' yum”.-.
‘:\' ‘-
.~, . 5

 

.. '13'...5.‘.‘.‘;‘Vy",'-21'£’ ‘,r:-.:.~I{»}~,n\v ',..1',‘»_.'«“'b'u_‘ .I w. * u r‘ :2: :11::'.::..:v..:.-'- "3 ~ .33: r; 7“?“- ‘-'"'vJ“-I‘i‘f€ji"31"3"““'1f-f"~"~"I"‘?‘-€"‘:'T‘-‘ M"? '. ““"jwmI"7915" " . w .' .. ” ' “h v " 1-
- ".1333.W‘Jfl‘ndnh"" ., - >. .. <,. . I , ‘ _ ‘ I ' ' ‘ r‘ - N- - u 3:.“ ma“. ""‘ ‘ ‘ v w ‘ . ' " '

 

A STUDY OF A PROCESS T0” meme-y » >_ ~ if
VERBAL INTERACTION PATTA'ERNSIOF
HIGH. SCHOOL GEOMETRY TEACHERS : j j, g »;_;_~ij;;;:-. .

Thesis for the. Degree of Ph. D“
MICHIGAN STATE UNIVERSITY
WILLIE ELBERT WILLIAMS . '

' 11972. 77 '

 

  
 
  

LIBR A R Y
Michigan State

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\l University

“‘3 1293 10385 8779 __

M

 

l

l

This is to certify that the l
thesis entitled

A Study of a Process to Modify Verbal
Interaction Patterns of High School
Geometry Teachers

presented by

Willie Elbert Williams

has been accepted towards fulfillment
of the requirements for

Ph.D. _degreein Secondary Education

 

 

[5/ mg m. FT gala

Major professor

Date e tember l 1972

0-7639

 

 

 

 

--.
--.' ”b q'
Q“ ‘r-

311:0!“ av g 3;
MG & SUNS' ..
BOOK BINDERV INC. ‘!

_ LIBRARY BINDERS
IPIIIOPOII.IICIIGAI l t

ABSTRACT

A STUDY OF A PROCESS TO MODIFY
VERBAL INTERACTION PATTERNS OF
HIGH SCHOOL GEOMETRY TEACHERS

BY

Willie Elbert Williams

PUIEOSG

The purpose of this study was to investigate
attempts to modify teachers' verbal behavior in a large
inner city high school in Cleveland, Ohio, characterized
by direct or indirect using Flanders' ten category system
of interaction analysis.

Much of the research on teacher effectiveness
has found that indirect teacher influence produced more
achievement gain, while direct teacher influence has pro-
duced less. The study was designed to determine if teachers
of direct influence patterns could be encouraged toward a

more indirect or flexible pattern of influence.

Procedure

 

Twelve instructional staff members, teaching geome-
try, were the subjects of this study. The study used all
the geometry teachers in one senior inner city high school,

containing grades 9-12.

Willie Elbert Williams

During the first week of school, all teachers were
introduced to Flanders' method of interaction analysis.

Two weeks later, each of the twelve geometry teachers was
observed (their initial observation taped), using Flanders'
ten category system of interaction analysis. The initial
observation was taped so that an adequate characterization
of their influence could be made as to direct or indirect.
Also, four of the tapes were analyzed a second time to
determine the consistency of the observational techniques.

The results of the initial observation revealed
three indirect and nine were direct. Five direct teachers
were randomly selected from the nine. The geometry students
of all teachers were automatically selected for each class—
room by computer assignment. Three groups were determined:
Group I, three indirect teachers (control); Group II, four
direct teachers (control); and Group III, five randomly
selected direct teachers (experimental). Only one observer
was used, and he had been trained in interaction analysis
through a program learning approach. The investigator was
this observer.

During the next two weeks a conference was held
(Treatment #1 for Group III) with each of the five teachers
of Group III. All conferences were designed for thirty
minutes of a forty minute period. Each teacher was given
a copy of the results of the initial observation to form a

basis for the discussion in the first conference. Each

Willie Elbert Williams

teacher was given his tape to play back on his own, a brief
explanation of Flanders' ten category method of interaction,
an explanation of acceptable limits for indirect influence
patterns, and details of how to become more indirect in his
teaching. The five teachers were observed again (Observa-
tion #JJ, and a conference was held (Treatment #2) within

a week. Each teacher was given a copy of the results of

the initial observation and observation #1 to form a basis
for discussion and comparison in the second conference.
Again, each teacher was given a brief explanation of Flanders'
ten category method of interaction, an explanation of accep-
table limits for indirect influence patterns, and details

of how to become more indirect in his teaching. No taping
was made of any other observation. At the end of six weeks
from observation #1, the five teachers were observed again
(observation #2) and a conference was held (Treatment #3)
within a week. Each teacher was given a copy of the results
of the initial observation, observation #1, and observation
#2 to form a basis for discussion and comparison in the
third conference. Again, each teacher was given a brief
explanation of Flanders' ten category method of interaction,
an explanation of acceptable limits for indirect influence
patterns, and details of how to become more indirect in his
teaching. At the end of twelve weeks from observation #1,
the final observation (observation #3) of the five teachers

selected for treatment was made.

Willie Elbert Williams

The treatment was the three conferences which pre-
ceded the three observations of the five teachers selected
at random from the original twelve.

The paired t-test was used to measure significant
differences, at the .05 level, of the mean I/D scores be-

tween Groups I, II, and III.

Findings

1. On a continuum from .56 to 2.50, the teachers.
in Group III moved up so that their mean I/D ratio did not
differ significantly with those in Group I.

2. Group III became more indirect in their influ-
ence pattern.

3. Group III failed to move up the continuum so
that there was a measurable significant difference between
them and the teachers of Group II.

4. Group III became more indirect with time, but
failed to reach the point of indirectness of Group I.

5. The mean percent of teacher-talk was 67.6.

6. The mean percent of student-talk was 19.1.

On the bases of the findings of this study, and
within its limitations, the following conclusions seem
justified:

1. Teachers can be encouraged toward a more in—

direct influence pattern of teaching.

Willie Elbert Williams

2. Direct teachers can be encouraged to-a more
indirect influence pattern using Flanders method of inter-
action analysis.

3. Students appeared to achieve more in teacher's
classes where the influence pattern is indirect.

4. Teacher's behavior in the classroom can be

affected by a conference using Flanders' method as a basis.

Discussion

 

In this study, the observational evidence that was
obtained does indicate that the teachers' behavior was
modified in the experimentally desired direction. However,
the hypotheses suggest that maybe a longer period should be
used across time, with more observations being made, if
there is to be a more appreciable change in teacher behavior
toward indirectness.

Samph found in his study (65), that teachers became
more indirect when an observer was present in their class-
room, whether they were informed of an observation prior
to its occurrence or not. This may be true, but the writer
believes that by also training teachers in the use of
Flanders' method of interaction, one can eliminate much of
the observer effect.

The findings that the mean percent of student-talk
was 19.1 is in accord with Stilwell's findings (66), teacher-
talk consumes approximately three times as much time as

student-talk.

Willie Elbert Williams

The achievement scores in the teachers' geometry
classes who became more indirect in their influence (Group
III) was higher than either of the other groups, but since
no pre-test was given, no statistical inference can be made

about the differences.

A STUDY OF A PROCESS TO MODIFY

VERBAL INTERACTION PATTERNS OF

HIGH SCHOOL GEOMETRY TEACHERS
BY

Willie Elbert Williams

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Secondary Education

1972

ACKNOWLEDGMENTS

I am greatly indebted to a number of peOple for
the completion of this study. Without their support and ~
constructive criticism, this endeavor would have been
impossible. However, I wish to particularly acknowledge
the contributions of:

Dr. William M. Fitzgerald, chairman of my doctoral
committee and friend, for his counSel, his trust, and his
invaluable advice and support.

I also appreciate the advice and support of the
other members of the committee: Dr. John Wagner, Dr. Glenn
Berkheimer, and Dr. Samuel Corl.

The mathematics teachers of Glenville High School,
without their support and cooperation, the experiment for
this study would not have been possible.

Dr. Andrew Porter for his valuable advice.

My beloved wife, Doris and my darling children:
Billy, Elaine, and Dorwyl for their support and under-
standing for the past three years. For, truly, without a
family that is willing to make sacrifices, this doctoral

program would have been highly impossible for me.

Willie Elbert Williams

ii

TABLE OF CONTENTS

Chapter Page
I. A DESCRIPTION OF PROBLEM AND THE STUDY . . . . 1
Introduction . . . . . . . . . . . . . . . . 1
Background of the Problem . . . . . . . . . 2
Statement of the Problem . . . . . . . . . . 3
Significance of the Problem . . . . . . . . 3
Limitations of the Study . . . . . . . . . . 6
Operating Procedure . . . . . . . . . . . . 7
Overview of the Study . . . . . . . . . . . 9
II. A REVIEW OF PERTINENT LITERATURE . . . . . . . 11
Introduction . . . . . . . . . . . . . . . . 11
Review of the Literature . . . . . . . . . . 14
Relation of Review of Literature
to the Present Study . . . . . . . . . . . 22
Summary . . . . . . . . . . . . . . . . . . 23
III. DESIGN OF THE STUDY . . . . . . . . . . . . . 25
Introduction . . . . . . . . . . . . . . . . 25
The Population . . . . . . . . . . . . . . . 25
Interaction Analysis . . . . . . . . . . . . 26
Measures . . . . . . . . . . . . . . . . . . 28
Methodology . . . . . . . . . . . . . . . . 28
Treatment #1 . . . . . . . . . . . . . . . 30
Treatment #2 . . . . . . . . . . . . . . . 31
Treatment #3 . . . . . . . . . . . . . . . 32
Statistical Hypoth ses . . . . . . . . . . . 33
Hypothesis 1 . . . . . . . . . . . . . . . 33
Hypothesis 2 . . . . . . . . . . . . . . . 34
Hypothesis 3 . . . . . . . . . . . . . . . 34
Summary . . . . . . . . . . . . . . . . . . 34
IV. ANALYSIS OF DATA . . . . . . . . . . . . . . . 36
Hypothesis 1 . . . . . . . . . . . . . . . . 38
Hypothesis 2 . . . . . . . . . . . . . . . . 39
Hypothesis 3 . . . . . . . . . . . . . . . . 40
Sub-Hypothesis 1 . . . . . . . . . . . . . 40

iii

Chapter Page

Sub-Hypothesis 2 . . . . . . . . . . . . . 41
Additional Comparisons . . . . . . . . . . . 42
Non-Verbal Behavior . . . . . . . . . . . . 45
Other Behavior . . . . . . . . . . . . . . . 46
Cases for Group III . . . . . . . . . . . . 46
Achievement . . . . . . . . . . . . . . . . 49
Summary . . . . . . . . . . . . . . . . . . 52

V. SUMMARY, FINDINGS, CONCLUSIONS, AND

RECOMMENDATIONS O O O O O O O O O O O O O O O 5 4
Introduction . . . . . . . . . . . . . . . . 54
sumnlary O O O O 0 O O O O O O O O O O O O O 54
Findings 0 O O O O O O I O I O O O O O O O O 55
DiSCUSSion O O O O C O O I O O O O O O O O O 57
Recommendation for Future Research . . . . . 58
Questions for Further Study . . . . . . . . 58

BIBLIOGRAPHY O O O O O O O O O O O O O O O O O O 60
APPENDICES O O O O O O O O O O O O O O O O O O O O 68
APPENDIX

A-l Notice Of Observation 0 o o o o o o o o o o 68

A-2 Conference Sheet I,

Flanders' Interaction Analysis System . . . 69

A-3 Initial Conference Sheet II . . . . . . . . 71

A-4 Conference Sheet . . . . . . . . . . . . . 72

B Non-Verbal Record Sheet . . . . . . . . . . 74

C-1 Verbal Categories of FIA . . . . . . . . . 76

C—2 Summary of Categories for FIA . . . . . . . 82
C-3 Procedure for Categorizing Teacher-Pupil

Interaction . . . . . . . . . . . . . . . 84

D-l Analysis Sheet . . . . . . . . . . . . . . 87

D-2 10 x 10 Data Matrix . . . . . . . . . . . . 88

iv

Table

4.1

LIST OF TABLES

Characteristics of the twelve teachers

after initial observation . . . .

Summary data for the paired t-test,
Group I, III C O O O O O O O O 0

Summary data for the paired t-test,
pre and post mean of Group III .

Summary data for the paired t-test,
Group II, III . . . . . . . . . .

Summary data for paired t-test, all
three groups . . . . . . . . . .

Comparison of percent of teacher-talk

Post-Percent of teacher-talk for Group III.

Summary data for mean percent of
teaCher-talk O O O O I O O O O 0

Summary of Pre and Post I/D scores
Of Group III C O O O O I O O O 0

Summary of Pre and Post Percent of
teacher-talk of Group III . . . .

