





































Action Research: An Early History in the U. S. 

Arthur W. Foshay 

Action research is a strategy that 
seeks to improve the quality of action 
through systematic examination of the 
processes and the consequences of 
action. In the United States, as 
applied to Education, it may be said 
to have had its beginnings in 1926, 
with the publication of a book by B. 
R. Buckingham, then Director of the 
Bureau of Educational Research at The 
Ohio State University, called Research 
for Teachers. Writing out of the 
"certainties" of the Scientific 
Movement in Education of the Twenties, 
Buckingham instructs teachers in the 
use and interpretation of standardized 
tests, and offers some advice on 
tool-making. The book had very little 
influence; it went out of print 
shortly. 

During the Thirties, extensive 
curriculum revision was attempted. The 
most prominent leader of this work was 
H. L. Caswell, first at Peabody 
College in Nashville, Tennessee, then 
at Teachers College, Columbia 
University, in New York. During that 
decade, Caswell made a number of 
discoveries, which he later summarized 
in his 1950 book, Curriculum 
Improvement in Public School Systems. 
One of these was that teachers would 
not adopt innovations from outside 
their classrooms. They had to "own" 
(Nelson Haggerson's term) an 
innovation before they would try it. 
This has been rediscovered by Bowe, 
Ball, and Gold, in a recent report of 
research on the consequences in The 
United Kingdom of the Educational 
Reform Act of 1988. They found that 
educational reforms must suffer 
"contextualizing" -i.e., being given a 
"local habitation and a name"--in the 
course of being adopted, even under 
official mandate. Teachers believe 
their own eyes, not someone else's. 

Caswell put this early finding to 
work when he founded the Horace 
Mann-Lincoln Institute of School 
Experimentation, first by creating 
what is now called a "school network" 
that extended across the country, and 
second by urging the staff of the new 
Institute to undertake classroom 
experimentation. 

Shortly after the end of World War 
II, Stephen M. Corey became Executive 
Officer of the new Institute. 
Influenced by Caswell, Corey took 
advantage of the work of John Collier, 
Director of the U. S. Bureau of Indian 

51 

Affairs, who had developed what was 
called "operations research", an 
approach that sought to improve action 
by applying systematic data-gathering 
to decisions in the field. The 
approach was immediately adopted by 
the military, and it attracted the 
attention of those involved in action 
in many fields. 

Corey made an important addition to 
operations research, however. 
Remembering the reluctance of 
classroom teachers to try someone 
else's practices, Corey suggested that 
the teachers be drawn in to the design 
of the classroom research, and that 
they participate in the gathering and 
interpretation of whatever data were 
collected. This approach he called 
cooperative action research, and that 
name held. 

Members of the Institute staff 
proceeded to do research with 
teachers, resul ting in several 
publications: Developing a Curriculum 
for Modern Living, by Florence 
Stratemeyer, and Ways of Studying 
Children, by Ruth Cunningham, both 
with members of the Denver school 
staff. Another was Children's Social 
Values, an Action Research Study, 
which I did in partnership with 
Kenneth Wann and the faculties of six 
elementary schools in Springfield, 
Missouri. 

The cooperative action research 
movement was summarized in a 1953 book 
by Stephen M. Corey, Action Research 
to Improve School Practices. Corey 
documents the origins of the idea, 
explains why it is necessary to 
conduct classroom experimentation with 
teachers if one seriously intends to 
have an effect on school practice, and 
provides illustrations of the process. 
A number of articles on action 
research appeared in the professional 
journals between 1949 and 1957. The 
1957 Yearbook of the Association for 
Supervision and Curriculum Development 
(ASCD), "Research for Curriculum 
Improvement", which I edited with 
James Hall, was largely devoted to 
action research. It contained chapters 

several well-known curriculum 
of that time: Corey, Matthew 

Miles (at that time a member of the 
Horace Mann-Lincoln Institute), Ross 

Hilda Taba, McKim, 
Herrick, among others. 



The chief limitation of cooperative 
action research, from the point of 
view of the educational researchers of 
that time, was that it was not 
possible to generalize from the 
examined population to others, because 
no attempt was made to see whether the 
examined population was representative 
of a larger population. In addition, 
since much of the research was 
designed and carried out by classroom 
teachers, who are not trained in 
research, the data often were flawed. 
For these reasons, the movement was 
ridiculed in the publications of AERA, 
and it did not spread. It disappeared 
as the members of the Institute staff 
scattered with the passage of time. 

During the Eighties, elements of 
the process reappeared, though with 
different names: "networking", 
"classroom research", "site-based 
management and inquiry", and, finally, 
under its original name, "action 
research" • 

It may be useful to consider what 
was learned about this process during 
its lifetime in the Fifties. 

For one thing, it did not turn out 
to be classroom experimentation. The 
work was descriptive, not 
experimental. What the teachers 
wanted, and could do well, was to 
examine, systematically, certain 
aspects of the children's behavior. 
The results of these examinations were 
not, directly, changes in school 
practice. The results were improved 
theory, or perhaps better, improved 
questions about children's 
development. For example, 
Stratemeyer's work in Denver 
contributed the concept of "persistent 
life situations" to the literature of 
child development, such as the 
changing answer (with development) to 
the question "What shall I eat?", or 
"How shall I deal with the opposite 
sex?" Stratemeyer proposed that 
questions such as these be addressed 
directly in the classroom, since they 
are of such basic importance. Again: 
in Springfield, we found that the 
teacher played a much greater role in 
the children's self-esteem than we had 
supposed. 

