C:\rsm\y520\readings\madsen68\madsen68.

fm 1
Journal of Applied Behavior Analysis, Vol. 1, pp. 139-150, Number 2 (Summer 1968).
Rules, Praise, And Ignoring:
Elements Of Elementary Classroom Control
Charles H. Madsen, Jr., Wesley C. Becker, and Don R. Thomas
1
Florida State University And University of Illinois
An attempt was made to vary systematically the behavior of two elementary school teachers to
determine the effects on classroom behavior of Rules, Ignoring Inappropriate Behaviors, and
showing Approval for Appropriate Behavior. Behaviors of two children in one class and one child
in the other class were recorded by observers, as were samples of the teachers behavior. Follow-
ing baseline recordings. Rules, Ignoring, and Approval conditions were introduced one at a time.
In one class a reversal of conditions was carried out. The main conclusions were that: (a) Rules
alone exerted little effect on classroom behavior, (b) Ignoring Inappropriate Behavior and show-
ing Approval for Appropriate Behavior (in combination) were very effective in achieving better
classroom behavior, and (c) showing Approval for Appropriate Behaviors is probably the key to
effective classroom management.
1 Modern learning theory is slowly but surely increasing its potential for impact upon social problems. As prob-
lems in social development and interaction are more closely examined through the methods of experimental
analysis, the importance of learning principles in everyday life becomes clearer. The potential contribution of
these developments to childrearing and education appears to be especially significant. This report is a part of a
series of studies aimed at demonstrating what the teacher can do to achieve a happier, more effective classroom
through the systematic use of learning principles. The study grows out of a body of laboratory and field research
demonstrating the importance of social reinforcers (smiles, praise, contact, nearness, attention) in establishing
and maintaining effective behaviors in children. Extensive field studies in experimental nursery schools by Wolf,
Bijou, Baer, and their students (e.g., Hart. Reynolds. Baer. Brawley, and Harris, 1968; Allen, Hart, Buell, Harris.
and Wolf, 1965; Bijou and Baer, 1963) provided a background for the extension of their work by the present
authors to special and typical elementary classrooms. In general, we have found to date that teachers with various
personalities and backgrounds can be trained systematically to control their own behavior in ways which will
improve the behavior of the children they are teaching. (Becker, Madsen, Arnold, and Thomas, 1967). We have
also found that teachers can create problem behaviors in the classroom by controlling the ways in which they
respond to their pupils (Thomas, Becker, and Armstrong, 1968; Madsen, Becker, Thomas, Koser, and Plager,
1968). It is hoped that field studies of this sort will contribute to more effective teacher training.
2 The present study is a refinement of an earlier study of Becker et al. (1967), in which the behavior of two children
in each of five classrooms was recorded and related to experimentally controlled changes in teacher behaviors.
The teachers were instructed and guided to follow a program which involved making classroom rules explicit,
ignoring disruptive behaviors unless someone was getting hurt, and praising appropriate classroom behaviors.
1
We wish to express our appreciation to the teachers involved, Mrs. Barbara L. Weed and Mrs. Margaret
Larson, for their cooperation in a study which involved using and applying procedures which at times
made their teaching duties very difficult. Gratitude is expressed to the Director of Elementary Education,
Unit District #116, Urbana, Illinois, Dr. Lowell M. Johnson, and to the principals of Thomas Paine and
Prairie Schools, Richard Sturgeon and Donald Hoiste. This study was supported by Grant HD-00881-05
from the National Institutes of Health. Reprints may be obtained from Wesley C. Becker, Bureau of Edu-
cational Research, 284 B Education Bldg., University of Illinois, Urbana, Illinois 61801.
.
C:\rsm\y520\readings\madsen68\madsen68.fm 2
Under this program, most of the severe problem children under study showed remarkable improvements in
classroom behavior. However, that study lacked certain controls which the present study sought to correct. First,
the teachers in the earlier study were in a seminar [begin page 140] on behavior theory and practice during base-
line conditions. Some target children improved during baseline, apparently because some teachers were begin-
ning to apply what they were learning even though they had been requested not to do so. Second, public relations
and time considerations did not make it possible to introduce the components of the experimental program one
at a time (rules, ignoring, and praise) to better study their individual contributions. Third, a reversal of teacher
behavior was not attempted. Such a reversal would more conclusively show the importance of teachers behavior
in producing the obtained changes. Fourth, extensive recordings of teacher behavior under all experimental con-
ditions were not undertaken in the earlier study. The present study attempted to deal with each of these prob-
lems.
