Log in
Subscribe
Aug 18, 2026
In the Dark About Daylighting
In the Dark About Daylighting
00:00
11:55
1×
Transcript
0:00
Hello, welcome back to Read the Room. I'm Katie, and it's been a long summer. Lots of traveling, lots of run-ins with place attachment, which was actually last month's newsletter topic.
0:12
I have been taking my preschooler to a bunch of new places this summer, um, even locally, and kept wondering how place attachment might work in little kids, and it turns out there's a brand-new article in the Journal of Environmental Psychology on exactly that, and it is so new that the full text isn't even up yet.
0:32
I'm keeping an eye on it. I'll share it with you when the time comes. The timing kinda feels like kismet. Today, we're shining a light, literally, on a landmark daylighting study.
0:42
I've seen this thing cited so many times that I have lost count. I've sat in project interviews where, like, well-intended designers use it to make a point.
0:53
It's been referenced in over 200 other scientific articles at least, and it is bullshit.
1:02
But before we get into it, enrollment just opened for the first two cohorts of my new course, Evidence-Based Design: Practical Foundations.
1:11
It's 12 weeks long with six modules, two rounds of one-on-one coaching with me, and you walk away with a portfolio-ready evidence-based design project. The first lesson is free. You can check it out.
1:24
The link is in the show notes, and you can get $100 off of registration with code READTHEROOM, all one word. Okay, onto the study. More light is better, right?
1:39
Well, that's at least what a utility company wanted you to think in 1999. Back in the late '90s, Pacific Gas & Electric and the California Energy Commission joined forces to fund a study.
1:53
They brought in the Heschong-Mahone Group, a well-known building science firm at the time, to investigate daylight on behalf of the utility. This was a big one, a big study.
2:05
The HMG team pulled test scores and demographic data on more than 21,000 students across three districts in California, Washington, and Colorado.
2:16
They went into some schools in each district and rated the classrooms on a scale of zero to five for daylight based on window size, skylights, and glazing using architectural drawings when they couldn't go into the schools.
2:32
After controlling for around 50 variables, which sounds impressive, right?
2:36
They found students in the most daylit classrooms progressed 20% faster in math and 26% faster in reading over a year compared to kids in the least daylit rooms. 20% in math, 26% in reading.
2:54
I have seen those numbers in two separate LEED and WELL project verification narratives. I've seen them in interviews.
3:03
If you've worked in this field for more than five minutes, [laughs] you have probably seen those numbers cited, too. There's a sketch in the original report that shows two classroom layouts side by side.
3:13
There's one packed with windows and skylights. There's one with almost none, so you can kind of see how they compared these conditions. I'll also drop that in the show notes if you want to see it.
3:23
It's also in the newsletter. So now, the obvious problem with this study. You've heard this one from me before, actually. Built environment research has a random assignment problem.
3:35
It's really, really hard to randomly sort people into different spaces, especially in places like schools and homes where quality varies a lot to begin with. I can't tell you to live in a lower quality home.
3:50
We actually encountered this back in issue one with Roger Ulrich's study. No one randomly assigned patients to a view of nature or a view of a brick wall. They retrospectively went back and looked at data.
4:03
That's what happened in this study, too. In this case, nobody randomly assigned kids to daylit or non-daylit classrooms, and peer reviewers before this article even came out caught an obvious issue really fast.
4:16
What if the more experienced or better-liked or just plain better teachers ended up in the nicer rooms with more windows? That caveat is actually addressed in the original article, but there's another wrinkle.
4:31
HMG only visited a sample of the actual classrooms to take real measurements. For the rest, they estimated daylight scores from architectural drawings, from photos, and maintenance records.
4:43
In fact, in their own report, they write that those estimates were not verified with on-site measurements since we could not visit enough classrooms under similar daylighting conditions.
4:57
To their credit, they acknowledge it and the HMG team went back and checked their own work. The original study came out in 1999, and they ran a reanalysis in 2001.
5:07
This time, they surveyed teachers directly, um, and they controlled for electric lighting as well. They didn't find evidence of assignment bias, at least not on the variables they measure.
5:20
And honestly, this is great, right? This is a nice example of peer review doing its job to improve the rigor of science. But it is still an observational study.
5:32
The authors went looking for problems after the results came in instead of designing the study to rule those problems out. This is not exactly the gold standard of how we do research.
5:43
When you're evaluating how strong a study is, you always have to check the funding source. I know I sound like a conspiracy theorist.