Summary of Pre and Post Percent of
student-talk in each class of
teachers of Group III . . . . . .

Summary of mean achievement scores in

each class in each group . . . .

Summary of means of each group and
overall mean . . . . . . . . . .

Page

36

39

39

40

41
42

43

43

44

44

45

50

51

CHAPTER I

A DESCRIPTION OF THE PROBLEM AND THE STUDY

Introduction

 

The cost of education is increasing and will in-
crease more. In recent years, there has been a lot of talk I
about accountability and performance contracting, and
parents are taking a harder look at what they are getting
for their money--increases in taxes. Hence, there will be
a greater effort in the future made by subject area super-
visors and department chairmen to up-grade teaching and
make teachers more responsible for the kinds of products
they are producing.

The general purpose of this study is to see if by
using Flanders' interaction analysis, one can improve the
role of the teacher in guiding his pupils learning. Thehii
relationship between teacher characteristics and student
behavior has been studied for many years. This research is
generally denoted as "teacher effectiveness". It appears
to be axiomatic, on the part of many educators, that the
teacher has some effect on his students. Students appear
to achieve better under one teacher than another teacher

even if all other aspects of the learning situation are

held constant. It is also realized that the teacher may
affect many aspects of student behavior other than achieve-
ment. However, achievement is an important goal of instruc-
tion, and this study will confine itself to the discussion
of achievement in terms of scores on standardized achieve-

ment tests.

Background of the Problem

 

As a department chairman in a large inner-city high
school in Cleveland, Ohio, one of the writer's main functions
is to evaluate teachers and instruction. This is being
done, as in many school systems, with instruments based on
value judgments. Much of this evaluation lacks systematic.
methods of inquiry. In recent years, researchers have

examined classroom verbal behavior--the sine qua non of

 

teaching--by systematic methods of inquiry. Several systems
have been constructed which tend to categorize this behavior.
Since it appears, by many educators, that it is axiomatic
that the teacher is the most influential person in the
classroom, and since talk is such a vital part of teaching,
and since the teacher's verbal behavior has a direct influ—
ence on the pupil's verbal behavior, it follows that teacher
talk is very important to education. It is further believed
that teachers can control their verbal behavior in order to
make their teaching less accidental, haphazard, and routine;

and consequently improve their teaching styles.

Statement of the Problem

 

The purpose of this study was to investigate the
influence of teachers verbal behavior in a large inner-city
high school in Cleveland, Ohio, where the teachers were
characterized as direct or indirect by using Flanders ten
category system of interaction analysis. Further, its
purpose was to investigate, through the case study technique,
the effects that these influences induced on student's
achievement in geometry at the school.

Since much of the research on teacher effectiveness
has found that indirect teacher influence produced more
achievement gains while direct teacher influence has pro-
duced less, the purpose was, also, to determine if the in-
fluence of the teacher was directed toward a more indirect
point of view in geometry; would students achieve more than
those under a more direct influence. Also, the purpose of
this study was to determine if teachers of direct influence
patterns could be directed toward a more indirect or flexible

influence.

Significance of the Problem

 

A more scientific method for evaluating teacher
effectiveness is a very important need in education today.
Educational systems are in a continuous search for ways to
improve teaching and create better conditions by which

children can learn more and better. Amidon (3) said,

"There is general agreement among educators that
teaching needs considerable improvement. Why with all
the effort put into the preparation of teachers, into
in-service education of teachers, and into teacher's
individual efforts to modify and improve their work
with youngsters, is not teaching far more effective
than it is? Why do researchers engaged in classroom
observation find that teachers are so controlling,
restrictive and inhibiting? Why is it that teachers
tend to do most of the talking (about 70 percent in one
average classroom, according to Flanders)? Teachers
would undoubtedly like to involve pupils more creatively
in the teaching process, would like pupils to partici-
pate more, ask more imaginative and thoughtful questions,
engage in more creative thinking. Why, then, do they
tend to teach as they themselves were taught as young-
sters, or as they see others teach in the schools,
rather than in ways they read about and talk about in
professional education courses? Why is it true today,
as it was in 1903 when John Dewey prepared this state-
ment, that 'The student adjusts his actual methods of
teaching, not to the principles which he is acquiring,
but to what he sees succeed and fail in an empirical
way from moment to moment. . . ." (3:p. 2-3)

This study was designed to affect a teacher's
verbal behavior at this school in such a way that they
would improve and grow, and their students would learn
more, and better.

For the purpose of this study the following assump-
tions are deemed necessary:

1. Most teacher influence is expressed through
verbal statements.

2. Most nonverbal influence is positively corre-
lated with the verbal influence.

3. Since the investigator was the only observer,
it was assumed that there was no observer effect.

4. Teaching can be submitted to systematic inquiry.

5. Teaching involves behavior which can be identi-
fied, and which teachers can systematically acquire.

6. Teachers want to improve their teaching.

Perhaps the best known system for analyzing inter-
action is attributed to Ned Flanders. Flanders' interaction
analysis is an instrument designed to observe instruction
in the classroom (10, p.l).

Flanders' basic interaction analysis contains ten
categories that are used to classify the statements of the
pupils and the teacher.* The ten categories include seven
assigned to teacher talk, two to student talk, and one to
silence or confusion. When the teacher is talking, the
observer must decide if the statement is: (l) accepting
student's feelings; (2) giving praise; (3) accepting,
clarifying, or making use of a student's ideas; (4) asking
a question; (5) lecturing, giving facts or opinions; (6)
giving directions; or (7) giving criticism. When a student
is talking, the observer must classify what was said into
one of two categories: (8) student response or (9) student
initiation. Silence and confusion is assigned to category
(10). All categories are mutually exclusive, yet totally
inclusive of all verbal interaction occurring in the class-

room (10, p. 33-35).

 

*Several modifications and extensions of Flanders
basic categories have been adapted, such as the Observa-
tional System for Instructional Analysis (Hough, 1967).

An indirect (category 1—4) - direct (category 5-7),
denoted by I/D ratio, focuses on the relative number of
indirect and direct teacher statements.

The hypotheses are given in testable form in
Chapter III, and stated briefly below:

First, is there a mean difference of the final I/D
ratio of the five direct teachers and the initial I/D ratio
of the three indirect teachers.

Second, is there a mean difference in the mean I/D
ratio of the initial observation of the total group and
the mean of the observations of the five selected teachers.

Third, is there a mean difference in the final I/D
ratio of the selected teachers and the initial I/D ratio

of the four teachers who were not selected.

Limitations of the Study

 

The purposes of this study have been clearly out-
lined in the statement of the problem. However, there are
some important limitations the writer wishes to mention
here.

1. The study is limited in its application to
geometry teachers at one senior high school in Cleveland,
Ohio.

2. This study is limited to the verbal behavior
of students and teachers in geometry classes at this

school.

3. Only two outcome variables, teacher influence
and student achievement, were evaluated in this study.

The limitations in this study are restricted to
the scope of the problem. This seems to be justified,
because in order to investigate any problem thoroughly, it
is necessary to focus ones attention upon as few factors
as possible.

However, the investigator feels that with the
limitations, and the information that can be obtained from
this study, it can be valuable, especially when this infor-

mation is correlated with similar studies.

Operating Procedure

 

At the beginning of the first semester, 1971-72,
the investigator introduced Flanders' method of interaction
analysis to the thirteen teachers of the department of
mathematics at the school, but only twelve of these teachers
taught geometry. Two weeks after the Opening of school,
each of the twelve geometry teachers were observed (their
initial observation taped) using Flanders' ten category
system of interaction analysis. The initial observation
was taped so that an adequate characterization of their
influence could be made as to direct or indirect. Also,
four of the tapes were analyzed a second time to determine

the consistency of the observational techniques.

From the initial observation, three teachers were
classified as indirect and nine were classified as direct,
using the classification direct if I/D ratio was less than
1.00, and indirect if I/D ratio was greater than or equal
to 1.00. The geometry students of all teachers are auto-
matically selected by a computer.

During the next two weeks a conference was held
with each of the five teachers of Group III. All conferences
were designed for thirty minutes of a forty minute period.
Each teacher was given a copy of the results of the initial
observation to form a basis for the discussion in the first
conference. Each teacher was given his tape to play back
on his own, a brief explanation of Flanders' ten category
method of interaction, an explanation of acceptable limits
for indirect influence patterns, and details of how to
become more indirect in his teaching. The five teachers
were observed again, and a conference was held within a
week. Each teacher was given a copy of the results of the
initial observation and observation 1 to form a basis for
discussion and comparison in the second conference. Again,
each teacher was given a brief explanation of Flanders'
ten category method of interaction, an explanation of
acceptable limits for indirect influence patterns, and
details of how to become more indirect in his teaching. No
taping was made of any other observation. At the end of

six weeks from observation 1, the five teachers were

observed again and a conference was held within a week.
Each teacher was given a copy of the results of the initial
observation, observation 1, and observation 2 to form a
basis for discussion and comparison in the third conference.
Again, each teacher was given a brief explanation of
Flanders' ten category method of interaction, an explana—
tion of acceptable limits for indirect influence patterns,
and details of how to become more indirect in his teaching.
At the end of twelve weeks from observation 1, the final
observation (observation 3) of the five teachers selected
for treatment was made.

The treatment was the three conferences which pre—
ceded the three observations of the five teachers selected

at random from the original twelve.

Overview of the Study

 

Chapter I included an introduction, background of
the problem, a statement of the problem, significance of
the problem, limitations of the study, and operating pro-
cedure. Chapter II includes a review of the literature
relevant to the present study.

Chapter III contains the experimental procedure
and design. A discussion and explanations of the variables,
and the techniques involved in measuring them is included
in the chapter also. Chapter IV contains an analysis of

the results of the statistical treatment of the study. It,

10

also, contains some non-verbal aspects of the teachers,
and case analysis of each of the teachers in Group III.
Chapter V contains a summary, findings, conclusions, and

recommendations for further study.

CHAPTER II

A REVIEW OF PERTINENT LITERATURE

Introduction

 

One of the most difficult problems in measuring
teacher effectiveness and student achievement in the class-
room is that of arriving at a satisfactory objective
measuring instrument that will objectively evaluate the
teacher's influence and how this influence relates to the
achievement of his students. Until about the last sixteen
years, the usual approach of this problem was to use rating
scales by supervisors or specially trained observers, but
despite of all these attempts to improve them, many such
ratings were based on value judgement, are still biased,
very subjective, and in many instances they are uninterpre-
table by anyone-—sometimes even the rater himself.

However disappointing these findings, several
promising trends seem to be emerging. Fey (9:p. 535-551),
who recently compiled a review of related research in
classroom teaching in mathematics, concluded in his summary:

. . . Recent research in mathematics teaching has
produced no major breakthrough in the search for per—
sonal characteristics, education or classroom behavior
of effective teachers. However, several promising
trends are emerging in the focus and techniques of
research."

11

12

The chief purpose of the present study was to use
Flanders' ten category system of interaction analysis to
analyze the verbal behavior in geometry classrooms, and
determine its effects on the achievement of the students
in these classes. Flanders' system was selected because
it appears to be one of the most objective methods developed
in the past twelve years, and geometry because of its highly
structural aspect. Further, it seemed most appropriate for
my design. This method seems to have great research possi-
bilities, and seems to be relatively independent of qualita-
tive judgement, and can be used reliably by trained
observers. The successful pursuit of the main objective
of this study involves six other basic problems:

1. Can the verbal behavior of teachers and their
students reliably and systematically be observed?

2. Does the verbal behavior of teachers observed
lend itself to any typical consistency?

3. After determining the relationship between the
teacher and student behavior, how does this attribute to
the amount of learning, or what manner would the teacher
be graded by his supervisor to determine his accountability?

4. Can Flanders' ten category method of interaction
analysis be used successfully in the classroom as a scien-
tific approach to measuring teacher effectiveness?

5. Can the level of instruction be raised through

the use of interaction analysis?

l3

6. Can a teacher's behavior be affected by a
conference?

There has been a lot of research in the past on
measuring the effectiveness of a teacher in the classroom,
but recently there is greater emphasis on a more scientific
approach to this research. The research indicates that
effective mathematics teaching must move beyond the measures
of ability, prior teaching experience, background, attitudes,
age, or knowledge of subject matter. Fey pointed out (what
seems to be the present trend in research):

". . . there is a growing realization that effective
teaching is a result of a complex interaction between
teacher ability, attitudes, and behavior; student
aptitudes and attitudes, and the structure inherent.in
mathematical topics. The traditional search for a
simple profile of a composite "good teacher" is giving
way to investigations that ask what kinds of teaching
style and subject matter organizations are most effeCe
tive for teaching a particular topic to some particular
student population". (8:p. 82)
Much of the previous research assumed that teacher effec-
tiveness is determined by one variable in the teacher-
learning situation--the teacher. However, Biddle and
Ellena states that:
"Unlike the factory worker, the teacher does not operate
upon the pupil in isolation from other agents. . . ..
. . . it is clear that competence involves a complex
interaction between properties and contextual factors
in the community, school and classroom". (7:p. 4-5)
However, the findings of Morsh, Burgess, and Smith revealed
that:
"Student gains can be reliably measured and that

student's ratings of their instructors' verbal facility
are correlated significantly with student gains". (24:p. l)

14

Many of these statements and the results of much

of the research on teacher effectiveness led to this study.

