In An Experiment with a Project Curriculum (1923) Ellsworth Collings reports on an experiment he directed as county superintendent from 1917 to 1921 at a rural school in McDonald County, Missouri, where the children themselves - not the teacher or the curriculum - determined the projects and topics they would study. Collings claims that the findings of his dissertation strengthened the case for the 'project method' as popularized by his doctoral adviser, William H. Kilpatrick, since the students at the 'experimental school' attained higher scores on standardized tests in writing, reading and arithmetic as well as in social skills, habits and attitudes than the students at the two 'control schools'. Collings's book is a classic of progressive education, and his story of how 10 students were successful in combating an outbreak of typhoid fever in their community is well known among historians and educators even today. A re-examination of the dissertation - in particular of the so-called 'typhoid project' - reveals, however, that the experiment never took place as described and that Collings reconstructed his data substantially in order to conform to Kilpatrick's frame of reference and to present convincing data on the possibility and superiority of child-centred education.
It was in the fall of 1920 that Ellsworth Collings and William Heard Kilpatrick met for the first time. Collings, a thirty-three-year-old county superintendent from the Midwest, had earned a bachelor of science and a master of arts degree from the University of Missouri. Now he aspired to a doctorate from Teachers College of Columbia University in New York City, with Kilpatrick--the educational philosopher, project pedagogue, and disciple of John Dewey--as his advisor. Collings had already conceived a proposal for his dissertation. He suggested reporting on an experiment he was currently directing in his school district. Kilpatrick gave his approval, both to Collings as his doctoral student and to the experiment as the dissertation topic.
Nine months later, in August 1921, Collings met Kilpatrick again. He had just resigned his position as county superintendent and was now enrolled at Teachers College as a full-time graduate student. At the conference, Collings presented Kilpatrick with a new outline. Kilpatrick was disappointed. 'Discuss with Collings his proposed dissertation,' he wrote in his diary (11 August 1921), 'and I succeed fully to my satisfaction and apparently to his, that he was working on a specious and useless plan.' Kilpatrick consulted the Committee on Higher Degrees about the proposal, and the members shared his view. '[They] agree to reject his new plan and go back to the original scheme,' Kilpatrick reported (Diary, 12 August 1921). '[Collings] is much pleased. . . . [He is] working along the project line.' A year later, Kilpatrick held the completed dissertation in his hands.
Having read Collings's work in its entirety, Kilpatrick was excited. 'He proves (in this instance) just what I have wished to believe but hardly dared to hope for,' Kilpatrick wrote (Diary, 13 July 1922). 'I am glad to have it for reference' (Diary, 21 December 1923). Conducted at a rural school in Bethpage, McDonald County, Missouri, from 1917 to 1921, Collings's experiment had centered on an attempt to realize the 'democracy of childhood.' There were no required courses, no prescribed curricula at the school. Instead, the children themselves--their interests and needs, the issues and problems which concerned them personally--determined the subjects they would study and the topics they would cover. The teacher served only as an advisor and facilitator. But Collings did not merely describe this 'project curriculum;' he also presented empirical data which were supposed to prove that the students at the 'experimental school' attained higher scores on standardized tests than the students at two traditionally conducted 'control schools.' For Kilpatrick, Collings's experiment confirmed his belief in child-centered education. 'It can now no longer be said that the theory won't work,' Kilpatrick announced proudly. 'It has worked. A régime of child purposing is feasible. We can lay aside school subjects as such and succeed,--succeed admirably' (1923: xxiii).
Collings's dissertation was published by Macmillan in December 1923 under the title An Experiment with a Project Curriculum. It was reviewed ten times, ran through five printings, came out in two foreign languages (Russian and German), and was read and studied by thousands of teachers and students all over the world--in Great Britain, India, and South Africa as well as in Brazil, Switzerland, and the Soviet Union (DeLima 1926, Arndt 1929, Pinkevitch 1929, McCallister 1931, Jacob 1932, Tenenbaum 1951). Since then, the book has lost nothing of its appeal and attraction. In the United States historians consider Collings's experiment as a classic of progressive education and as a forerunner of the 'Eight-Year Study' (Alberty and Alberty 1962, Bleeke 1968, Hines 1972, Willis et al. 1993), while in Europe--where the project method has experienced a tremendous resurgence during the last twenty years--educators regard his report on how ten students successfully combated an outbreak of typhoid fever in their community as an inspiring illustration of both Dewey's concept of learning and Kilpatrick's method of teaching (Knox 1961, Rasborg 1968, Röhrs 1977, Schäfer 1988). German teachers in particular revere what is called the 'typhoid project' as the model for social education and the standard by which school reform, curriculum development, and student participation are to be measured even today (Pütt 1982, Krauth 1985, Frey 1993).
Yet the 'typhoid project,' as Collings presents it in An Experiment with a Project Curriculum, is fiction. It never took place as described. The children themselves determined neither the project's content nor its direction. Little about its design or implementation lay in their hands. Contrary to what the report suggests, the 'typhoid project' was planned in advance. The teacher prepared the lessons by selecting the subject matter and material and giving thought to what questions he would ask, what discussions he would pursue, and what activities he would propose. Collings had no doubts about the purpose and value of standard curricula, structured courses, or standardized tests. Indeed, the 'experiment' in Bethpage ran along traditional lines. There was none of the 'free and spontaneous' learning that educators have admired and tried to duplicate in their own schools for more than half a century. To put it simply, Collings stretched the truth and twisted the facts. Kilpatrick's essay on the 'project method' of 1918 is not, of course, the foundation for Collings's experiment with the 'project curriculum' of 1917. The 'typhoid project' does not provide proof that the impossible is possible and that nine- to twelve-year-olds are able to organize school and to change society by themselves. We have been the victims of deception. [...]