People came in worried about the refinery, the highway, the asphalt plant. They left more focused on their own kitchens, bedrooms, and cleaning routines.
Participants weren't asked to run experiments. They were asked to carry the monitor and sync the data daily. Half of them ran experiments anyway.
Raquel suspected her partner's humidifier was making her ill. The monitor showed it was. (Existing research backs her up — ultrasonic humidifiers can emit PM2.5.) Susan had a long-running disagreement with her partner about whether their air purifier was worth running. The data settled it. Margaret moved her monitor around the kitchen until she found the spot where VOCs spiked highest — the pan rack — and changed how she stored cookware.
These weren't trained researchers. Many were recruited through door-to-door flyering. Some told interviewers compensation was their only motivation. The authors argue this matters: when a low-cost monitor puts real-time data into someone's hand, they tend to start asking their own questions — not because they were taught to, but because they could.
The paper doesn't only celebrate. It identifies four limits worth surfacing.
First, the Air Quality Score (AQS) in the Atmotube app is a composite — it combines particulate matter (well-established health effects) with TVOCs (an aggregate parameter, not toxicologically specific). Participants sometimes saw their AQS drop and didn't know which pollutant was responsible. The authors recommend that monitor designers help users distinguish between scientifically established pollutants and screening-level indicators. We take that point.
Second, personal monitoring naturally narrows attention to the individual sphere. Some participants came in wanting to document neighborhood-level industrial pollution and ended up focused mostly on their own homes. Useful for behavior change; less useful, on its own, for structural advocacy.
Third, a portion of participants (16%) expressed powerlessness about outdoor air quality they couldn't control. Awareness without agency can be its own kind of burden.
Fourth, novelty wore off. Repeat participants engaged less deeply the second time around. Personal monitoring works best as a learning tool, not as continuous background surveillance.
This is a study about Atmotube PRO, but the takeaway isn't really about a device. It's about what changes when data becomes local, fast, and yours. Composite indices, color-coded gauges, in-app explanations — these design choices shape what people notice, what they test, and what they change. The Westbrook team's work is a useful mirror for anyone building these tools.
Two things stand out for us. The behavior-change finding (40%) is consistent with what we hear from users in other contexts: once you see your own kitchen on the screen, the abstraction of "air quality" becomes concrete and actionable. And the limitations section is a roadmap for what we keep working on — particularly around helping users understand what's driving a score change, and connecting personal data back to community-level context.
If you're a researcher considering a similar community-monitoring study, the SJEQ-D paper is worth reading in full. Methodology, interview guide, and participant quotes are all detailed there.
Read the full study in Frontiers in Environmental Science →
