Spaced Repetition: Spitzer's 3,600 Iowa Students to Alzheimer's Recall

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!

Forgetting, the episode argues, is the brain's energy-saving cleanup system, and spaced repetition is the way around it. The story starts in the 1880s with the German scientist Hermann Ebbinghaus, whose forgetting curve showed losses of up to 70 percent within days and who proposed that reviews at intervals could flatten it. In 1932 C. A. Mace's book The Psychology of Study called for reviews at gradually increasing intervals of one, two, four and eight days. In 1939 H. F. Spitzer tested more than 3,600 sixth graders in Iowa on science facts and found expanding test schedules drastically improved retention. Schools largely ignored the result for decades, in part because tracking individual schedules by hand was a logistical nightmare. The mechanism is desirable difficulty: a harder retrieval strengthens the memory, provided the first review comes soon enough.

Later research pushed beyond facts. A 2007 study by Hal Pashler and colleagues found spacing helped students learn the equation of a line, and that delayed feedback on wrong answers improved later scores. A 2023 pilot study spread neurosurgery residents' simulation practice over six weeks, and they outperformed peers trained in massed blocks. From the late 1980s researchers including Cameron Camp applied spaced retrieval to Alzheimer's patients with strong results, and in 2012 Jeff Small explained why it works for dementia. Tools followed: the Leitner box, Paul Pimsleur's graduated interval recall, and software algorithms that predict when each memory will fade.

• Ebbinghaus memorized thousands of nonsense syllables to the tick of a metronome to measure forgetting without prior knowledge.
• In a 2003 study, a caregiver used phone-based spaced retrieval to help an Alzheimer's patient recall her grandchild's name five days later.
• Sebastian Leitner's 1970s box had sections 1, 2, 5, 8 and 14 centimeters deep, and cards were reviewed only when a section filled.
• Pimsleur's 1967 audio system repeated words after five seconds, 25 seconds, two minutes and ten minutes, stretching to two years.
• Meta-analyses still debate expanding versus uniform intervals, often finding similar long-term results.