I posted the idea of installing very bright lights on LW five years ago and Eliezer commented there so I give myself credit for at least making that spontaneous idea more likely. And it happens to be the case I’ve been thinking about the failings of light boxes for SAD in the meantime.
What happened is that a few people experimented with light therapy, got succcess with 2500lux for two hours, decided two hours per day was infeasible outside the lab, found that they could get the same result dividing the time but multiplying the light intensity and then… just… stopped. They did studies with 10000 lux boxes and that’s a relatively expensive study so you better cooperate with a producer of such boxes. So you get some type of kickback and suddenly nobody’s interested in studying whether stronger, cheaper lights are even better. Light boxes became a medical device and magically became just as expensive as medical insurers would tolerate. That they don’t work for everyone was expected, because no depression treatment works for everyone (maybe except electroconvulsive therapy, and Ketamine but that was later). And LEDs only became cheap enough recently (five years ago they still weren’t clearly the cheapest option), so going much beyond 10000 lux presented enough of a technical challenge to make further trials pretty expensive, until recently.
Right now, you could probably do a study with SAD sufferers who have tried light boxes and found them insufficient. Give them a setup that produces like 40000 lux and fits in a normal ceiling fixture so they can have it running while they do things, rather than have to make time to sit in front of it. For a double blind control design, maybe give one group twice the brightness of the other? Have your participants log every day how much time they spent in the room with the lamp running, and how much time they spent outside. Don’t give them money but let them keep the lamp if they continue mailing their fillled out questionnaires. Should be doable at a hundred dollars per participant, and without ever physically meeting them. You still need six figures to run that study at a large enough size, and no light box maker is going to fund you.
I posted the idea of installing very bright lights on LW five years ago and Eliezer commented there so I give myself credit for at least making that spontaneous idea more likely. And it happens to be the case I’ve been thinking about the failings of light boxes for SAD in the meantime.
What happened is that a few people experimented with light therapy, got succcess with 2500lux for two hours, decided two hours per day was infeasible outside the lab, found that they could get the same result dividing the time but multiplying the light intensity and then… just… stopped. They did studies with 10000 lux boxes and that’s a relatively expensive study so you better cooperate with a producer of such boxes. So you get some type of kickback and suddenly nobody’s interested in studying whether stronger, cheaper lights are even better. Light boxes became a medical device and magically became just as expensive as medical insurers would tolerate. That they don’t work for everyone was expected, because no depression treatment works for everyone (maybe except electroconvulsive therapy, and Ketamine but that was later). And LEDs only became cheap enough recently (five years ago they still weren’t clearly the cheapest option), so going much beyond 10000 lux presented enough of a technical challenge to make further trials pretty expensive, until recently.
Right now, you could probably do a study with SAD sufferers who have tried light boxes and found them insufficient. Give them a setup that produces like 40000 lux and fits in a normal ceiling fixture so they can have it running while they do things, rather than have to make time to sit in front of it. For a double blind control design, maybe give one group twice the brightness of the other? Have your participants log every day how much time they spent in the room with the lamp running, and how much time they spent outside. Don’t give them money but let them keep the lamp if they continue mailing their fillled out questionnaires. Should be doable at a hundred dollars per participant, and without ever physically meeting them. You still need six figures to run that study at a large enough size, and no light box maker is going to fund you.
Why isn’t a light box maker willing to pay $100,000 as a marketing expense?