METHOD
Procedures
3 Teachers in a public elementary school volunteered to participate in the study. After consultation with teachers
and observation of the children in the classroom, two children with a high frequency of problem behavior were
selected for study in each class. Previously developed behavioral categories (Becker et al., 1967) were modified
for use with these particular children and baseline recordings were made to determine the frequency of problem
behaviors. At the end of the baseline period the teachers entered a workshop on applications of behavioral princi-
ples in the classroom which provided them with the rationale and principles behind the procedures being intro-
duced in their classes. Various experimental procedures were then introduced, one at a time, and the effects on
the target childrens behaviors observed. The experiments were begun in late November and continued to the end
of the school year.
Subjects
4 Classroom A. There were 29 children in Mrs. As middle-primary (second grade) room who ranged in school
progress from mid-first-grade level to early-third-grade level. Cliff and Frank were chosen as the target children.
5 Cliff was chosen because he displayed no interest in school. In Mrs. As words, he would sit throughout entire
work periods fiddling with objects in his desk, talking, doing nothing, or misbehaving by bothering others and
walking around the room. Lately he has started hitting others for no apparent reason. When Cliff was required to
stay in at recess to do his work, he would complete the work in a short time and it was usually completely accu-
rate. I was unable to motivate him into working on any task during the regular work periods. Cliff is the son of a
university professor who was born in Europe and immigrated when Cliff was 5-yr old. Cliff scored 91 on an early
(CA 5-3) intelligence test. This score was discounted by the examiner because of language problems. His group
IQ scores rose steadily (CA 5-9, IQ 103; CA 6-2, IQ 119; CA 7-1, IQ 123). His achievement scores indicated a low
second-grade level at the beginning of the present study. Cliff was seen by the school social worker throughout
the entire first grade and throughout this entire study.
6 Cliff was observed early in the year and it was noted that he did not respond once to teachers questions. He
played with his fingers, scratched himself repeatedly, played in his desk, paid no attention to the assignment and
had to stay in at recess to finish his work. Almost continually he made blowing sounds and talked to himself. On
occasions he was out of his seat making noises and talking. He would leave the room without permission. Before
the study began the observers made the following notes: What a silly kid, writing on the bottom of his shoes,
writing on his arms, blowing kisses at the girls. He was vying for the attention of the girl behind him, but she
ignored him. . . . Poor Cliff! he acts so silly for his age. He tried to talk to the other kids, but none of them would
pay attention to him. . . . Cliff seems concerned with the little girl beside him (girl behind him last week). He has
a sign on his desk which reads, Do you love me?. . . .
C:\rsm\y520\readings\madsen68\madsen68.fm 3
7 Frank was described by his teacher as a likable child. He had a record of misbehavior in the classroom and intense
fighting on the play ground. He was often out of his seat talking to other children and did not respond to disci-
pline. If someone was reprimanded for doing something, Frank would often do the same thing. Test scores indi-
cated an IQ of 106 (Stanford-Binet) and achievement level just under beginning second grade at the start of
school (average California Achievement Test [begin page 141] scores 1.6 grades). The school psychologist noted
that Franks mother was a person who willingly permitted others to make decisions for her and did not seem
able to discipline Frank. Father was absent from the home during the entire year in the Air Force.
8 Classroom B. Twenty children were assigned to Mrs. Bs kindergarten room. Two children were observed initially;
one moved from the community shortly after baseline was taken, leaving only Stan for the study.
9 Stan was described as coming from a truly pathetic home environment. The mother was not married and the
family of four children subsisted on state aid. One older brother was enrolled in a special class for the educable
retarded. At the beginning of the year, Stans behavior was characterized by the teacher as wild. She reported
that, Stan would push and hit and grab at objects and at children. He had no respect for authority and appar-
ently didnt even hear directions. He knew how to swear profusely, and I would have to check his pockets so I
would know he wasnt taking home school equipment. He would wander around the room and it was difficult to
get him to engage in constructive work. He would frequently destroy any work he did rather than take it home.