5:51
In this case, a utility company paid for research that happened to just build the case for daylighting incentives and the energy code changes they wanted anyways. That doesn't automatically make the findings wrong.
6:05
It's just important context There is a follow-up too, and it's actually a pretty good one. HMG ran a better-controlled study, in my opinion, in 2003, this time with office workers.
6:19
Workers with a view of the outdoors scored 10% better on a cognitive acuity test and 16% better on a memory test compared to workers with no view at all.
6:30
Even though the research design was better, the results aren't super compelling. Are you a better employee actually if your memory is 16% better? Questionable, but I'll take it for now.
6:43
But here's the part that really gets me when I was looking into this work.
6:47
When they isolated the effect of daylight specifically, not just the view, just daylight, a 10% increase in measured daylight predicted only a.4% improvement in cognitive accuracy, and it had no measurable effect on memory at all.
7:05
So we're not even in the realm of 20%, not 26%, basically nothing. This is actually a weaker relationship than the famous spurious correlation between the divorce rate in Maine and the per capita margarine consumption.
7:19
I am not kidding. That is a real example that statisticians use, and they are correlated nearly one to one. Regardless, an authority actually weighed in a few years later.
7:30
A research review out of Lawrence Berkeley National Lab's Indoor Air Quality Program looked at the daylighting and performance literature as a whole.
7:40
Systematic reviews sit at the top of the evidence hierarchy, so this carries more weight than any single study.
7:47
When we look at the results from multiple studies together, we get a better conclusion, and they concluded that they had too few studies with results that were too inconclusive, so they couldn't draw a conclusion either way.
8:01
Are we shutting the windows on the whole idea of daylighting? No. Anecdotally, we all know that windows matter. Maybe it's the view of nature, maybe it's positive distraction, maybe it's the light itself.
8:13
I am not going to argue against the value of windows. I'm rather here to argue about the evidence for windows, and there's much better evidence out there now, so please stop citing the HMG study. It's old.
8:28
Let's get into the better evidence. Four studies, real fast. First, treating the office as a lab. Mayo Clinic's Well Living Lab tested this for real.
8:39
10 people worked their normal jobs, but in a simulated office for 14 weeks, and they cycled through blackout shades, mesh shades, and a dynamic tint in randomized order.
8:51
Daylight and view access improved cognitive performance and satisfaction and reduced eye strain compared to the blackout condition. Second, an interesting new way to measure.
9:05
A very recent University of Manchester study tracked 60 adults' actual daylight exposure using wearables instead of estimating it from window size. More daytime light exposure, better cognitive performance.
9:21
It's a small sample, just like the Mayo Clinic study, but it is a genuinely more objective way to collect data. Third, a better school study.
9:30
A 2020 study used a massive EU data set, 2,670 kids across 155 classrooms in 12 European countries. They found daylighting linked to math and logic scores, with window-to-floor ratio mattering the most.
9:48
It was bigger and way more geographically diverse than the original US study. It's still correlational, but it's more diverse.
9:57
Fourth, for the research nerds, the original HMG study is worth reading too, even if it's just to practice spotting the confounds and alternative explanations yourself.
10:07
See what issues you can find in their research design, and notice how they tried to fix some of them the second time around.
10:15
All of these studies are linked in the show notes if you don't want to go down the rabbit hole yourself. So in closing, daylight is genuinely a good thing for a lot of reasons.
10:26
Circadian rhythm, views of nature, better energy use most of the time. We should all be pushing our clients towards better daylight interventions. But none of that depends on a poorly designed study from 25 years ago.
10:43
So if you are in the position to recommend a glazing strategy or a window-to-floor ratio and you're backing it up with the HMG study, you're standing on shaky ground.
10:56
Instead of citing a correlational study from 1999 funded by a utility, grab one of the four studies that I just walked through. Good design doesn't need inflated data to be justified.
11:11
Good data is out there for good design. Go use it. If you want to go deeper on spotting strong evidence and using this type of science in your own design practice, check out the free preview lesson for my new course.
11:27
It's linked below. Use code READTHEROOM, one word, for $100 off enrollment. That's it for this one. If this was useful, send it to one person in your firm who needs to hear it.
11:39
And if you want to work with me on a project or a speaking engagement, my info is in the show notes, or head over to katiepedito.com. Talk soon. [chair creaks]
Read the Room
Listen on
Apple Podcasts
Apple Podcasts
Pocket Casts
Pocket Casts
Castro
Castro
Recent episodes
I Just Need Some Space
Jul 15, 2026
Your House was Designed for an English Detective
Jun 2, 2026
