Most people assume that longer study sessions produce better results. Sit down for an hour, cover more material, retain more. It is a reasonable assumption. It is also wrong.
A consistent finding across cognitive psychology and language acquisition research is that short, frequent sessions outperform long, infrequent ones. Not by a small margin. By a significant one.
The study that settled it
In 1978, Alan Baddeley and David Longman ran a study with Post Office workers learning to type. They divided participants into groups with different session structures: one hour a day, two hours a day, four hours a day, and a 30-minute twice-daily group.
The group doing 15 minutes twice a day, a total of 30 minutes, learned faster and retained more than the groups doing one or two hours in a single block. The four-hour group performed worst of all. More time, worse outcome.
The results suggest that performance was a decreasing function of session length, indicating that distributed practice is superior to massed practice.
Baddeley, A.D. & Longman, D.J.A. (1978). The influence of length and frequency of training session on the rate of learning to type. Ergonomics, 21(8), 627–635.
The mechanism is attention. Human concentration is not linear. It peaks early in a session, around 10 to 20 minutes in, and then begins to drop. By the 45-minute mark of a demanding cognitive task, performance is measurably lower than it was at the 10-minute mark. Working memory is finite. Once it fills, new information competes with what is already there and much of it is lost.
Spaced repetition and the long view
Baddeley and Longman showed short sessions are better per unit of time. Harry Bahrick and colleagues showed that spacing those sessions over months produces dramatically better long-term retention.
Bahrick's 1993 longitudinal study tracked retention of Spanish vocabulary over 56 years. Participants who had originally studied in many small, spaced sessions retained far more vocabulary decades later than those who had crammed. The gap was not marginal. Vocabulary learned in massed sessions decayed quickly. Vocabulary learned with spacing held for a lifetime.
Nicholas Cepeda and colleagues (2008) reviewed 254 studies on spacing effects and found a consistent relationship: the optimal gap between study sessions grows as the retention interval grows. For long-term retention of a language over years, daily short sessions are close to optimal. Weekly long sessions are not.
The implication is stark. The hour-a-week approach, which describes most adult language classes, is one of the worst possible ways to structure learning for retention. You acquire material in the session. You forget most of it before the next session. You re-learn it. Repeat indefinitely without meaningful progress.