10 The difficult home situation was made manifest during the month of March. Stan had been absent for two weeks
and it was reported that his mother was taking her children out of public school and placing them in a local paro-
chial school. Investigation by school personnel indicated that Stans mother had moved the children into a rela-
tives home and had gone to the hospital to have another illegitimate baby. A truancy notice was filed for all four
children including Stan. Following legal notice the children were returned to school.
Rating of Child Behavior
11 The same rating schedule was used in both classrooms except that Isolate Play was added to the list of Inappro-
priate Behaviors for the kindergarten. Since the children were expected to be involved in structured group activi-
ties during observation periods, going off by oneself to play with the many toys or materials in the room was
considered inappropriate by the kindergarten teacher. Inappropriate Behavior was defined as the occurrence of
one or more of the behaviors listed under Inappropriate Behavior in Table 1 during any observation interval.
12 Observers were trained in the reliable use of the rating schedule before baseline recordings began. Training con-
sisted of practice in use of the rating schedule in the classroom. Two observers would each rate the same child for
20 min and then return to the research office to compare their ratings and discuss their differences with their
supervisor. Training was continued until reliability was above 80% on each behavior code. Training lasted
approximately two weeks. Reliability was determined periodically throughout the study by dividing the number
of agreements by the number of agreements plus disagreements. An agreement was defined as a rating of the
same behavior class in the same observation interval. Average reliability over children, behavior classes, and days
for the 69 occasions (out of 238) on which it was checked was 81%. Single day reliabilities ranged from 68% to
96%. Reliabilities were checked in each phase of the study.
13 Instructions to observers followed those used by Becker et al. (1967). In essence, the observers were not to
respond to the children, but to fade into the background as much as possible. Teachers, as well as children,
quickly learned not to respond to the observers, although early in the study one observer was attacked by a kin-
dergarten child. The observer did not respond to the behavior and it quickly disappeared. Experimental changes
were initiated without informing observers in an attempt to control any observer bias. However, the changes were
often dramatic enough that observer comments clearly reflected programmed changes in teachers behavior.
14 The target children were observed for 20 min per day, three days a week. In the middle primary class, observa-
tions were taken when the children were engaged in seat work or group instruction. In the kindergarten class,
observations were made when structured activities, rather than free play, were expected. Each observer had a
clipboard, stopwatch, and rating sheet. The observer would watch for 10 sec and use symbols to record the
occurrence of behaviors. In each minute, ratings would be made in five consecutive 10-sec intervals and the final
C:\rsm\y520\readings\madsen68\madsen68.fm 4
10 sec would be used for recording comments. Each behavior category could be rated only once in a 10-sec inter-
val. [begin page 142]
15 The primary dependent variable was percentage of intervals in which an Inappropriate Behavior occurred. Since
the varieties of Inappropriate Behavior permitted a more detailed analysis with the schedule used, the presenta-
tion of results is focussed on them, even though functionally their converse (Appropriate Behavior) was the main
behavior being manipulated.
Ratings of Teacher Behavior
16 Ratings of teacher behavior were obtained to clarify relationships between changes in teacher behavior and
changes in child behavior. Recordings of teacher behavior were also used by the experimenters to help the teach-
ers learn the contingent use of Approval and Disapproval Behaviors. The teacher rating schedule is presented in
Table 2. Teacher behaviors were recorded by subclasses in relation to child behaviors. That is, the record would
show whether a teacher response followed Appropriate child classroom behavior or whether it followed one of
Table 1: Behavioral Coding Categories for Children
I. Inappropriate Behaviors
A. Gross Motor. Getting out of seat, standing up, run-
ning, hopping, skipping, jumping, walking
around, moving chair, etc.
F. Turning Around. Turning head or head and body to look at
another person, showing objects to another child, attending to
another child. Must be of 4-sec duration, or more than 90
degrees using desk as a reference. Not rated unless seated. If
this response overlaps two time intervals and cannot be rated
in the first because it is less than 4-sec duration, then rate in
the interval in which the end of the response occurs.