Review of the Literature

 

This review is concerned with teacher verbal
behavior over the past twenty years, and how this behavior
affects the achievement of students. However, none of the
research reviewed concludes that any one pattern of teacher
behavior is superior to another under all conditions, but
many do conclude that students taught by teachers with a
more indirect approach achieve more; while those taught by
a more direct approach achieve less. The present study
came about, basically, as the result of the latter state-
ments.

Also, this study deals with a more recent scientific
approach in analyzing the verbal behavior in the classroom--
Interaction Analysis-swith the idea that interaction is one
of the best reflections of classroom climate.

The earliest systematic studies of spontaneous
pupil and teacher behavior were those of H. H. Anderson (5)
as early as 1939. These studies were based on the observa—
tion of "dominative" and "integrative" behavior of teachers.

The findings of Anderson, e£_§l, are based on the
study of preschool, primary, and elementary school class-
rooms involving five different teachers and extending over

several years. Taken altogether, their imaginative research

15

has produced a series of significant findings. Those impor-
tant here are that if one type of contact predominates,
domination stimulates further domination, and integration
stimulates further integration. It is the teacher's prin-
cipal behavior pattern that spreads among pupils, and is
taken over by them even when the teacher is no longer in

the room.

About a year after Anderson started his work,
Lippitt and White (18) in 1940 working with Kurt Lewin
carried out laboratory experiments to analyze the effects
of adult leaders influence on boy's groups.

The pattern»Lippitt and White named "authoritarian
leadership" is similar to Anderson's dominative contacts;
"democratic leadership" was similar to integrative contacts;
while "laissez-faire leadership" consisted of irregular and
infrequent integrative contacts with a lack of adult initia—
tive that is seldom found in a classroom and was not present
in the Anderson studies.

Most of the conclusions of the Lippitt and White
study confirmed or extended the general conclusions of
Anderson, et_§l.

These two studies aroused considerable interest in
the analysis of teacher behavior. Additional research
revealed minor variations of the central theme. Withall
(44) in 1949 showed that a simple classification of the

teacher's verbal statements into seven categories produced

16

an index of teacher behavior almost identical to the inte-
grative-dominative I/D ratio of Anderson. He spoke of the
classroom as teacher-centered vs. learner—centered.

Perkins (37) in 1964, using Withall's technique,
studied groups who were studying child growth and develop-
ment occurred when group discussions were free to focus on
that tOpic; groups with an integrative type of leader were
able to do this more frequently than were pupils led by a
dominative type of leader.

In a large sectional study, which did not use
observation of spontaneous teacher behavior, Cogan (33) in
1963 administered a single paper-and-pencil instrument of
987 eighth-grade students in 33 classrooms. The instruments
contained three scales; a) a scale assessing student per-
ceptions of the teacher, b) a scale on which students
reported how often they did extra required school work,
and c) a scale on which students reported how often they
did extra non-required school work. Cogan's first scale
assessed traits that he developed in terms of Murray's list
of major personality needs. There were two patterns of
teachers on his scale. The items of one pattern were
grouped as "dominative", and "affiliative", and "hurturant".
These are close to Anderson's dominative and integrative
patterns. Cogan found that students reported doing more
assigned, and extra schoolwork when they perceived the
teacher's behavior as falling into the integrative pattern

rather than the dominative pattern.

17

One earlier research project does supply evidence
that teachers are flexible in their use of integrative and
dominative contacts. Mitzel and Rabinowitz (36) in 1953
observed four teachers using Withall's technique, and
organized their data to permit an analysis of variation
among teachers, visits, and observers. This concluded
that teachers adapt their influence to the immediate situa-
tion. They classified the influence of the teachers as
hostile or supportive, which also, follows Anderson's
pattern of dominative-integrative.

In a laboratory type experiment Amidon and Flanders
(11) have shown that dependent prone junior high school
students are more sensitive than average students to
differences in patterns of teacher influence, and that
dependent prone students learned less geometry when exposed
to a more rigid, direct pattern of influence compared with
a more indirect pattern.

There is no doubt in the mind of the writer that
all of these investigators are referring to highly similar,
even identical, dimensions of behavior, reliably measurable,
and important in educational theory. However, the important
question is left unanswered by the studies reviewed thus
far. Namely, since both integrative and dominative types
of statements are used by all teachers, including the most
excellent teachers, what are the consequences of these

different types of statements used under different conditions:

18

This question introduces the idea of flexibility of teacher
behavior--that a teacher may need to vary his behavior
under different conditions in order to achieve the desired
consequences.

In some recent research, a number of other studies
have investigated the relationship between patterns of
teacher influence and pupil achievement. The findings of
recent studies which parallel Flanders' study (11) in 1960
are consistent with his findings. However, some of this
research has been done in other subject matter areas and
grade levels.

Nelson (25) in 1966, in a study carried out at the
elementary school level, found that indirect teacher
influence was positively correlated to pupil achievement
on written language tests. She also found that direct
teacher influence patterns appeared to inhibit pupils'
development of written language skills.

LaShier (54) in 1966, obtained similar results
when working with student teachers who were teaching a
six-week unit in biological science. LaShier found that
pupils of indirect student teachers achieved more than
pupils of direct student teachers.

Two of the more significant of the recent studies
were completed at Temple University by Furst (13) in 1966,
and Soar (42) in 1967. Furst was perhaps the first person

to attempt the replication of a study by re-analyzing a set

19

of audio tapes that had been analyzed earlier in a study
using another observational system. In this study, she
used the tapes of a cognitive study of the classroom by
Bellack (6) in 1967, analyzing them in terms of the inter-
action analysis categories. In using interaction analysis,
she was able to identify certain relationships between
teacher patterns and student achievement. Furst found that
above-average student achievement was positively correlated
to indirect teacher influence. She also found that the
amount of student talk was positively correlated to student
achievement.

Soar (42) in 1966, in one of the largest studies
yet conducted on interaction analysis found that indirect
teaching produced greater growth in reading comprehension
in elementary school pupils than direct teaching. He found
that children who had been in classes taught by indirect
teachers also advanced an average of five and one-half
months in reading comprehension during the summer vacation;
while children who had been in direct teachers' classes
advanced three months in the same period. These results
further seem to support the fact that the influence of the
teacher on learning persists even after the formal class-
room experience is completed.

Weber (69) in 1967 studied pupils' creativity
levels after a three year experience with either an indirect

or direct teacher, and he found that indirect teaching

20

produced higher pupil creativity scores than direct teach-
ing. This study was conducted in a unique situation in
which children spent the first, second, and third grades
with the same teacher. The Torrance creativity test was
used to measure the children's growth in creativity.

Powell (60) in 1968, using a design similar to
Weber's found that children who had an indirect teacher
scored significantly higher on arithmetic tests than
children who had a direct teacher for the same period of
time.

Urbach (68) in 1966, studied the recurring patterns
of teaching using Flanders ten category interaction obser-
vation techniques did exist for each teacher.

Stilwell (66) in 1967 studied, developed and ana-
lyzed a category system for systematic observation of
teacher-pupil interaction during a geometry problem-solving
activity. He found that teacher talk consumes approximately
three times as much time as student talk, and also, that
this system is able to differentiate between teachers.

Samph (65) in 1968 studied observer effects on
teacher behavior. He used Flanders' system of interaction
analysis to record teacher verbal behaviors. He found that
teachers became more "indirect" when an observer was present
in their classroom, whether they were informed of an obser-

vation prior to its occurrence or not.

21

Kysilka (53) in 1969 studied the verbal teaching
behaviors of mathematics and social studies teachers in
eight and eleventh grades. The instrument she used was
OScAR,5V developed by Medley, et_gl, (36) in 1968. She
found that mathematics teachers talked significantly more
than social studies teachers (p<.05).

Geeslin (49) in 1971 reported on an investigation
of the relationship between characteristics of mathematics
teachers and student achievement. He found that there was
very little correlation in the teacher's characteristics,
and his effectiveness. He used the National Longitudinal
study of mathematical abilities data bank as a data source.
However, in using this data, there were more factors con—
sidered than basically those of the verbal characteristic.

Roland (63) in 1971, found that when teachers and
pupils had had training in classroom analysis, the teachers
would exhibit more indirect teaching behaviors and students
more participation in class than those classrooms with only
teachers taking the training.

Gabehart (48) in 1971 found that an intensive in-
service training program designed to develop a positive
social—emotional climate was partially successful using
Flanders system of Interaction Analysis.

Reed (61), in 1970 found that a four week verbal
and written feedback program was not effective in bringing

about measurable teacher flexibility change when measured

22

10 weeks after the program was completed. This measurement
was made by the students who had been trained to use
Flanders Interaction Analysis.

Relationship of Review of Literature
to the Present Study

 

An adequate measurement of teacher effectiveness
will certainly lead to more effective means of accountabil-
ity, more positive directions toward improvement of in-
struction, and consequently to explicit ways by which
students may learn more and better.

Keeping these ideas in mind, and considering the
fact that this study deals with the modification of teach-
er's verbal behavior and the achievement of high school
students in geometry, one could ask, "Can you modify a
teacher's behavior, and what effect does this have on
students achievement?", "Can systematic inquiry and experi-
mentation be used effectively in changing a teacher's
verbal behavior?", or "Can a teacher adequately produce
patterns of influence that will successfully increase his
students learning?" These questions have not been answered
by previous research. The research was primarily concerned
with the classification of teacher influence and how this
influence affected the achievement of students in the
classroom.

Most of the studies reviewed were helpful in carry-

ing out this study, because they contributed to the

23

conceptualization of teacher verbal behavior and the effects

of this behavior on the achievement of students.

Summary

These studies indicate that there seems to be some
relationship between the verbal behavior of the teacher
and the achievement of students in the classroom. The
literature was also reviewed in the area of analyzing the
teacher's verbal behavior using Flanders and other methods
of interaction analysis. Some conclusions from the liter-
ature which are considered important to this study are:

1. According to Anderson (1939), Withall (1949),
Mitzel and Rabinowitz (1953), Flanders (1960), and others,
the research on teacher effectiveness during the last
twenty years has lent itself toward a more scientific
approach to evaluation of teaching.

2. A vital factor affecting learning in the class-
room is talk--by the student or by the teacher. Amidon &
Hunter (3: p. 1,9), Biddle & Ellena (7: p. 129), Hughes
(14: p. 91), and Amidon & Giammatteo (14: p. 187).

3. There appears to be evidence that teachers
recognize the need for more student involvement in the
classroom as revealed by Amidon & Hunter (3: p. 2,103).

4. There appears to be a need for a more viable
objective measuring device than a rating scale if teachers

are to be held accountable.

24

5. Patterns of verbal behavior do exist for each
teacher according to the findings of Urbach (68) and
Moskowitz (14: p. 275).

6. Students tend to achieve more under teachers
who are more indirect in their approach; while tending to
achieve less when the approach is more direct. According
to: Nelson (25), LaShier (54), Soar (42), Weber (69),
Powell (60), and Amidon & Flanders (30).

7. Further, the review of the literature gives
much support to the fact that the extent to which the
teacher challenges students to support ideas, and the
amount of spontaneous student discussion may be related to
student gains, according to the findings of: Nelson (25),
Amidon & Flanders (11), Morsh and Burgess (24), Furst (13),
and Soar (42).

8. Students appear to achieve better under one
teacher than another teacher even if all other aspects of
the learning situation remain constant according to Amidon
and Flanders (la).

9. Interaction analysis has highlighted the
possibility of the importance of teacher influence of pupil
behavior, learning and attitudes by associating influence
with the teacher's verbal behavior.

10. Achievement appears to be a vital goal of

instruction.

CHAPTER III

DESIGN OF THE STUDY

Introduction

 

As chairman of the mathematics department in the
largest high school in Cleveland, Ohio, the writer was
interested in a more scientific approach in teacher evalua—
tion. Therefore, this study was designed to determine if
a teacher's verbal behavior could be changed toward a more
indirect influence; thus resulting in better achievement
scores on a geometry test of students in that school. The
teacher's verbal behavior was measured by observations of
teacher's classroom interaction by the department chairman
who was trained to use Flanders Interaction Analysis

technique.

The Population

 

The population for this study consists of
twelve instructional staff members (all geometry teachers),
teaching geometry in a large inner-city high school of
Cleveland, Ohio. All twelve of the teachers were used in
the study.

All 12 teachers were observed and characterized as

having direct or indirect influence. Three teachers were

25

26

classified as indirect and nine were classified as direct.
Then five teachers were selected at random from the nine
direct teachers and were provided with the treatment.