One of the most durable 
contributions to practice made by 
these projects can best be called 
"tool-making". Cunningham's Ways of 
Studying Children was just such a 
compilation of tools for gathering 
useful information about children's 
social development, their attitudes, 
and their achievement. When I returned 

52 

to Springfield five years after the 
completion of the project there, I 
found that what remained in use was 
the sociometric devices we had found 
and developed. It is safe to say that 
nothing remains in Denver of 
Stratemeyer's work, though the ideas 
she put forth have entered curriculum 
theory in various forms. 

The other contribution, which was 
limited to the members of the 
participating staffs, is an array of 
interesting and pleasant professional 
memories. The staffs clearly were 
pleased to carryon hours of 
professional thought and talk with the 
university consultants--but the talk 
ended when the consultants left; 
conversation in the teachers' lounges 
returned to gardening, an occasional 
anecdote about a difficult child, and 
personalities. 

Not so for the consultants. They 
remembered and reported what had been 
discovered. However, since the 
findings could not be generalized to 
unexamined populations, they were 
generally ignored. 

Nevertheless, the findings remain 
intriguing, For example, one 
elementary school staff in Springfield 
"hunched" (hypothes ized) that 
children's social interactions became 
more complex as they grew older. They 
defined "complex" as the number of 
steps required to bring the 
interaction to completion. The 
hypothesiS was rejected when a teacher 
observed a seven-year-old girl going 
through at least thirty steps, and 
taking more than a month, to prevent 
another girl from being invited to a 
third girl's birthday party. Or, 
again, we had speculated a good deal 
about attitude change. Gordon Allport 
had defined attitudes as follows: an 
attitude is a disposition to find a 
given referent aversive or valuable. A 
child may like or dislike Math, for 
example. Moreover, any attitude has a 
degree of strength, or intensity. A 
child likes or dislikes Math (a 
referent) in some degree--from much to 
little. Or perhaps the child doesn't 
care one way or the other--he is 
apathetic. Thus, attitudes have 
valence (direction), and . We 

that a change in 
from, say, an aversion to a value 
would be gradual. The aversion would 
become weaker and weaker, until the 
attitude ic, whereupon its 
referent acquire-gradually 

value. Not so. What we 
found was that it was very, very 
difficult to change an attitude at 



all, but that if it did change, it 
flipped; the valence might change, but 
the intensity remained constant. The 
change matched exactly what we had 
heard about religious conversions. The 
child who changed liked the referent 
as much as he/she had disliked it, and 
presumably vice versa. 

In another study, we contrasted 
physical and verbal aggression, 
supposing that response to verbal 
aggression would increase as the 
children grew older. Not so. Second 
graders responded to verbal aggression 
more than did the sixth graders. 

In these studies during the late 
Forties and early Fifties, we were 
much influenced by the social 
psychologists who had followed after 
Kurt Lewin, who had found American 
classrooms to be as authoritarian as 
Nazi Germany, whence he had recently 
fled. These psychologists were 
primarily concerned with group 
behavior, and we were, too. Hence the 
teachers' desire to study the social 
behavior of children in Springfield, 
and Cunningham's book about the group 
behavior of boys and girls. 

So influenced, we used the tools 
that had been developed for such 
study: projective devices, such as 
unfinished stories, and attitude 
questionnaires, such as "things that 
make me feel important", or "things 
that make me want to strike back". We 
also used what Herbert Stolz once 
called "systematic, incidental 
observation", and sociometric 
analyses. We never did get round to 
experimentation to improve academic 
performance. The prevailing talk in 
educational circles of that time was 
heavily influenced by the external 
situation: the war against 
dictatorship had just ended, and we 
were deeply concerned with the 
preservation and improvement of 
democracy at home. The Cold War was 
just beginning. I had to include an 
explanatory paragraph in my social 
values book to defend it against the 
right-wing extremism that had appeared 
in Pasadena, in Ohio, and in 
McCarthyism. Springfield, Missouri, 
where the project was undertaken, was 
in the American Bible Belt--a seat of 
religious fundamentalism--and the 
teachers there consequently cared 
deeply about questions of social 
morality. It showed in the schoolyard, 
where, instead of playing "cops and 
robbers", the children often 
"Jesus and the Pharisees". seems 
hard to believe, but I was told that 
it was common. 

53 

As I observe the present scene, I 
find almost no residue from the 
Cooperative Action Research movement 
of the Forties and Fifties. It may be 
well, therefore, for me to offer a 
purely subjective account of what 
those who now attempt action research 
should bear in mind. 