B. Object Noise. Tapping pencil or other objects, clap-
ping, tapping feet, rattling or tearing paper, throw-
ing book on desk, slamming desk. Be conservative,
only rate if you can hear the noise when eyes are
closed. Do not include accidental dropping of
objects.
G. Other Inappropriate Behavior. Ignores teachers question or
command. Does something different from that directed to do,
including minor motor behavior such as playing with pencil or
eraser when supposed to be writing, coloring while the record
is on, doing spelling during the arithmetic lesson, playing with
objects. The child involves himself in a task that is not appro-
priate. Not rated when other Inappropriate Behaviors are
rated. Must be time off task.
C. Disturbance of Others Property. Grabbing objects
or work, knocking neighbors books off desk,
destroying anothers property, pushing with desk
(only rate if someone is there). Throwing objects at
another person without hitting them.
H. Mouthing Objects. Bringing thumb, fingers, pencils or any
object in contact with the mouth.
D. Contact (high and low intensity). Hitting, kicking,
shoving, pinching, slapping, striking with object,
throwing object which hits another person, poking
with object, biting, pulling hair, touching, patting,
etc. Any physical contact is rated.
I. Isolate Play. Limited to kindergarten free-play per period. Child
must be farther than 3 ft from any person, neither initiates or
responds to verbalizations with other people, engages in no
interaction action of a non-verbal nature with other children
for the entire 10-sec period.
E. Verbalization. Carrying on conversations with
other children when it is not permitted. Answers
teacher without raising hand or without being
called on; making comments or calling out
remarks when no questions have been asked: call-
ing teachers name to get her attention; crying,
screaming, singing, whistling, laughing, coughing,
or blowing loudly. These responses directed to
teacher or children
II. Appropriate Behavior
Time on task; e.g., answers question, listens, raises hand, works on
assignment. Must include while 10-sec interval except for Turning
Around responses of less than 4-sec duration.
C:\rsm\y520\readings\madsen68\madsen68.fm 5
the categories of Inappropriate Behavior. Responses to all children were rated. Teacher behavior was scored as the
frequency of occurrence of a specified class of behavior during a 20-min interval. Teacher ratings were either
recorded during one of the periods when a target child was being rated by another observer, or immediately
thereafter when only one observer made both ratings. Teacher behavior was rated on the average of once a week,
except during experimental transitions, when more frequent ratings were made. The number of days teacher
behavior was rated under each condition is given in Table 3. Most recorded teacher behavior (about 85%) fell in
the Verbal Approval or Disapproval categories. For this reason we have used the term Praise interchangeably with
Approval Behaviors and Criticism interchangeably with Disapproval Behaviors.
17 Reliability of measures of teacher behavior were checked approximately every other rating day (21 of 42 occa-
sions for the two teacher [begin page 143]
18 by dividing the agreements as to time interval and behavior codes by the agreements plus disagreements. Average
reliability over behavior classes, teachers, and days was 84% with a range from 70% to 96% for individual day
measures.
Experimental Conditions
19 In the middle-primary class (Class A) the experimental conditions may be summarized as consisting of Baseline;
Table 2: Coding Definitions for Teacher Behaviors
Appropriate child behavior is defined by the child rating
categories. The teachers rules for classroom behavior
must be considered when judging whether the childs
behavior is Appropriate or Inappropriate.
I. Teacher Approval following Appropriate Child Behav-
ior
B. Criticism. Critical comments of high or low intensity, yelling,
scolding, raising voice. Examples: thats wrong, dont do that,
stop talking, did I call on you, you are wasting your time, dont
laugh, you know what you are supposed to do.
A. Contact. Positive physical contact such as embrac-
ing, kissing, patting, holding arm or hand, sitting
on lap.
C. Threats. Consequences mentioned by the teacher to be used at
a later time. If _____________ then ________________ com-
ments.
B. Praise. Verbal comments indicating approval, com-
mendation or achievement. Examples: thats good,
you are doing right, you are studying well, I like
you, thank you, you make me happy.