This was the final division of the twelve teachers
into three distinct groups:

Group I, the three teachers who were indirect
already; Group II, the four direct teachers who were not
selected for the treatment group; and Group III, the five
direct teachers who were selected at random from the

original nine direct, and provided the treatment.

Interaction Analysis

Flanders' interaction analysis was the instrument
used to categorize teacher verbal behavior. This system
is concerned with analyzing the influence pattern of the
teacher. The purpose is to record a series of acts in
terms of pre-determined concepts. The concepts in this
case refer to the teacher's control of the student's free-
dom of action. The system of categories are designed to
enable an observer, using the instrument, to distinguish
acts of the teacher that encourage the student's freedom
of action from those that restrict them. Both those acts
that encourage, and restrict the student's freedom are
recorded (11).

There are ten behavior categories designed in the

Flanders' Interaction analysis system. Seven of the

27

categories are assigned to teacher talk and two to student
talk. The tenth category covers pauses, short periods
of silence and talk that is confusing or noisy (Appendix C).

Of the seven categories assigned to teacher talk,
categories one through four represent indirect influence:

1. Acceptance of feeling

2. Praise or encouragement

3. Accepting ideas

4. Asking questions

Direct influence is categorized under five through
seven:

5. Lecture

6. Giving directions

7. Criticizing or justifying authority (1: p. 6-11)
The system of categories is designed for situations in
which the teachers and the students are actively discussing
(12: p. 21).

Teacher flexibility is a measure of of Indirect/
Direct verbal influence ratio. A measurement of teacher
influence can be calculated in either of two ways: The
I/d ratio consists of dividing the total of categories
1,2,3, and 4 by the totals of categories 5,6, and 7
(12: p. 35).

The second way is using the i/d ratio, which merely
excludes categories 4 and 5 and becomes categories l+2+3

divided by 6+7. By eliminating categories 4 and 5, lecture

28

and asking questions, it gives evidence as to the direct
or indirect approach a teacher uses for motivation and
control.

Only I/D ratio was used in making comparisons of
the mean teacher flexibility score of the three groups.
The I/D ratio is more inclusive because it makes use of

all teacher statements (12: p. 74).

Measures

During each observation, some non-verbal character-
istics of teachers and students were recorded to determine
if any relationships existed that would reflect any signi-
ficant difference in the calculation of the verbal behavior

(see Appendix B).

Methodology

 

The writer was the observer, and he learned obser-
vation and recording techniques of the Flanders' interaction
analysis by taking a programmed learning course of Educa-
tional Consulting Associates, Inc. (19). The writer spent
the summer of 1971 learning the technique. The training
design may also be found in Appendix C, concerning the
categories, their numbers, and meaning were given to each
teacher prior to the initial observations. This occurred
during the first week of school. During the second week

of the next two weeks, the initial observation was made on

29

all twelve geometry teachers within a three-day period.
These twelve observations were taped in order that reliable
characterizations could be made. These tapes were analyzed,
the results recorded on data matrices; then the computations
were made on I/D, i/d, teacher talk, student talk, and
silence and confusion. Four of the tapes were analyzed
again to determine the reliability and consistency of the
observational techniques. These twelve teachers were
characterized according to their I/D ratio--indirect if
I/D was greater than or equal to 1.00, and direct if I/D
was less than 1.00. This characterization resulted in
three teachers being classified as indirect, and nine being
classified as direct. A random sample of five direct
teachers were selected from the nine. This divided the
teachers into three groups: Group I, the three indirect
teachers (control); Group II, the four direct teachers who
were not selected (control); and Group III, the five
randomly selected direct teachers (experimental). The
teachers in Group III were provided the treatment. The
results of the first observations provided information for
the first treatment of Group III.

The treatments were designed for a thirty minute
conference of a forty minute period with each teacher of

Group III.

30

Treatment #1.

 

At the beginning of the conference, each teacher
was given a brief description of Flanders' Interaction
Analysis System (Appendix A-Z), a copy of their data matrix
with the results (Appendix D), and a conference sheet
(Appendix A-4). The conference followed the format of
conference sheet III (Appendix A-4). It began by discus—
sing the percent of teacher talk, student talk, percent of
silence and confusion, and the I/D ratio (interpreting
each score using Flanders' guide lines in (la: p. 31—65).
Also, questions that could be useful to the teacher in
interpreting his own results were discussed from the sheet
with emphasis being placed on the fact that the questions
represented average percentages, and they reflected current
practice--not the best or most desired practice.

The last part of the discussion delt with showing
the teachers how their I/D ratio was computed, and how they,
using the seven categories of teacher influence, could
increase (looking at their data matrices, and Appendix C-l)
their I/D ratio. Finally, the teachers were given a copy
of their initial observation on tape to play back in addi-
tion to keeping the results of the initial observation for
their own follow-up.

Treatment number one used the analysis of the

initial observation.

31

Treatment #2.

 

The beginning of the second set of conferences
began by answering questions that teachers had about the
last conference. Then each teacher was given a brief
description of Flanders' Interaction Analysis System
(Appendix A—2), a copy of their data matrix for observation
1 (Appendix D), and a conference sheet (Appendix A-4).

This sheet listed the results of the initial observation
and observation 1. Using the format of conference sheet
III (Appendix A-4), the observer discussed (comparing re-
sults of last two observations) teacher talk, student talk,
percent of silence and confusion, and the I/D ratio. The
scores of observation 1 were interpreted using Flanders'
guide lines in (la: p. 31-65). Also, questions that could
be useful to the teacher in interpreting his own results
were discussed from the sheet with emphasis being placed
on the results being average percentages, and reflecting
current practice--not the best or most desired practice.

The last part of the conference delt with discussing
how the I/D ratio of observation 1 related to the data
matrix of observation 1, and the seven teacher categories
of Appendix C-l. Also, how his I/D ratio could be increased.

Finally, teachers were given a copy of the results
of observation l--with comparisons to initial observa-

tion--for their own follow—up.

32

Treatment number two used the analysis of initial

observation, and observation 1.

Treatment #3.

 

The beginning of the third set of conferences began
by answering questions that teachers had from the results
of the last two conferences. Then each teacher was given
a brief description of Flanders' Interaction Analysis System
(Appendix A-2), a copy of their data matrix of observation
2 (Appendix D), and a conference sheet (Appendix A-4). This
sheet listed the results of the initial observation, obser-
vation 1, and observation 2. Using the format of conference
sheet III (Appendix A-4), the observer discussed (comparing
results of the last three observations) teacher talk,
student talk, percent of silence and confusion, and the I/D
ratio. The scores of observation 2 were interpreted using
Flanders' guide lines in (1a: p. 31-65). Also, questions
that could be useful to the teacher in interpreting his own
results were discussed from the sheet. Emphasis was, again,
placed on the results being average percentages, and re-
flecting current practice--not the best or most desired
practice.

The last part of the conference delt with discus-
sing how the I/D ratio of observation 2 related to the data
matrix of observation 2, and the seven teacher categories

of Appendix C-l. Also, how his I/D ratio could be increased.

33

Finally, teachers were given a c0py of the results
of observation 2--with comparisons to initial observation,
observation 1, and observation 2--for their own follow-up.

Treatment number three used the analysis of initial
observation, observation 1, and observation 2.

Observation 3 (final) was made after treatment
number three.

A paired t-test was used to measure significant

difference of the mean I/D ratios between the groups.

Statistical Hypotheses

The problem of assessing the verbal influence of
teachers characterized as having direct or indirect influ-
ence was approached through the testing of three basic
hypotheses. For the purpose of data analysis the hypotheses

are stated in the null form.

Hypothesis 1

 

H01 The mean I/D ratio of Group I, indirect at the
beginning, shall not be significantly different
from final I/D ratio of Group III.

Hon -M =0

1 3

Legend: M1 = Indirect teachers mean; M3 = Five

randomly selected teachers mean.

34

Hypothesis 2

 

Ho

2

The mean pre I/D ratio of Group III, five randomly
selected, shall not be significantly different
from the mean post I/D ratio of Group III.

HO: Pr - P0 = 0

Legend: Pr = Pre-I/D ratio mean; Po = Post—I/D

ratio mean.

Hypothesis 3

 

Ho

3

The mean I/D ratio of Group II, the four direct
teachers who were not selected, shall not be
significantly different from the final I/D ratio
of Group III.

HO: M2 - M3 = 0

Legend: M2 four direct teachers not chosen mean;

M

3 five direct teachers chosen mean.

Summary

This study was designed to determine if a teacher's

verbal behavior could be changed toward a more indirect

influence resulting in better achievement scores on a

geometry test of students in an inner city high school in

Cleveland, Ohio. Twelve geometry teachers of this school

were the subjects of this study. Teacher flexibility was

measured by I/D ratio of the Flanders' Interaction Analysis

instrument.

35

Teachers were characterized by their I/D ratios
and placed in three groups. Group I, three teachers who
were indirect at the beginning; Group II, four teachers
who were direct at the beginning, but not selected to
provide treatment; Group III, five direct teachers selected
at random from the nine direct teachers, the experimental
group to be studied over a five month period.

The observer was trained to use Flanders' Inter-
action Analysis by a programmed learning approach. Each
of the five teachers selected to study were observed three
times (besides the initial observation) with a conference
preceding each observation.

The group's mean I/D ratio were compared in a
paired t-test to determine the significant difference, if
any, in the change of teacher verbal behavior resulting

from the treatment.

CHAPTER IV

ANALYSIS OF DATA

The problem of assessing the verbal influence of
teachers characterized as having direct or indirect influ-
ence was approached by means of three basic hypotheses.

For the purpose of data analysis, all hypotheses were stated
in the null form. The initial characterization of the

twelve teachers is found in Table 4.1.

Table 4.1 Characteristics of the twelve teachers after
initial observation.

 

 

 

Teacher I/D i/d Percent Percent
Teacher-talk Student-talk
A .59 16.02 72 16
B .50 12.25 69 18
C .49 14.12 68 15
D .69 15.13 71 20
E 2.43 6.35 65 30
F .61 12.02 69 19
G .61 13.82 71 14
H 2.17 4.18 61 29
I .59 10.04 71 17
J .48 15.26 60 12
K .63 11.00 68 16
L 2.89 5.91 70 24

36

37

After the twelve teachers were characterized and
a random sample of five from the nine direct teachers were
made, the groups (for the purpose of group I/D comparison)

were as follows:

 

 

 

 

 

Group I Group II Group III
I/D Ratio I/D Ratio I/D Ratio
Sl 2.43 S4 .79 88 .59
82 2.17 35 .91 $9 .50
S3 2.89 S6 .69 S10 .49
s7 .48 S11 .51
$12 .63

 

Observe that each of the teachers in Group III
(below) increased their I/D ratio over time except teachers
S11 and $12 in the last observation. Sll's failure to be
consistent was probably due to an unexpected ringing of the
fire be11--the class never regained .its composure, and
teacher 812 had a discipline problem which upset the class
very much.

The I/D ratios for Group III for the three

observations after each of the three treatments were as

 

 

follows:
Observations
01 02 03
$8 .57 1.14 1.39
89 .64 .69 1.01
S10 .50 .58 2.39
Sll 1.04 1.19 .76
S12 .56 1.03 .67

 

 

 

 

 

 

38

To test the validity of the characterization of
Group I, II, and III at the beginning, the t-test at
a = .05 was applied, and found that the mean of Group I
was significantly higher than the mean of either Group II
or Group III. It was also found that there existed no
significant difference in the means of Group II or Group
III at the beginning.

The test statistic was the paired t-test:

 

 

 

—-— A A2
X1 X2 2 (“1'1)512+ (n2-l)82
t = where Sp = _
/Sp2(% + % ) nl+n2 2
1 2
Decision Rule: At a = .05 and n1 + n2 - 2 degrees of
freedom, reject HO 1f t 2 t(a,n +n _2)

l 2

or if t S -t
(a, nl+n2-2)

Hypothesis 1

H01 The mean I/D ratio of Group I, indirect at the be-
ginning, shall not be significantly different from

the mean of the final I/D ratio of Group III.

The mean I/D score of Group I is 2.50 and the final
mean I/D score of Group III is 1.23. The computed t value
of this analysis is 2.442. This value is less than the
tabular value of t with 6 degrees of freedom (2.447),

a = .05. The null hypothesis is not rejected on the basis

of this evidence. There exist no measurable significant

39

difference in the mean I/D scores between Group I and
Group III. The summary data for the paired t-test is found

in Table 4.2.

Table 4.2 Summary Data for the Paired t-test.

 

 

Group Mean
I 2.50
III 1.23

 

 

 

 

Hypothesis 2

 

H02 The mean post I/D ratio of Group III, five randomly

selected, shall not be significantly different

from the pre I/D ratio of Group III.