First, remember that teachers have 
so far found educational research of 
little use as applied in the 
classroom. It follows that action 
research, if it is to improve the 
quality of action, has to come from 
the inside out, not the top down or 
the outside in. It has to originate in 
classrooms, not in constructed lab 
situations. Second, remember the 
function of theory: it exists to 
clarify reality. It follows that 
outside consultants can bring theory 
to bear on actuality in the classroom, 
provided that the theory clarifies the 
actuality. For example, we found 
Allport's definition of attitudes very 
useful in understanding the attitudes 
of the children we faced in 
Springfield. 

Third, keep it simple, or the 
teachers will be skeptical. For 
example, I once worked with some 
teachers in a suburb of Columbus, 
Ohio; they worried that children in 
the Third Grade couldn't seem to deal 
with "word problems", such as "Johnny 
went to buy a dozen apples at five 
cents each. What did the apples cost 
Johnny?" I suggested that the teachers 
assign as homework that the children 
make up their own word problems. It 
worked. The children enlisted their 
parents. Some of the problems they 
brought in could not be solved because 
there wasn't enough information 
(Johnny went to the store to buy some 
apples. How much did they cost?) Some 
were complicated with useless detail; 
some were too simple; and so on. The 
performance of the children on the 
"word problems" in the textbook 
improved dramatically. Nobody 
quantified anything to support the 
conclusion that it worked. It was 
obvious. 

Fourth, learn how to be an 
effective classroom consultant. The 
human relationship between a 

consultant and an 
classroom teacher requires 

direct attention by the consultant. 
The relat will always be 
polite, but the teacher can be 

to test the consultant's 
sincerity constant For one thing, 
teachers usual with the 
assumption that university 



consultant hasn't been in a classroom 
for years, and knows nothing of the 
realities of teaching in the lower 
schools. Don't parade your knowledge. 
Instead, seek the teacher's opinions, 
and pay close and respectful attention 
to them, because the teacher knows the 
immediate situation better than you 
ever can. 

Fifth, learn how the teacher 
thinks. I once was asked to find out 
how a certain kindergarten teacher 
managed, year after year, to produce 
such independent children. The 
children entered more or less 
infantile, and emerged self-respecting 
young people, full of initiative and 
independence. I asked her how she did 
it. Her response was to tell me story 
after story about individual children 
she had known. She was about sixty 
years old, and she seemed to remember 
every child she had ever faced. I 
didn't know how to listen to her 
stories, so I learned nothing that 
way. I sat in her classroom every 
morning for a month, and it finally 
dawned on me that she gave the 
children no alternative to being 
independent. If there was noise in the 
hall during a class planning session, 
she asked "Does anyone else hear the 
noise in the hall?", and one of the 
children got up and closed the door. 
When a boy brought her a boat he was 
carving and asked if it was all right, 
she asked, "Are you satisfied?", and 
he returned to work on it some more. 
When a child asked her to tie his 
shoe, she asked "If I tie the second 
knot, can you tie the first?" Of 
course the children became 
independent. They had no alternative. 
I was looking for technique, and I 
finally saw some. If I had learned how 
to listen to her anecdotes, I might 
have learned much more than I did. I 
might have gone beyond technique to 
ideas. But I didn't, and I'm the 
poorer for it. 

Finally, be modest in reporting 
action research. We live in a time 
when exaggerated claims for 
educational accomplishments are almost 
the norm. If our main 
audience--teachers in service--is to 
pay attention to what we report, we 
have to explain to them how they may 
verify what we say, and we must 
include the false starts and the 
mistakes in the report. Education is 
not yet a science; there is very 
little that we know without doubts. 
Action research continues to be 
promising, but-we must be aware of its 
limitations as well as its 
possibilities. 

54 

References 

Bowe, Richard, and Stephen J. Ball, 
with Anne Gold. Reforming Education 
and changing Schools. London and 
New York, Routledge, 1992. 

Buckingham, B. R. Research for 
Teachers. New York, Silver Burdette 
and Company, 1926. 

Caswell, Hollis L., and Associates. 
Curriculum Improvement in Public 
School Systems. New York, Bureau of 
Publications, Teachers College, 
Columbia University, 1950. 

Collier I John. "United States Indian 
Administration as a Laboratory of 
Ethnic Relations." Social Research, 
12:265-303, May 1945. 

Corey, Stephen M. Action Research to 
Improve School Practices. New York, 
Bureau of Publications, Teachers 
College, Columbia University, 1953. 

Cunningham, Ruth. Ways of Studying 
Children. New York, Bureau of 
Publications, Teachers College, 
Columbia University, 1949. 

Foshay, Arthur W., editor. Research 
for Curriculum Improvement, the 
1957 Yearbook of the Association 
for Supervision and Curriculum 
Development. Washington, D. C., The 
Association, 1957. 

Foshay, Arthur W., and Wann, Kenneth 
D. Children's Social Values, an 
Action Research Study. New York, 
Bureau of Publications, Teachers 
College, Columbia University, 1953. 

Foshay, Arthur W. "Considerateness 
and Agression." Educational 
Research Bulletin, v. 32, no. 4 
(April 8, 1953), pp. 27-58. 

Stratemeyer, Florence, and Associates. 
Developing a Curriculum for Modern 
Living. New York, Bureau of 
Publications, Teachers College, 
Columbia University, 1947. 