D. Facial attention. Frowning or grimacing at a child.
C. Facial attention. Smiling at child. IV. Teacher Disapproval following Inappropriate Child Behavior
Same codes as under III.
II. Teacher Approval following Inappropriate Child
Behavior
Same as codes under I.
V. Timeout Procedures
a
A. The teacher turns out the lights and says nothing.
B. The teacher turns her back and waits for silence.
C. The teacher stops talking and waits for quiet.
D. Keeping in for recess.
E. Sending child to office.
F. Depriving child in the classroom of some privilege.
a.These are procedural definitions of teacher behaviors possibly involving the withdrawal of reinforcers as a consequence of disruptive
behaviors which teacher could not ignore.
III. Teacher Disapproval following Appropriate Child
Behavior.
A. Holding the child. Focibly holding the child, put-
ting child out in the hall, grabbing, hitting, spanking,
slapping, shaking the child.
VI. Academic Recognition
Calling on a child for an answer. Giving feedback for academic
correctness.
C:\rsm\y520\readings\madsen68\madsen68.fm 6
introduction of Rules; Rules plus Ignoring deviant behavior; Rules plus Ignoring plus Praise for appropriate
behavior: return to Baseline; and finally reinstatement of Rules, Ignoring, and Praise. In the kindergarten class
(Class B) the experimental conditions consisted of Baseline; introduction of Rules; Ignoring Inappropriate Behav-
ior (without continuing to emphasize rules); and the combination of Rules, Ignoring, and Praise.
20 The various experimental procedures were to be used by the teachers for the classroom as a whole throughout the
day, not just for the children whose behavior was being recorded, and not just when observers were present.
21 Baseline. During the Baseline period the teachers conducted their classes in their typical way. No attempt was
made to influence their behavior.
22 Rules. Many people would argue that just telling children what is expected should have considerable effect on
their behavior. We wished to explore this question empirically. Teachers were instructed individually and given
written instructions as follows:
The first phase of your participation in the use of behavioral principles to modify classroom
behaviors is to specify explicit rules of classroom conduct. When this is done, there is no doubt as
to what is expected of the children in your classroom. However, do not expect a dramatic shift in
classroom control, as we all know that knowing the prohibitions does not always keep people from
sin. This is the first phase in the program and inappropriate behavior should be reduced, but per-
haps not eliminated. The rules should be formulated with the class and posted in a conspicuous
location (a chart [begin page 144] in front of the room or a special place on the chalkboard where
they will not be erased). Go over the rules three or four times asking the class to repeat them back
to you when they are initially formulated and use the following guidelines:
(a) Make the rules short and to the point so they can be easily memorized.
(b) Five or six rules are adequate. Special instructions for specific occasions are best given when
the occasion arises. Children will not remember long lists of rules.
(c) Where possible phrase the rules in a positive not a negative manner (for example, Sit quietly
while working, rather than, Dont talk to your neighbors). We want to emphasize positive
actions.
(d) Keep a sheet on your desk and record the number of times you review the rules with the class
(strive for at least four to six repetitions per day). Remember that young children do not have the
retention span of an adult and frequent reminders are necessary. Let the children recite the rules as
you ask them, rather than always enumerating them yourself.
(e) Remind the class of the rules at times other than when someone has misbehaved.
(f) Try to change no other aspects of your classroom conduct except for the presentation of the
rules at appropriate times.
23 Teacher tally sheets indicated that these instructions were followed quite explicitly. The average number of pre-
sentations of rules was 5.2 per day.
24 Ignoring Inappropriate Behavior. The second experimental phase involved Ignoring Inappropriate Behavior. In
Class A, repetition of rules was also continued. Individual conferences to explain written instructions were given
both teachers. Both teachers were given the following instructions:
The first aspect of the study was to make expectations explicit. This you have been doing over the
past few weeks. During the next phase of the study you should learn to ignore (do not attend to)
behaviors which interfere with learning or teaching, unless of course, a child is being hurt by
another, in which case use a punishment which seems appropriate, preferably withdrawal of some
positive reinforcement. Learning to ignore is rather difficult. Most of us pay attention to the viola-
tions. For example, instead of ignoring we often say such things as the following: Johnny, you
know you are supposed to be working; Sue, will you stop bothering your neighbors; Henrietta,
you have been at that window for a long time; Jack, can you keep your hands off Bill; Susie, will
C:\rsm\y520\readings\madsen68\madsen68.fm 7
you please sit down; Alex, stop running around and do your work; Jane, will you please stop
rocking on your chair.
