“Corsi” most commonly refers to the Corsi-Rosenthal box, a DIY air purifier made from box fans and furnace filters that gained widespread attention during the COVID-19 pandemic. It can also refer to the Corsi block-tapping task, a clinical test used to measure spatial memory. This article covers both, starting with the air purifier most people are searching for.
The Corsi-Rosenthal Box: A DIY Air Purifier
The Corsi-Rosenthal box (often called a CR box) is a homemade air cleaner built from four or five MERV 13 furnace filters arranged in a cube shape, with a standard 20-inch box fan mounted on top pulling air through the filters. It’s named after Richard Corsi, an indoor air quality engineer now at UC Davis, and Jim Rosenthal, a filtration expert and CEO of a filter company. The two helped popularize the design in 2020 as a low-cost way to remove airborne particles, including virus-carrying aerosols, from indoor spaces.
The concept is simple: air gets pulled through the filter walls of the cube, trapping particles, and clean air exits through the fan on top. Because the cube shape exposes a large surface area of filter material, airflow stays high even though the filters themselves are denser than what a box fan normally pushes air through.
What You Need to Build One
According to UC Davis, the standard build requires four two-inch MERV 13 filters, a 20-inch box fan, a piece of cardboard (the fan’s own packaging works) to seal the bottom of the cube, duct tape, and scissors. The cardboard forms the base, the four filters stand upright to create the cube walls, and the fan sits on top blowing upward. The whole thing can be assembled in about 20 minutes.
Total cost runs around $35 for a basic setup using two-inch filters. A version with thicker five-inch filters costs closer to $120 but delivers higher airflow. Either way, it’s a fraction of what commercial air purifiers cost.
How It Compares to Commercial HEPA Purifiers
The CR box performs surprisingly well against devices costing five to ten times more. Clean air delivery rate (CADR), the standard measure of how much filtered air a purifier produces per minute, tells the story clearly. A basic four-filter CR box running on its lowest fan speed delivers an estimated 471 cubic feet per minute (cfm). On high speed, that jumps to 659 cfm. Lab measurements from UC Davis found even higher numbers: 600 cfm on low and 850 cfm on high.
For comparison, a Coway Airmega 300, a well-regarded HEPA purifier retailing for around $549, produces about 265 cfm at maximum speed. A $233 Airthereal model tops out at 354 cfm. So a $35 CR box on its lowest setting already outperforms commercial units running flat out.
There’s a catch, though. MERV 13 filters are less efficient than true HEPA filters at catching the smallest particles. HEPA purifiers trap 95% or more of particles down to 0.3 microns (the hardest size to capture). MERV 13 filters catch about 50% or more of particles in the 0.3 to 1.0 micron range, per EPA ratings, and around 67 to 73% at 0.3 microns in the configurations tested. The CR box compensates by moving so much more air that its total clean air output still matches or exceeds HEPA units. You’re filtering a larger volume of air slightly less efficiently, which in practice cleans a room just as well or better.
Safety and Noise
A common concern is whether strapping filters to a box fan could cause the motor to overheat. The EPA partnered with Underwriters Laboratories (UL) to test this directly. UL ran five different box fan models with clean, smoke-loaded, and dust-loaded filters attached, including an extreme scenario where both sides of the fan were blocked for seven straight hours. None of the setups caught fire. External fan temperatures stayed near room temperature under normal use, and even in the worst-case blocked scenario, all measured temperatures remained below recognized safety thresholds.
One important caveat: these tests used fans manufactured after 2012, which include fused plugs and thermal cutoffs as standard safety features. Older fans were not tested and pose known fire risks. If you’re building a CR box, use a newer fan.
Noise is a more realistic downside. Box fans are louder than commercial purifiers designed with noise dampening, especially on higher speeds. Running the CR box on its lowest setting helps, and since even that setting delivers strong CADR, there’s rarely a need to crank it up.
When to Replace the Filters
Filters in a CR box last between 1,000 and 4,000 hours of actual run time on a medium setting. That range depends heavily on the environment. In a clean, uncarpeted office running eight hours a day, five days a week, filters can last about 25 months. In a dusty room with carpet and pets, or a space with lots of foot traffic, plan on closer to eight months at six hours of daily use.
A common mistake is swapping filters too early because they look dirty. Discoloration is a sign the filters are working, not that they’re spent. Unlike HVAC filters that restrict airflow when clogged (raising energy costs), CR box filters can accumulate visible grime and still function within their rated lifespan.
The Corsi Block-Tapping Task
If you searched “what is Corsi” in a psychology or neuroscience context, you’re likely looking for the Corsi block-tapping task, a widely used clinical test for spatial working memory. It was developed in the early 1970s and consists of nine blocks (or squares on a screen, in digital versions) arranged on a board. An examiner taps a sequence of blocks, and the person being tested reproduces that sequence, either in the same order (forward) or reversed (backward).
The sequences start short, typically two blocks, and grow longer until the person can no longer keep up. A healthy adult typically reaches a span of about seven blocks. Research on children shows that spatial span capacity grows through childhood and plateaus around age 14, when eighth graders average about 6.9 blocks, statistically indistinguishable from young adults averaging 7.1.
Clinically, the forward version is used to assess short-term spatial memory, while the backward version targets working memory more broadly. The test appears in diagnostic batteries for Alzheimer’s disease, Parkinson’s disease, schizophrenia, ADHD, and traumatic brain injury, all conditions that can impair the ability to track and recall spatial information. Digital tablet versions now allow more precise timing measurements, which can help detect subtle neurological changes earlier than the traditional wooden-block setup.