The mean pre I/D score of Group III is .56, and
the mean post I/D score of Group III is 1.23. The computed
t value of this analysis is 2.61. This value is greater
than the tabular value of t with 8 degrees of freedom
(2.306), a = .05. The null hypothesis is rejected on the
basis of this evidence. There exist a measurable signifi-
cant difference between the mean post I/D scores of Group
III, and the mean pre I/D scores of Group III. The summary

data for the paired t-test is found in Table 4.3

Table 4.3 Summary Data for paired t-test.

 

Group III Mean

 

Pre .56
Post 1.23

 

 

 

 

40

Hypothesis 3

 

HO3 The final mean I/D ratio of Group III, randomly
selected, shall not be significantly different from
the mean I/D ratio of Group II, the four direct
teachers who were not selected.

The final mean I/D score of Group III is 1.23, and
the mean I/D score of Group II is .59. The computed t
value of this analysis is 1.98. This value is less than
the tabular value of t with 7 degrees of freedom (2.365),

a = .05. The null hypothesis is not rejected on the basis

 

of this evidence. There exist no measurable significant
difference between the final mean I/D score of Group III,
and the mean I/D score of Group II. The summary data for

the paired t-test is found in Table 4.4.

Table 4.4 Summary Data for paired t-test.

 

 

Group Mean
II .59
III 1.23

 

 

 

 

Sub-Hypothesis I

 

Sub HOlThe overall mean I/D score of Group III shall not
be significantly different from the mean I/D score
of Group I.
The overall mean I/D score of Group III is .94, and
the mean I/D score of Group I is 2.50. The computed t value

of this analysis is 10.4. This value is greater than the

41

tabular value of t with 6 degrees of freedom (2.447),

a .05. The null hypothesis is rejected on the basis of
this evidence. There exist a measurable significant
difference between the overall mean I/D score of Group III,

and the mean I/D score of Group I.

Sub-Hypothesis 2

 

Sub H02 The overall mean I/D score of Group III shall not
be significantly different from the mean I/D score

of Group II.

The overall mean I/D score of Group III is .94, and
the mean I/D score of Group II is .59. The computed t value
of this analysis is .41. This value is less than the tab-
ular value of t with 7 degrees of freedom (2.365), a = .05.
The null hypothesis is not rejected on the basis of this
evidence. There exist no measurable significant difference
between the overall mean I/D score of Group III, and the
mean I/D score of Group II. The summary data for sub

hypotheses 1 and 2 is found in Table 4.5.

Table 4.5 Summary data for paired t-test

 

 

Group Mean
I 2.50
II .52

III .94

 

 

 

 

42

Additional Comparisons

 

The percentage of teacher-talk was recorded and

analyzed. The results of the teacher-talk at the beginning

was:

 

 

Group Mean
I 65.3

II 67.8
III 69.6

 

 

 

The mean teacher—talk of all the teachers at the beginning
was 67.6. The mean teacher-talk of Group III for the final
observation was 67.6. There was no measurable significant
difference between the means of teacher-talk. For purposes
of percent of teacher-talk comparisons, the data for the
groups were as follows: (Table 4.6, Table 4.7, and Table

4.8).

Table 4.6 Comparison of percent of teacher-
talk.

 

 

Pre-Percent of teacher-talk

 

Group I Group II Group III
S1 65 S4 71 ' 88 72
82 61 S5 69 89 69
S3 70 86 71 S10 68
S7 60 S11 71

S 68

 

43

Table 4.7 Post-Percent of teacher-talk
for Group III.

 

 

 

Subjects Percent of Teacher-Talk
SB . 70
$9 58
S10 71
S11 71
$12 68

 

Table 4.8 Summary data for mean percent
of teacher-talk

 

 

 

 

Group Mean
Pre Post
I 65.3
II 67.8
III 69.6 67.6

 

Recall that the results of the three observations
of each of the five teachers in Group III appeared on page
31. The summary data for comparing the pre I/D scores, and
post I/D scores of each teacher in Group III are found in
Table 4.9. The summary data for comparing the pre and post
percent of teacher-talk of each teacher in Group III are

found in Table 4.10.

44

 

 

 

 

 

 

 

Table 4.9 Summary of pre and post I/D
scores of Group III.
Subjects Pre Post
88 .59 1.39
89 .50 1.01
S10 .49 2.39
811 .61 .76
812 .63 .67
Table 4.10 Summary of pre and post
percent of teacher-talk
of Group III.
Subjects Pre Post
S8 72 70
89 69 58
S10 68 71
S11 71 71
812 68 68

 

The summary data for comparing the pre and post

percent of student-talk in each of the classes of the

teachers of Group III are found in Table 4.11.

The mean

student-talk of students of teachers in Group III at the

beginning was 15.8; while the mean student-talk at the end

was 21.4.

There was a measurable significant difference

between the pre and post means of student-talk of the

teachers of Group III.

45

Table 4.11 Summary of pre and post percent
of student-talk in each class
of teachers of Group III.

A

 

 

Subjects pre post
S8 16 22
$9 18 21
S10 15 22
S11 14 18
812 16 24

 

Since interaction analysis is a specialized research
procedure which provides information about a few of the
many aspects of teaching, some non-verbal characteristics
were recorded during each observation (see Appendix B).
These characteristics would not be revealed by the verbal
communication, and the investigator feels that they may be
important to the total complex called teaching. A composite

of the results is recorded below:

Non-Verbal Behavior

 

Teachers were teaching basically the same kind of
material during the observations:
Observation, Initial The Introduction to polygons

Observation #1 Relationships of polygons
and Drawing conclusions

Observation #2 Proofs

Observation #3 Proving triangles congruent

46

Class Formation Taught as a unit
Preparation of Teachers Generally Good
Content Presentation Generally Good

Other Behavior

 

Teacher Student
Stands Behind desk, Answered most
at chalkboard of time Right
Moves about Most of time Looks around
(unattentively) Some
Looks at
Plans Sometime Ignores
Instructor Very Little
Smiles or
Laughs Most of time Slumps Very Little

Demonstrates At chalkboard Yawns, stretches Very Little
Uses Chalk- To illustrate Sleeps or dozes Very Little

board key ideas
Group answers Some

Cases For Group III

 

The following is a discussion of the cases of each
teacher in Group III throughout the study:

Teacher 88 became more indirect with an increase

 

of .80 in I/D ratio. Teacher-talk was decreased by 2%;
while student-talk increased 6%. She was very interested
and enthusiastic about the study. She could not wait to
have the Opportunity to discuss her observations in the
conference and see her results. She studied the Flanders'
information thoroughly. She worked hard at involving her

students more. The observer felt that this teacher really

47

wanted to have her students learn more. She was very
dynamic and constantly used techniques to get her students
to talk and become involved with the class discussion.

Her students' average score on the achievement test was
the same as the mean.

Teacher 89 became more indirect with an increase

 

of .51 in I/D ratio. Teacher-talk decreased by 11%; while
student-talk increased by 3%. He had heard about Flanders'
method, and thought he knew it. The flow in his class was
kind of slow, and it took students longer to react to his
questions, but after they became involved, it was hard to
stop them. He was always on the defensive in his confer-
ences. He could not seem to accept some of the results.
This was due to the fact that he was trying to work in the
frame work of F1anders--and thought he was doing a good
job. He would come back invariably to check something.

He had several discipline problems that got the class out
of order on several occasions, but he recovered quite well.
His students' average score on the achievement test was
the same as the mean.

Teacher Slo became more indirect with an increase

 

of 1.90 in I/D ratio. Teacher-talk increased by 3%; while
student-talk increased by 7%. There was a very small per-
centage of silence and confusion. He was more responsive
to the conferences and was concerned with specific things

that he could do to involve his students more. This

48

teacher checked his tape with his students--they discussed
it afterwards. He worked hard at becoming more indirect
in his influence. He was very cooperative--much better
than the others. His students appeared ready to get
involved. It was evident that he had gotten his students
worked up. They scored as well on the achievement test

as any class in this group. His students' average score
on the achievement test was 2 above the mean.

Teacher S11 became more indirect with an increase

 

of .15 in I/D ratio. Teacher-talk remained the same; while
student-talk increased by 4%. She was very responsive to
the study and was very concerned about the results. She
was more indirect on her first two observations involving
the treatment than any of the other teachers in Group III
(1.04 and 1.19 respectively). The result of the last ob-
servation does not justify her hard work, interest and
enthusiasm. Someone rang the fire gong on the last obser-
vation and it took a lot out of her, beacuse she wanted so
bad to have a good performance. She wanted me to nullify
the last and come again--I did not. Her students' average
score on the achievement test was 2 points above the mean.
Teacher S became more indirect with an increase

12

of .04 in I/D ratio. Teacher-talk remained the same; while

 

student-talk increased by 8%. There was less percentage
of silence and confusion from pre to post. He was very

interested in the results of the observations. However,

49

the observer did not feel he worked very hard at getting
more involvement from his students - he believed he was a
good teacher. He talked quite a bit and many times students
questions could shake him up. He had a little something

in his voice that at times students questions and answers
seemed to have a personal attack toward him. There was a
large percentage of silence and confusion in his class at
times, but he was able to eliminate some in the last obser-
vation and increase student involvement. His low score

on the last observation probably attributed to the disci-
pline problem he had, that broke the class up, and it

never came back together that period. His students'
average score on the achievement test was 3 points above
the mean.

Each of the five teachers in Group III became more
indirect in their influence pattern from pre to post--some
more than others. In cases where the I/D ratio remained
the same, the change in talk is reflected in a change in
the amount of silence and confusion. Each of the five
teachers talked more than the average teacher (61% of the

time).

Achievement

 

The investigator was more interested in modifying
teacher behavior than how this behavior attributed to

changes in students. However, he will discuss the (more

50

or less) inconsequential aspects of the students' achieve—
ment scores in Geometry.

At the end of the course, all geometry students
were administered a standardized test--The Cooperative
Mathematics Test in Geometry, Form A. These tests were
scored by a computer, and the results recorded. The mean
of all geometry students in the school was 135. The
summary data for comparing the mean scores of the classes
in each group are found in Table 4.12.

Table 4.12 Summary of Mean Achievement Scores in each
class in each group.

 

 

Student Mean Scores

 

 

 

 

Group I Group II Group III
S1 131 S4 131 $8 135
s2 134 35 133 39 135
S3 138 S6 134 S10 137
S7 135 S11 137
$12 138

 

 

The mean achievement score of the students of teachers in
Group I is 134.3; this is .7 below the overall mean. The
mean achievement score of students of teachers in Group II
is 133.3; this is 1.7 below the overall mean. The mean

achievement score of students of teachers in Group III is

51

136.4; this is 1.4 higher than the overall mean. The
summary data for comparing the means of each group with

the overall mean is found in Table 4.13.

Table 4.13 Summary of Means of each group
and overall mean.

 

 

 

Group Mean Student Achievement
I 134.3
II 133.3
III 136.4
Overall 135.0

 

Because of the consistent relationship that appears
to exist between patterns of classroom interaction and
pupil achievement, an interest has developed in its use
as a tool in teacher education. The results of this study
do indicate that the teacher's behavior in Group III was
modified, and the results of the achievement scores indi-
cate a higher achievement in Group III than in either
Group I or Group II. However, since no pre-test was admin—
istered or any adjustments made on initial ability, the
investigator is unable to statistically or experimentally
determine what these differences are attributed to--nor
can the determination of the significance of these differ-
ences be made. The indication of this study that the
teacher's behavior was modified gives support to the find-

ings of Gunnison's experiment in 1968 (50).

52

The investigator believes that studies of natural
teaching do suggest promising variables, but it is only
through careful controlled experimental studies that we
can determine which teaching variables (if any) enhance

student achievement.

Summary

The hypotheses of this chapter were analyzed by
using the paired t-test. The first two hypotheses were
divided into subhypotheses, each testing the overall I/D
score of Group III with the pre I/D score of Group I and

II.

HCl The mean I/D ratio of Group I,
indirect at the beginning, shall (Failed to Reject)
not be significantly different
from the final I/D ratio of Group
III.
H The mean post I/D ratio of Group
III, five randomly selected, (Rejected)
shall not be significantly
different from the pre I/D ratio
of Group III.
H The final mean I/D ratio of Group
III, randomly selected, shall not (Failed to Reject)

be significantly different from

the mean I/D ratio of Group II,

53

the four direct teachers who
were not selected.
Sub HCl The overall mean I/D score
of Group III shall not be (Rejected)
significantly different from
the mean I/D score of Group I.
Sub H0 The overall mean I/D score
of Group III shall not be (Failed to Reject)
significantly different from
the mean I/D score of Group II.

Some additional comparisons were made, and it was
found that there was no significant difference in the
percent of teacher-talk of the twelve teachers in the
beginning and the final observations of Group III. How-
ever, there was less teacher-talk among teachers of Group
III at the end than there was at the beginning. Also, the
students participated more in the classes of the teachers
of Group III at the end than at the beginning.