Behaviors which are to be ignored include motor behaviors such as getting out of seat, standing
up, running, walking around the room, moving chairs, or sitting in a contorted manner. Any verbal
comment or noise not connected with the assignments should also be ignored, such as: carrying on
conversations with other children when it is not permitted, answering questions without raising
hands or being called on, making remarks when no questions have been asked, calling your name
to get attention, and extraneous noises such as crying, whistling, laughing loudly, blowing noise, or
coughing. An additional important group of behaviors to be ignored are those which the student
engages in when he is supposed to be doing other things, e.g., when the child ignores your instruc-
tions you are to ignore him. Any noises made with objects, playing with pencils or other materials
should be ignored, as well as, taking things from or disturbing another student by turning around
and touching or grabbing him.
The reason for this phase of the experiment is to test the possibility that attention to Inappropriate
Behavior may serve to strengthen the very behavior that the attention is intended to diminish.
Inappropriate Behavior may be strengthened by paying attention to it even though you may think
that you are punishing the behavior.
25 Praise for Appropriate Behavior. The third phase of the experiment included individual contacts with teachers to
encourage and train Praising of Appropriate Behavior. The Praise instructions to the teachers were as follows:
[begin page 145} The first phase included specifying explicit rules, writing them on the board and
reviewing them 4-6 times per day. The second phase was designed to reduce the amount of atten-
tion paid to behaviors which were unwanted by ignoring them. This third phase is primarily
directed toward increasing Appropriate Behaviors through praise and other forms of approval.
Teachers are inclined to take good behavior for granted and pay attention only when a child acts up
or misbehaves. We are now asking you to try something different. This procedure is characterized
as catching the child being good and making a comment designed to reward the child for good
behavior. Give praise, attention, or smile when the child is doing what is expected during the par-
ticular class period in question. Inappropriate Behavior would not be a problem if all children were
engaging in a great deal of study and school behavior, therefore, it is necessary to apply what you
have learned in the workshop. Shape by successive approximations the behavior desired by using
praise and attention. Start small by giving praise and attention at the first signs of Appropriate
Behavior and work toward greater goals. Pay close attention to those children who normally engage
in a great deal of misbehavior. Watch carefully and when the child begins to behave appropriately,
make a comment such as, Youre going a fine job, (name). It is very important during the first few
days to catch as many good behaviors as possible. Even though a child has just thrown an eraser at
the teacher (one minute ago) and is now studying, you should praise the study behavior. (It might
also decrease the rate of eraser throwing.) We are assuming that your commendation and praise are
important to the child. This is generally the case, but sometimes it takes a while for praise to
become effective. Persistence in catching children being good and delivering praise and attention
should eventually pay off in a better behaved classroom.
Some examples of praise comments are as follows:
I like the way youre doing your work quietly (name).
Thats the way I like to see you work _____________.
Thats a very good job ________________________.
Youre doing fine _________.
You got two right__________________, thats very good (if he generally gets no answers right).
In general, give praise for achievement, prosocial behavior, and following the group rules. Specifi-
cally, you can praise for concentrating on individual work, raising hand when appropriate,
responding to questions, paying attention to directions and following through, sitting in desk and
studying, sitting quietly if noise has been a problem. Try to use variety and expression in your com-
C:\rsm\y520\readings\madsen68\madsen68.fm 8
ments. Stay away from sarcasm. Attempt to become spontaneous in your praise and smile when
delivering praise. At first you will probably get the feeling that you are praising a great deal and it
sounds a little phony to your ears. This is a typical reaction and it becomes more natural with the
passage of time. Spread your praise and attention around. If comments sometimes might interfere
with the ongoing class activities then use facial attention and smiles. Walk around the room during
study time and pat or place your hand on the back of a child who is doing a good job. Praise quietly
spoken to the children has been found effective in combination with some physical sign of
approval.
