The following is taken from Mark Fettes' article on Esperanto education. The entire article can be found at http://infoweb.magi.com/~mfettes/f-r2.html
The Teachers College Study (1924-1935)
The Teachers College Study actually consists of three related experimental studies, taking place in 1924, 1928 to 1931 and 1934-1935, respectively (see Charters, n.d.; Division of Psychology, 1933). Directed and supervised by the well-known educational psychologist Edward L. Thorndike, and involving the assistance of Drs. Laura Kennon and Helen Eaton, the Teachers College Study focused on such issues as the learnability of auxiliary languages in general (and Esperanto in particular), as well as on the propaedeutic effects of Esperanto study on future language learning. Involving subjects ranging in age from 8 to 65, in a variety of different types of classroom settings, the Teachers College Study results indicated that, given the same amount of exposure, students will learn more Esperanto than they would of a national/ethnic language. In fact, the 1933 report included the claim that:
An average college senior or graduate in twenty hours of study will be able to understand printed and spoken Esperanto better than he understands French or German or Italian or Spanish after a hundred hours of study.
(Division of Psychology, 1933: 6-7)
This assertion appears to be the basis for the common assertion that Esperanto is roughly five times easier to learn than a national/ethnic language. The report, in fact, makes an even stronger claim:
On the whole, with expenditures of from ten to a hundred hours, the achievement in the synthetic language will probably be from five to fifteen times that in a natural language, according to the difficulty of the latter. (Division of Psychology, 1933: 7)
The empirical foundation for these claims, however, is weak at best, being based on very limited test results that do not adequately take into account compounding variables. However, the general claim that Esperanto is easier to learn than a national/ethnic language -- a claim that can offer in its defense both common sense and impressive anecdotal evidence -- does receive reasonably compelling support from the Teachers College Study. As the 1933 report notes,
Two facts are certain. Any one of the reputable "synthetic" languages is very much easier to learn than any natural language; but its learning is none the less a very substantial task. (Division of Psychology, 1933: 6)
As for the propaedeutic effects of Esperanto study, the Teachers College did suggest that there were positive outcomes associated with the study of Esperanto for both English and French learning. The report summarizes the results as follows:
So at the end of two years the English scores are more or less the same, though slightly higher for the Esperanto group; the French scores show 9.7 points more progress for the Esperanto group, which had studied French during one of the two years, than was made by the non-Esperanto group, which had studied French during the two years ...
(Division of Psychology, 1933: 30)
While these results do appear to be impressive, it should be kept in mind that they were based on a very small sample, and that the study did not involve carefully maintained experimental and control groups. In fact, the authors of the report themselves were clearly very hesitant to draw any substantial conclusions from their data. At best, one can say that, as far as the propaedeutic effects of Esperanto study, the Teachers College Study is marginally supportive of the positive claims that are often made.