The average achievement score was higher in Group
III than in either Group I or Group II, but since no pre-
test was given no statistical inferences could be made

about the differences.

CHAPTER V

SUMMARY, FINDINGS, CONCLUSIONS,
AND RECOMMENDATIONS

Introduction

 

This chapter draws from the preceding chapters
the central elements of the study. A summary of the study
will be presented along with the findings based upon the
data obtained. Some general conclusions are noted along

with recommendations for further study.

Summary

It was the purpose of this study to determine if
a teacher's verbal behavior could be changed toward a more
indirect influence resulting in better achievement scores
on geometry tests of students in a large inner-city high
school in Cleveland, Ohio. The study used a case study
approach, and Flanders' method of interaction analysis to
analyze the teachers verbal behavior.

The review of the literature revealed that students
taught by teachers who use an indirect influence in their
classrooms, achieved more; while those in classrooms of

direct influence achieved less (10).

54

55

All of the geometry teachers in the school were
involved in the study. They were divided into two groups
at the start by using Flanders' method of interaction
analysis (11). This division resulted into three teachers
with indirect influence, and nine teachers with direct
influence. This was done on the basis of the teacher's
I/D ratio score, i.e., if a teachers' I/D ratio was less
than 1, they were characterized as direct, and if the I/D
ratio was greater than or equal to 1, they were character-
ized as indirect. Five direct teachers were selected from
the nine direct teachers to form three final groups. Group
I, three indirect teachers; Group II, four direct teachers
not selected for treatment; and Group III, five direct
teachers who would receive the treatment. The treatment
involved a conference about Flanders' method of inter-
action analysis, with the idea of trying to direct the five
direct teachers toward a more indirect influence.

The achievement of the students was measured by
COOperative Mathematics Tests, Form A, Geometry of the
Educational Testing Service (see Appendix E).

The results were analyzed by using a paired t-test

with an alpha level of .05.

Findings
Within the parameters of this study, the following

findings are presented.

56

1. On a continuum from .56 to 2.50 the five direct
teachers moved up so that their I/D ratios did not differ
significantly with those of the indirect teachers.

2. The five direct teachers became more indirect
in their influence pattern.

3. The five direct teachers failed to move up the
continuum so that there was a measurable significant differ-
ence between them and the four direct teachers who did not
receive the treatment.

4. The five teachers in Group III became more
indirect with time, but failed to reach the point of in-
directness of Group I, those teachers who were indirect at
the beginning.

5. The mean percent of teacher talk was 67.6.

6. The mean percent of student talk was 19.1.

On the basis of the findings of this study and
within its limitations, the following conclusions seem
justified.

1. Teachers can be directed toward a more indirect
influence pattern of teaching.

2. Direct teachers can be made more indirect using
Flanders' method of Interaction analysis.

3. Students appeared to achieve more in teachers'
classes who possess an indirect influence pattern of teaching.

4. Teacher's behavior in the classroom can be

affected by using Flanders' method of interaction analysis.

57

Discussion

 

In this study, the observational evidence that was
obtained does indicate that the teachers' behavior was
modified in the experimentally desired direction. However,
the hypotheses suggest that maybe a longer period should
be used across time, with more observations being made, if
there is to be a more appreciable change in teacher behavior
toward indirectness.

Samph found in his study (65), that teachers became
more indirect when an observer was present in their class-
room, whether they were informed of an observation prior
to its occurrence or not. This may be true, but the
writer believes that by also training teachers in the use
of Flanders' method of interaction, one can eliminate much
of the observer effect.

The findings that the mean percent of student-talk
was 19.1 is in accord with Stilwell's findings (66),
teacher-talk consumes approximately three times as much
time as student-talk.

The achievement scores in the teachers' geometry
classes who became more indirect in their influence (Group
III) was higher than either of the other groups, but since
no pre-test was given, no statistical inference can be

made about the differences.

58

Recommendations for Future Research

The following recommendations are a direct result
of the findings of this study and as a result of working
as a department chairman with all of the teachers involved
in this study.

1. One can learn interaction analysis through a
program approach well enough to implement it in the class-
room.

2. The observations should be spread over a
longer interval of time and with more frequency.

3. This study should be replicated with certain
modifications using the idea of #2, and by considering
some non-verbal characteristics such as the form in Appen-
dix B.

4. Other subjects different from geometry should
be used following this basic format.

5. Since the primary purpose is to improve in-
struction, one could give the teachers involved a thorough

training in interaction analysis before beginning the study.

Questions for Further Study

 

This study seems to answer a few of the questions
prOposed in Chapter II i.e. 4, can Flanders' ten category
method of interaction analysis be used successfully in the

classroom as a scientific approach to measuring teacher

59

effectiveness?; 5, can the level of instruction be raised
through the use of interaction analysis?; and 6, can a
teacher's behavior be affected by a conference? Also, this
study and previous research seem to suggest that the answer
to #1 is yes.

Therefore the two important questions for further
study are:

1. Does the verbal behavior of teachers observed
by Flanders' method of interaction analysis lend itself
to any typical consistency?

2. After determining the relationship between the
teacher and student behavior, how does this attribute to
the amount of learning, or what manner would the teacher

be graded by his supervisors to determine his accounta-

bility?

BIBLIOGRAPHY

1a.

BIBLIOGRAPHY

A. Books

Amidon, E. and Flanders, N., The Role of the Teacher in
the Classroom, Minneapolis: Association for
Productive Teaching, Inc.) 1967.

 

 

Amidon, E. and Flanders, N., "Measuring Dependence
Proneness in the Classroom." Bureau of Educa-
tional Research, No. Ber 60-66, University of
Minnesota, Minneapolis.

 

Amidon and Hunter, "Interaction Analysis: Recent
Developments," In Amidon—Hough eds., Interaction
Analysis: Theory, Research, and Applications,
Massachusetts: Addison-Wesley, 1962, pp. 388-
392.

 

Amidon and Hunter, Im roving Teaching: The Analysis of
Classroom Verpal Interaction, Holt, Rinehart,
and Winston, Inc., New York, 1967.

 

 

Anderson, H. H. et a1., "Studies of Teachers' Classroom
Personalities, III," Applied Psychology Mono-
graph No. 11, Stanford, Ca1if.: Stanford Uni-
versity Press, 1946, 156 pp.

 

Anderson, H. H., 1939, "The Measurement of Domination
and of Socially Integrative Behavior in
Teachers' Contacts with Children." In Amidon-
Hough eds., Interaction Analysis: Theory, Re-
search, and Applications, Massachusetts: AdIEon-
Wesley, 1967, pp. 4-23.

 

 

Bellack, A. and others, TheyLanguage of thg Classroom,
New York: Bureau of’Pfiblications, Teachers
College, Columbia Univ., 1967.

 

Biddle, B. and Ellena, W., Contemporary Research on
Teacher Effectiveness, New York: Holt, Rinehart,
and Winston, Inc., 1964.

 

6O

10.

ll.

12.

l3.

14.

15.

16.

17.

61

Fey, James T. "Classroom Teaching of Mathematics,"
Studies in Mathematics, Vol. XIX, Review of Re-
cent Research in Mathematics Education, Stanford:
School Mathematics Study Group, I969a.

 

 

 

Fey, James T. "Research in Classroom Teaching of Math—
ematics," Review of Educational Research, 39,4
(October, 1969b), 535—551.

 

Flanders, N. A., Analyzing Teacher Behavior, 1970,
Addison-Wesley, Reading, Mass.

 

Flanders, Ned A., Interaction Analysis in the Classroom:
A Manual for Observers, Minneapolis: College of
Education, UniVersity of Minnesota, 1960a, 36 pp.

 

Flanders, N. A., "Teacher Influence in the Classroom,"
In Amidon-Hough, eds. Interaction Analysis:
Theory, Research, apdApplicgtion, Massachusetts:
Addison-Wesley, 1967, pp. 103-115.

 

 

Furst, Norma, "The Effects of Training in Interaction
Analysis on the Behavior of Student Teachers in
Secondary Schools," In Amidon-Hough, eds. Inter-
action Analysis: Theory, Research, and Applica-
tion.

 

Gage, N. S. (ed), Handbook of Research on Teaching,
Chicago: RandiMcNally, 1963.

 

Good, Carter V. "The Individual and Case Study: Diag—
nosis and Therapy," Essentials of Education Re-
search, New York: Appleton-Century-Crofts, 1966,
310-344.

 

Hough, John B., "An Observation System for the Analysis
of Classroom Instruction," Amidon and Hough, eds.
Interaction Analysis: Theory, Research, and Ap-
plications. Massachusetts: Addison-Wesley, 1967.
pp. 150-157.

 

 

Hough, J. and Ober, R., "The Effect of Training in In-
teraction Analysis on the Verbal Teaching Behav-
ior of Pre-Service Teachers," Amidon-Hough, eds.,
Interaction Analysis: Theory, Research, and A -

liCations, Massachusetts: Addison-Wesley, 1967,
329-345. ~

 

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

62

Lewin, K., Lippitt, R., White, R., "Patterns of Agres-
sive Behavior in Created Social Climates," In
Amidoanough, eds., Interaction Analysis: Theory
Research, and Applications, Massachusetts: iii
Addison-Wesley, 1967, pp. 24-46.

 

 

 

 

Olson, Miles C. Learning Interaction Anal sis: A Pro-
grammed Approach, Englewood, CoIo., Educational
Consulting Associates, Inc., 1970.

 

Remmers, H. H. "Rating Methods in Research on Teaching,"
In Gage, N. L. (ed.), Handbook of Research on
Teaching, Rand McNally, 1963, pp. 329-378.

 

Whitla, Dean K., Handbook of Measurement and Assessment
in Behavioral Soiences, Reading, Mass., Addison-
Wesley Publ. Co., 1968.

 

Mitzel, H. E. and Rabinowitz, W., "Assessing Social-
Emotional Climate in the Classroom by Withall's
technique," Psychology Monogram, 1953, 67, No.
18.

Morsh, Joseph E., Systematic Observation of Instructor
Behavior, Report AFPTRC—TN-56-52, Lackland Air
Force Base, Texas, 1956.

 

 

Morsh, Joseph E., Burgess, George G., Student Achieve-
ment as a Measure of Instructor Effectiveness,
Report No. 7950)iTask 77243, LaCkIand Air Force
Base, Texas, 1954.

 

 

Nelson, Lois., "Teacher Leadership, An Empirical Ap-
proach to Analyzing Teacher Behavior in the
Classroom," Classroom Interaction Newsletter,
2:31-32, November, 1966.

 

Flanders, Ned A. "Some Relationships Between Teacher
Influence Pupil Attitudes, and Achievement," A
Chapter submitted to the AASA, the NSBA, and the
NBA-Classroom Teachers Division for incorporation
into a book titled The Evaluation of Teaching.

 

Flanders, Ned A., Teacher Influence, Pupil Attitudes and
Achievement: Studies in Interaction Analysis,
Final Report, Cooperative Research Project No.
397, U.S. Office of Education, 1960b, p. vi to
121.

63

28. Simon, Anita and Boyer, E.G. (eds.) Mirrors for Behav-
ior: An Antholo of Classroom Observation Instru—
ments, Vol. IV. i21 S. Broad St., Philadelphia,
Penn., Research for Better Schools, Inc., 1967.

 

 

 

29. Torrance, E. P., and Parent, E., Characteristics of
Mathematics Teachersphat Affect Students' Learning.
COOperative Research Project No. I020, Minnesota
School Mathematics and Science Center, Institute
of Technology, University of Minnesota, 1966.

(ERIC: ED 010 378).

 

C. Periodicals

 

30. Amidon, E. and Flanders, N. "The Effects of Direct and
Indirect Teacher Influence on Dependent-Prone Stu-

dents Learning Geometry," Journal of Educational
Psychology, Vol. 52, No. 6:286-291, 1961.

 

31. Barr, A. S. "Measurement of Teaching Ability," Journal
of Educational Research, Vol. XXVIII (April, 1935),
561-569.

32. Bolton, Floyd B. "Evaluating Teaching Effectiveness
Through the Use of Scores on Achievement Tests,"
Journal of Educational Research, Vol. 38, 691-696.

33. Cogan, Morris L. "Research on the Behavior of Teachers:
A New Phase," Journal of Teacher Education, 14:

34. Hughes, Marie M. "Teaching is Interaction," The Ele-
mentary School Journal, Vol. LVIII (May, I958) 457-
464.

 

35. Jayne, C. D. "A Study of the Relationship Between
Teaching Procedures and Educational Outcomes,"
Journal of Experimental Education, Vol. XIV (1945),
101-134.

36. Medley, Donald M., Mitzel, Harold E. "A Technique for
Measuring Classroom Behavior," Journal of Educa-
tional Psychology: Vol. 49, No. 2, 1958, 86-93.

 

 

37. Perkins, Hugh V. "A Procedure for assessing the class-
room behavior of students and teachers," American
Educational Research Journal, Vol. 1, No. 4, 249-
260, November, 1964.