General Rule: Give praise and attention to behaviors which facilitate learning. Tell the child what
he is being praised for. Try to reinforce behaviors incompatible with those you wish to decrease.
26 The teachers were also instructed to continue to ignore deviant behavior and to repeat the rules several times a
day.
27 Additional training given teachers consisted of: (a) discussion of problems with suggested solutions during
weekly seminars on behavior analysis, and (b) specific suggestions from the experimenter on possible alternative
responses in specific situations based on the experimenters observations of the teachers during experimental
transitions, or based on observer data and notes at other times when the data showed that the teachers were not
on program.
28 [begin page 146] Additional cues were provided to implement the program. Cards were placed on the teachers
desks containing the instructions for the experimental phase in which they were engaged.
29 Reversal. In Class A the final experimental conditions involved an attempt to return to Baseline, followed by a
reinstatement of the Rules, Praise, and Ignore condition. On the basis of the earlier observations of Teacher A, we
were able to specify to her how frequently she made disapproving and approving comments. The success of this
procedure can be judged from the data.
RESULTS
30 Percentage of observation intervals in which Inappropriate Behaviors occurred as a function of conditions is
graphed in Fig. 1 and 2. Major changes in Inappropriate Behaviors occurred only when Praise or Approval for
Appropriate Behaviors was emphasized in the experimental procedures. A t test, comparing average Inappropri-
ate Behavior in conditions where Praise was emphasized with those where Praise was not emphasized, was signif-
icant at the 0.05 level (df = 2).
31 Before examining the results more closely, it is necessary to inspect the data on teacher behavior. Table 3 gives the
frequency of classes of teacher behaviors averaged within experimental conditions. Since day-to-day variability
of teacher behavior was low for the measures used, these averages fairly reflect what went on.
32 Introduction of Rules into the classroom had no appreciable effect on Inappropriate Behavior.
33 Ignoring Inappropriate Behaviors produced inconsistent results. In Class A the children clearly became worse
under this condition; in Class B little change was apparent. Both teachers had a difficult time adhering to this
condition, and Teacher A found this phase of the experiment very unpleasant. Table 3 shows that Teacher A was
only able to reduce critical comments from an average of one per 1 min to an average of three in 4 min. Teacher B
cut her critical comments in half. In view of these difficulties, the present results cannot be taken as a clear test of
the effects of responding with Disapproval to inappropriate Behaviors.
34 The failure to eliminate Disapproval Reactions to inappropriate Behaviors in Phase Three of the experiment,
adds some ambiguities to the interpretation of the Phase Four data for Teacher A. The Rules, Ignore, and Praise
condition for Teacher A involved both a reduction in critical comments (Ignoring) as well as a marked increase in
Praise. As demonstrated previously (Becker et al., 1967), this
C:\rsm\y520\readings\madsen68\madsen68.fm 9
C:\rsm\y520\readings\madsen68\madsen68.fm 10
[begin page 148] combination of procedures is very effective in reducing inappropriate classroom behaviors, but
we still lack a clear isolation of effects. The data for Teacher B are not confounded with a simultaneous shift in fre-
quency of Disapproval and Approval Reactions, but they are made less interpretable by a marked shift in Aca-
demic Recognition (defined in Table 2) which occurred when the shift in Praise was made. Since Academic
Recognition does not show any systematic relations to level of Appropriate Behaviors elsewhere in the study, we
are not inclined to interpret this change as showing a causal effect. A best guess is that the effective use of Praise
gave the teacher more time to focus on academic skills.
35 The reversal operation for Teacher A quite clearly shows that the combination of Praising and Ignoring exerts a
strong control over Appropriate Behaviors.
36 As with Academic Recognition, no attempt was made to control how frequently the teacher used procedures
labelled Timeout (defined in Table 2). The frequency data reported in Table 4 indicates that during Baseline,
Teacher A, especially, used Timeout procedures to try to establish control (usually turning off the lights until
the children were quiet). The changes in the frequency of use of Timeout procedures are not systematically
related to the behavior changes graphed in Fig. 1 and 2.