 

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

64

Pickett, Laurel A. "Can the level of instruction be
raised through the use of interaction analysis?,"
Educational Leadership, 597-600.

 

Rolfe, J. F. "The Measurement of Teaching Ability,
Study Number Two," Journal of Egperimental Educa-
tion, V01. 14' 52-740

Rosenshine, Barak. "The Stability of Teacher Effects
Upon Student Achievement," Review of Educational
Research, Vol. 40, No. 5, (December,’i970), p. 647.

Silberman, Charles B. "You May Not Agree, but...even
Student Teaching is Dismal," Todayls Education,
January 1971, 22-63.

 

Soar, R. S. "Optimum Teacher—Pupil Interaction for
Pupil Growth," Educational Leadership, Vol. 26,
1968, 275—280.

 

Turner, R. L. "Pupil Influence on Teacher Behavior,"
Classroom Interaction News Letter, 1967, 3 (No. 1),
5-8. Available from Anita Simon (ed.), Research
for Better Schools, Inc., Suite 1700/1700 Market
St., Philadelphia, Penn., 19103.

 

Withall, J. "Development of a technique for the mea-
surement of socio-emotional climate in classrooms,"
Journal of Experimental Education, 1949, 17, 347-
361.

 

D. Unpublished Materials

 

Daniels, John Wm., "Effects of Interaction Analysis
Upon Teaching Assistants and Student Achievement
In College Mathematics," Dissertations Abstracts,
A31:2768, 1970.

 

Dougherty, J. Dwight, "The Function of the Mathematics
Department Head in a Large Urban High School."
Unpublished doctoral dissertation, New York Univ.,
1954.

Eicher, Charles E., "The Effects of Instruction in In-
teraction Analysis of the Verbal Teaching Behavior
of Elementary Mathematics Methods Students," Disser-
tation Abstracts, A31:262, 1970.

 

48.

49.

50.

51.

52.

53.

54.

55.

56.

65

Gabehart, David P., "An Investigation of the Effects
of an In-Service Training Program for Elementary
Teachers upon Selected Teacher-Pupil Perceptions
and Relationships." Unpublished Doctoral disser-
tation, West Virginia Univ., 1971.

Geeslin, Wm. B., "An Investigation of the Relationship
Between Characteristics of Mathematics Teachers
and Student Achievement." Paper presented at the
annual meeting of the National Council of Teachers
of Mathematics, Anaheim, Calif., April, 1971.
SMSG, Stanford University.

Gunnison, John P., "An Experiment to Determine the Ef—
fects of Changing Teacher Classroom Behavior Through
Training of Student Teachers in the use of the
Flanders Interaction Analysis System." Unpublished
Ph.D. dissertation, 1968.

Hill, Wm. M., "The Effect on Verbal Teaching Behavior
of Learning Interaction Analysis as an In-Service
Education Activity." Unpublished doctoral disser-
tation, Ohio State Univ., 1966.

Johns, Joseph P., "The Relationship Between Teacher Be-
haviors and the Incidence of Thought-Provoking
Questions by Students in Secondary Schools." Un-
published doctoral dissertation, Univ. of Michigan,
1966.

Kysilko, Marcella L., "The verbal Teaching Behaviors
of Mathematics and Social Studies Teachers in Eighth
and Eleventh Grades." Unpublished Ph.D. disserta-
tion, The University of Texas, 1969.

LaShier, W. 8., Jr., "The Use of Interaction Analysis
in BSCS Laboratory Black Classrooms." Paper read
at the National Science Teachers Association meet-
ings, New York City, April 3, 1966.

Limberg, Ronald K., "Relationships Between Indirect-
Direct Teaching, Teacher characteristics and School
Climate." Unpublished Ph.D. dissertation, Michigan
State University, 1971.

McCall, Kendrick M., "Modification of Teacher Behavior
and Self-Perception: Effects of Dissonance Coded
Feedback." Unpublished Ph.D. dissertation, The
State University of Rutgers, 1969.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

66

Nelson, Lois Ney, "The Effect of Classroom Interaction
on Pupil Linguistic Performance," Dissertation Ab-
stracts, 25:789, 1964.

 

Pankratz, Roger, "Verbal Interaction Patterns in Class-
room of Selected Physics Teachers." In Amidon,
Edmund J., and Hough, John B., (eds.), Interaction
Analysis: Theory, Research and Applications.
Addison, 1967, 402 pp.

 

Parrish, H. Wayne, "A Study of the Effects of In-Service
Training in Interaction Analysis on the Verbal Be-
havior of Experienced Teachers." Unpublished Ph.D.
dissertation, University of Oregon, 1968.

Powell, E. R., "Some Relationships between Classroom
Process and Pupil Achievement in the Elementary
School." Unpublished doctoral thesis, Temple Uni-
versity, 1968.

Reed, Dick, "The Effect of Pupil-Teacher Feedback on
Teacher Flexibility As Perceived by Students."
Doctoral dissertation, Michigan State University,
1970.

Rogers, Wm. J., "A case study of the North Dearborn
Heights Teacher's Strike." Unpublished Ph.D.
dissertation, Michigan State University, 1968.

Roland, Mark V., "Modification of Teacher and Pupil
Classroom Behavior Through Training with an Instru-
ment For Classroom Analysis Indicators of Quality."
Unpublished Doctoral disseration, Columbia Univ.,
1971.

Rosenshine, B., "Teaching Behaviors Related to Pupil
Achievement." Paper presented at the annual meet-
ing of the National Council for the Social Studies,
Washington, D.C., 1968. School of Education,
Temple University.

Samph, Thomas, "Observer Effects on Teacher Behavior."
Unpublished Ph.D. dissertation, The University of
Michigan, 1968.

Stilwell, Merle E., "The Development and Analysis of a
Category System for Systematic Observation of
Teacher-Pupil Interaction During Geometry Problem
Solving Activity." Unpublished Ph.D. dissertation,
Cornell University, 1967.

67

67. Storlie, T. R., "Selected Characteristics of Teachers
Whose Verbal Behavior is Influenced by an In-
Service Course in I. A." Doctoral dissertation,
Univ. of Minnesota, 1961, 209 pp.

68. Urbach, Floyd D., "A study of Recurring Patterns of
Teaching." Unpublished Ph.D. dissertation, The
University of Nebraska, 1966.

69. Weber, W. A., "Teacher and Pupil Creativity." Unpub-
lished doctoral thesis, Temple Univ., Philadelphia,
1967.

APPENDICES

APPENDIX A-l

NOTICE OF OBSERVATION

Date:

 

Math Teacher:

 

I mentioned to you at the beginning of the school year that
I would be looking at all geometry classes as my experiment
for my dissertation. I will be in your geometry class

at period.

 

What I will be doing, is noting the verbal discourse that
takes place in your geometry classroom, use Flanders'

Method of Interaction Analysis to analyze this verbal behav-
ior. This will require me to tape the discourse. These
visits will in no way constitute your evaluation, by me,
here as a teacher. I will always let you know when I am
coming for this purpose, and I will always talk to you after
I have analyzed the discourse.

Note: If you have a test planned for this day, let me know
and I will come a different day.

Willie E. Williams
Chairman
Dept. of Mathematics

68

APPENDIX A-Z

CONFERENCE SHEET I

FLANDERS INTERACTION ANALYSIS SYSTEM

What are the things that go on between teacher and students
in a classroom setting that determine the "climate" of that
relationship? How has the teacher used his influence to af-
fect classroom interaction? What changes should a teacher
make to produce a classroom atmosphere of the kind desired:
Interaction Analysis, the system suggested by this program,
can provide answers to each of these questions.

Interaction Analysis looks at only one aspect of teaching:
the pedagogical moves which the teacher makes during his
conduct of a class and their relationship to what the stu-
dents do in responding to or provoking those moves. Look-
ing at student-teacher interaction is important, for inter-
action is one of the best reflections of classroom climate,
and it is also a frequent cause of a positive or negative
climate.

This program presents a technique which permits a close ex-
amination of what both teacher and students do in terms of
frequency (e.g., How much time did the teacher spend in lec-
turing?) as well as in terms of sequence, (e.g., What hap-
pened after the teacher asked a question; or what did the
teacher do to provoke that student comment?)

ASSUMPTIONS. Classroom activities which have been selected
for this system are based on the assumption that there are
three broad kinds of things which go on in almost any class-
room:
1. Teacher actions which either encourage or restrict
students.
2. Student actions, either in response to the teacher or
self-initiated.
3. Silence or confusion.

Interaction Analysis assumes that the verbal behavior of

teacher and students is the most important indicator of the
affective tone of the classroom. While there are certainly

69

70

a multitude of other indications, research using this sys-
tem has tended to verify this highly important point. Be-
cause of this verbal emphasis, the system is especially use-
ful to the individual teacher in self—assessment, since there
is no need for an outside observer to "code" what goes on in
the classroom. The teacher need only start his tape recorder
and proceed with his class. He can listen to, code, and
analyze his class by playing the tape in the quiet of an
empty classroom, an office, or his own home.

INTERACTION ANALYSIS, THEN, DEALS WITH CLASSROOM TALK. For
purposes of indentification, student talk is divided into
two categories--one which is induced by teacher questions,
the other which is volunteered by the students and unsolic-
ited by the teacher. Many other kinds of division are pos-
sible; however, in the interest of simplicity and to provide
the most meaningful analysis, these two broad categories
were decided upon.

Teacher talk, on the other hand, probably needs a much broad-
er investigation, since it is the teacher's influence and
it's effect which are of primary interest. If a teacher is
to examine his own verbal behavior in the classroom, he will
be interested specifically in what he did to bring about a
student response and what he did following a student's com-
ments. (19: p. 9)

TOTAL SYSTEM

 

 

Actor Behavior Category Description
1 Accepting student
feelings
Encouraging _ 2 Praising
or 3 Accepting and/or using
Indirect student ideas
Teacher % _4 Asking questions
Restricting or __5 Lecturing
Direct 6 Giving Directions
“' _1 Criticizing of justi-
___ d 8 fyigg :uthoritg
_ Respon ing Stu en Respon 3
Student Initiating 9 Student initiates

response
Silence or confusion

H
O

Silence or
Confusion

APPENDIX A—3

INITIAL CONFERENCE SHEET II

 

 

 

Teacher Date Visit

I/D Ratio i/d Ratio Percent of Teacher-talk
Percent of Student—talk . Percent of Silence and/or
COMMENTS:

71

APPENDIX A-4
CONFERENCE SHEET III

Teacher Date Visit

 

 

Remarks:

1st time 2nd time‘3rd time

 

Percent of Teacher Talk

 

Percent of Student Talk
% of S.T. self initiated

 

 

Percent of Silence and/or confusion

 

*I/D Ratio

 

**i/d Ratio

 

 

 

 

QUESTIONS THAT MAY BE USEFUL TO YOU AS A CLASSROOM TEACHER
Special Note: Average percentages reflect current prac-
tice, not the best or most desired practice.

1. Do I do too much of the talking in the classroom? (Aver-
age about 60%).

2. Am I typically a direct or indirect teacher? I/D less than

1.00, direct; and I/D greater than or equal to 1.00, in-

direct.

How do I react to student verbal behavior?

How much time do I spend in lecturing? (Average about 40%).

Do I spend enough time in the extension of student ideas?

Do students tend to resist my influence?

Do I accept, clarify, and use student emotion?

How effectively do I use praise: (Average between 1% and

2% of total talk.)

9. How effective am I in communicating subject-matter to my
pupils? (Average between 35% and 40% in lecture).

10. How effectively do I use criticism in my teaching? (Aver-
age twice as much criticism as praise).

11. Is there adequate pupil participation in my classroom?

(Average about 24% of total verbal behavior).

GDQOWob-w

72

 

 

Categories 1 2 + 3
. * -_-
NOTE' I/D Categories 5 6 + 7*
**i/d = Categories 1 2 + 3
Categories 6 7

This focuses on the
relative number of

indirect and direct
teacher statements.

This is called a re-
vised I/D ratio and

is employed to find
the kind of emphasis
given to motivation
and control in a par-
ticular classroom.

The i/d ratio elim-
inates the effects of
categories 4 and 5,
asking questions and
lecturing, and gives
evidence about whether
the teacher is direct
or indirect in his ap-
proach to motivation
and control.

SEE ATTACHED SHEET (Your verbal interaction analysis on my

 

last visit)

 

Note: Diagonal cells of the matrix are called "steady-
state cells," all other cells are transitional
cells representing movement from one category to
another. Heavy loading in the diagonal cells in-
dicate extended uses of categories 1-10.

APPENDIX B

NON VERBAL RECORD SHEET

Teacher

Course: Geometry.

General Room Condition

Date

Visit Class Size

What was topic of lesson

 

 

Instructional Aids

Class formation (groups, units, etc.)

 

Preparation of teacher

 

Content presentation

 

Other materials

Chalkboard diagrams

 

Distractions

 

Time in settling down to work

Passing out/collecting materials (time) .

OTHER BEHAVIOR

TEACHER

I.

II.

III.

IV.

Stands

a. Behind desk

b. At chalkboard

c. With book in hand
d. Other

Moves About

a. Sometime

b. Most of time
c. Never

Looks at notes (plans)
a. Sometime

b. Most of time

c. Never

Smiles or laughs
a. Sometime

b. Most of time
c. Never

II.

III.

IV.

74

STUDENT

How answered most of time
a. Right
b. Wrong

Looks around (unattentatively)
a. None

b. Some

c. Frequently

Ignores Instructor
a. None

b. Some

c. Frequently

Slumps

a. None

b. Some

c. Frequently

TEACHER

V.

VI.

Demonstrates

a. Instructional aids
b. Gestures

c. At chalkboard

d. Other

Uses chalkboard
a. Key term (idea)
b. Key diagram

75

STUDENT

V.

VI.

VII.

Yawns, stretches
a. None

b. Some

c. Frequently

Sleeps or dozes
a. None

b. Some

c. Frequently

Group Answers
a. None
b. Some
c. Frequently

APPENDIX C-l

VERBAL CATEGORIES OF FLANDERS
INTERACTION ANALYSIS

Indirect Teacher Behavior

 

Category 1, Acceptance of Feeling. The teacher ac-
cepts feelings when he says he understands how the children
feel, that they have the right to have these feelings, and
that he will not punish the children for their feelings.
These kinds of statements often communicate to children both
acceptance and clarification of the feeling.

Also included in this category are statements that
recall past feeling, refer to enjoyable or uncomfortable
feelings that are present, or predict happy or sad events
that will occur in the future.

In our society people often react to expressions of
negative feelings by offering negative feelings in return.
Acceptance of these emotions in the classroom is quite rare;
probably because teachers find it difficult to accept nega-
tive emotional behavior. However, it may be just as diffi-
cult for them to accept positive feelings. Feelings ex-
pressed by students may also be ignored by the teacher if
he considers the classroom to be a place where people are

concerned primarily with ideas rather than feelings.

76

77

Category 2, Praise or Encouragement. Included in
this category are jokes that release tension, but not those
that threaten students or are made at the expense of indie
vidual students. Often praise is a single word: "good,"
"fine," or "right." Sometimes the teacher simply says,

"I like what you are doing." Encouragement is slightly dif-
ferent and includes statements such as, "Continue," "Go ahead
with what you are saying," "Uh huh; go on; tell us more about
your idea."

Category 3, Accepting Ideas. This category is quite

 

similar to Category 1; however, it includes only acceptance
of student ideas, not acceptance of expressed emotion. When
a student makes a suggestion, the teacher may paraphrase the
student's statement, restate the idea more simply, or sum-
marize what the student has said. The teacher may also say,
"Well, that's an interesting point of view. I see what you
mean." Statements belonging in Category 3 are particularly
difficult to recognize; often the teacher will shift from
using the student's idea to stating the teacher's own idea.

Statements belonging in Category 3 can be identified
by asking the question, "Is the idea that the teacher is now
stating the student's or is it the teacher's?" If it is the
student's idea, then this category is used; if it is the
teacher's, another category must be employed.

Category 4, Askingguestions. This category includes

 

only questions which the teacher expects an answer from the

78

pupils. If a teacher asks a question and then follows it
immediately with a statement of opinion, or if he begins
lecturing, obviously the question was not meant to be an-
swered. A rhetorical question is not categorized as a ques-
tion. An example of another kind of question that should
not be classified in Category 4 is the following: "What in
the world do you think you are doing out of your seat,
John?" With prOper intonation the question is designed to
get John back in his seat; if such is the case, it must be
categorized as criticism of the student's behavior (Cate-
gory 7.)

Questions that are meant to be answered are of sev-
eral kinds. There are questions that are direct in the
sense that there is a right and wrong answer. The question,
"What are 2 and 2?" is a question that limits the freedom of
the student to some extent. Although he can refuse to an-
swer, give the wrong answer, or make a statement of another
kind, in general, this kind of question focuses the student's
answer more than does a question such as, "What do you think
we ought to do now?" Questions, then, can be either narrow
and restrict the student in his answer, or they can be very
broad and give the student a great deal of freedom in answer-
ing. All questions, however broad or narrow, which require
answers and are not commands or criticism, fall into Cate-

gory 4.

79

Direct Teacher Behavior

 

Category 5, Lecture. Lecture is the form of verbal

 

interaction that is used to give information, facts, opin-
ions, or ideas to children. The presentation of material
may be used to introduce, review, or focus the attention of
the class on an important topic. Usually information in the
form of lecture is given in fairly extended time periods,
but it may be interspersed with children's comments, ques-
tion, and encouraging praise.

Whenever the teacher is explaining, discussing, giv-
ing opinion, or giving facts or information, Category 5 is
used. Rhetorical questions are also included in this cate-
gory. Category 5 is the one most frequently used in class-
room observation.

Category 6, Giving Directions. The decision about
whether or not to classify the statement as a direction or
command must be based on the degree of freedom that the stu-
dent has in response to teacher direction. If he says,
"John, go to the board and write your name," he is giving a
direction or command. When he says, "John, I want you to
tell me what you have done with your reader," he is still
giving a direction.

Category 7, Criticizing_or Justifying Authority. A
statement of criticism is one that is designed to change stu-

dent behavior from nonacceptable to acceptable. The teacher

80

is saying, in effect, "I don't like what you are doing. Do
something else." Another group of statements included in
this category are those that might be called statements of
defense or selfejustification. These statements are partic-
ularly difficult to detect when a teacher appears to be ex-
plaining a lesson or the reasons for doing a lesson to the
class. If the teacher is explaining himself or his author-
ity, defending himself against the student, or justifying
himself, the statement falls in this category.

Other kinds of statements that fall in this category
are those of extreme self-reference or those in which the
teacher is constantly asking the children to do something as
a special favor to the teacher.

Categories 1 through 4, those of indirect teacher
influence, and categories 5 through 7, those of direct teach-
er influence, have been described. They are all categories
of teacher talk. Whenever the teacher is talking, the state-
ments must be categorized in one of the first seven cate-
gories. If the observer decides that with a given statement
the teacher is restricting the freedom of children, the
statement is tallied in Categories 5, 6, or 7. If, on the
other hand, the observer decides that the teacher is expand-
ing freedom of children, the category used is either 1, 2,
3, or 4.

There are three additional categories for use in

classroom interaction:

81

Category 8, Student Talk-Response. This category
is used when the teacher has initiated the contact or has
solicited student statements, when the student answers a
question asked by the teacher, or when he responds verbally
to a direction the teacher has given. Anything that the
student says that is clearly in response to initiation by
the teacher belongs in Category B.

Category 9, Student Talk-Initiation. In general,

 

if the student raises his hand to make a statement or to
ask a question when he has not been prompted to do so by
the teacher, the appropriate category is nine.

Distinguishing between Categories 8 and 9 is often
difficult. Predicting the general kind of answer that the
student will give in response to a question from the teacher
is important in making this distinction. If the answer is
one that is of a type predicted by the observer (as well as
the teacher and class), then the statement comes under Cate-
gory 8. When in response to a teacher-question the student
gives an answer different from that which is expected for
that particular question, then the statement is categorized
as a nine.

Category 10, Silence or Confusion. This category
includes anything else not included in the other categories.
Periods of confusion in communication, when it is difficult
to determine who is talking, are classified in this category.

A summary of these categories, with brief definitions

for the use of the observer will follow.

APPENDIX C-2

SUMMARY OF CATEGORIES FOR
INTERACTION ANALYSIS

 

T E A C H E R T A L K

INDIRECT INFLUENCE

*ACCEPTS FEELING: accepts and clarifies the
feeling tone of the students in a nonthreat-
ening manner. Feelings may be positive or
negative. Predicting or recalling feelings
is included.

 

*PRAISES OR ENCOURAGES: praises or encourages
stfident action or behavior. Jokes that re-
lease tension, but not at the expense of an-
other individual; nodding head, or saying
"um hm" or "go on" are included.

 

*ACCEPTS OR USES IDEAS OF STUDENTS: clarify-
ing, Building, or developing ideas suggested
by a student. As teacher brings more of his
own ideas into play, shift to Category 5.

 

*ASKS QUESTIONS: asking a question about con-
tent or procedure with the intent that a stu-
dent answer.

 

 

 

DIRECT INFLUENCE

 

*LECTURING: giving facts or opinions about
content or procedures; expressing his own
ideas, asking rhetorical questions.

 

*GIVING DIRECTIONS: directions, commands, or
orders with which a student is expected to
comply.

 

*CRITICIZING OR JUSTIFYING AUTHORITY: state-
ments intended to change student behavior
from nonacceptable pattern; bawling someone
out; stating why the teacher is doing what
he is doing; extreme self-reference.

 

82

83

 

 

TALK

STUDENT

 

8. *STUDENT TALK - RESPONSE: talk by students in
response to teacher. Teacher initiates the con-
tact or solicits student statement.

9. *STUDENT TALK - INITIATION: talk by students,
which they initiate. If "calling on" student
is only to indicate who may talk next, observer
must decide whether student wanted to talk. If
he did, use this category.

10. *SILENCE OR CONFUSION: pauses, short periods of
silence, and periods of confusion in which com-
munication cannot be understood by the observer.

 

 

*There is NO scale implied by these numbers. Each
number is classificatory; it designates a particular
kind of communication event. To write these numbers
down during observation is to enumerate--not to
judge a position on a scale.

APPENDIX C-3

PROCEDURE FOR CATEGORIZING
TEACHER-PUPIL INTERACTION

The Flanders' system of interaction analysis was
originally used as a research tool and continues to serve
this function. As such, it is employed by a trained ob-
server in order to collect reliable data regarding class—
room behavior as a part of a research project.

As it is described in this by Flanders, the system
is meant to be used as an in-service training device for
teachers. It may be employed by a teacher either as he ob-
serves someone else teach or as he categorizes a tape re-
cording of his own classroom behavior. In either case the
method is the same.

Every three seconds the observer writes down the
category number of the interaction he has just observed.
He records these number of the interaction he has just ob-
served. He records these numbers in sequence in a column
(row). He will write approximately 20 numbers per minute;
thus, at the end of a period of time, he will have several
long columns (rows) of numbers. The observer preserves
this sequence of numbers that he has recorded. It is im-

portant to keep the tempo as steady as possible, but it is

84

85

even more crucial to be accurate. He may also wish to write
down marginal notes from time to time, which can be used to
explain what has been happening in the classroom.

No matter whether he is using a live classroom or a
tape recording for his observations, it is best for the ob-
server to spend 5 to 10 minutes getting oriented to the sit-
uation before he actually begins to categorize. He then
has a feeling for the total atmosphere in which the teacher
and pupils are working. After he has begun to get the feel-
ing of the classroom interaction, he begins to record the
interaction.

The observer st0ps classifying whenever the class-
room activity is changed so that observing is inappropriate
as, for instance, when there are various groups working
around the classroom, or when children are working on work-
books or doing silent reading. He will usually draw a line
under the recorded numbers, make a note of the new activity,
and resume categorizing when the total class discussion con-
tinues. At all times the observer notes the kind of class
activity he is observing. The reading group in the elemen-
tary school is obviously different from an informal discus-
sion period, a review of subject matter, a period of super-
vised seat work, teacher-directed discussion, introduction
of new material, or evaluation of a unit that has been com-

pleted. Such diverse activities may be expected to show

86

different types of teacher-pupil interactions even when
guided by the same teacher. A shift to new activity should

also be noted.

 

 

 

 

 

Teacher

 

87

APPENDIX D-l

ANALYSIS SHEET

Class size Week

 

 

 

 

l min.

 

2 min,

 

 

 

 

 

3 min,

 

 

 

 

 

4 min.

 

 

 

 

 

 

5 min.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6 min.
F I
7 min.
I
8 min.
I I I
9 min.
I
10 min.
| I
11 min.

 

 

 

 

 

12'min.

Page

 

 

J

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 3U, 45 13 min.
II I I I.
I4 min,
I II
15 min.
I I
16 min.
I J I
17 min.
II
18 min.
I L
19 min.
II
20 min.
I III
21 min.
I III
22 min.
I I I I
23 min.
I I I I
24 min.

 

 

88

APPENDIX D-2
10 X 10 DATA MATRIX

Teacher Date Class size Week

 

I/D ratio i/d ratio % Teacher talk % Student talk

% Silence or Confusion

DATA MATRIX

Category 1 2 l3 4 5 6 7 8 9 101 Total

1

 

 

 

 

 

 

 

 

 

10

 

 

 

 

 

 

 

 

 

 

 

 

 

 

"IIIIIIIIIIIIIIII