37 In summary, the main results indicate: (a) that Rules alone had little effect in improving classroom behavior, (b)
the functional status of Ignoring Inappropriate Behavior needs further clarification, (c) the combination of
Ignoring and Praising was very effective in achieving better classroom behavior, and (d) Praise for Appropriate
Behaviors was probably the key teacher behavior in achieving effective classroom management.
38 The effects of the experimental procedures on individual classes of behavior for the two children in Class A are
presented in Table 4. The data in Table 4 illustrate that with a few exceptions the effects on individual classes of
behavior are similar to those for Inappropriate Behavior as a whole.
Teacher A Behavior Classes Baseline I Rules
Rules +
Ignore
Rules +
Ignore +
Praise I Baseline II
Rules +
Ignore +
Praise II
Approval to Appropriate 1.2 2 0 18.2 2.5 12.5
Approval to Inappropriate 8.7 0.8 20 1.2 4 5.1
Disapproval to Inappropriate 18.5 20.5 15.7 4.1 9.8 3.5
Disapproval to Appropriate 0.9 0.7 1 0.3 0.9 0
Timeout 3.3 1.4 1.7 0.4 0 0.1
Academic Recognition 26.5 23.6 46.3 52.4 45.4 45.6
Days Observed 15 8 3 11 4 9
Teacher B Behavior Classes Baseline I Rules
Rules +
Ignore
Rules +
Ignore +
Praise I Baseline II
Rules +
Ignore +
Praise II
Approval to Appropriate 19.2 14.1 19.3 35.2
Approval to Inappropriate 1.9 0.9 0.3 0
Disapproval to Inappropriate 16.9 22.1 10.6 10.8
Disapproval to Appropriate 0 0 0 0
Timeout 1.5 1.5 0.3 0.4
Academic Recognition 14.5 5.1 6.3 35.6
Days Observed 8 6 6 10
Experimental Conditions
Table 3: Teacher Behavior -- Averages for Experimental Conditions (Frequency per 20-min Observation).
Experimental Conditions
C:\rsm\y520\readings\madsen68\madsen68.fm 11
DISCUSSION
Technical Considerations
39 The problems of gaining good data and maintaining adequate experimental control in an ongoing classroom in a
public school have not all been recognized as yet, much less solved. The greatest difficulty encountered was that
of maintaining stable control over some important variables while others were being changed. When these vari-
ables involve aspects of teacher behavior, the problem becomes one of helping the teacher maintain discrimina-
tive control over her own behavior. Daily feedback [begin page 149] from the experimenter, based on the
observer ratings. can help in this task (i.e., show the teacher the up-to-date graph of her behavior). Also, provid-
ing the teacher with a small counter to help monitor her own behavior can be helpful (Thomas, et al., 1968). Most
difficult to control in the present study was teachers Disapproving Reactions to Inappropriate Behaviors during
the Ignore Phase of the experiment. Teacher A became very upset as her classroom became worse. One solu-
tion to this problem might be a pre-study in which the teacher is trained in effective management techniques,
and then taken through a series of short periods where both Approval and Disapproval are eliminated and one or
the other reinstated. The teacher would then have confidence that she can effectively handle her class and be bet-
ter able to tolerate short periods of chaos (if such periods did occur). She would also have had sufficient training
in monitoring her own behavior to permit more effective control.
40 No attempt was made to program the frequency of various classes of Academic Recognition behaviors. Since
such behavior may be important in interpreting results, and was found to vary with some experimental condi-
tions, future work should strive to hold this behavior constant also.
41 The present study emphasized the importance of contingencies between student and teacher behaviors, but did
not measure them directly. While producing similar effects on two children in the same classroom and one child
in another classroom, and showing correlated changes in teacher behaviors (including a reversal operation),
more powerful data are potentially obtainable with a different technology. Video-tape recordings could enable
the use of present coding techniques to obtain contingency data on all classroom members over longer observa-
tion periods. Just as the children adapted to the presence of observers, a class could be adapted to the presence of
a TV camera man. Costs could be trimmed by saving only some sample tapes and reusing others after reliability
ratings are obtained. The current observation procedures (short of having an observer for each child) cannot
readily be extended to include simultaneous coding of teacher and child behavior without over-taxing the
observers. The present findings, and related studies in this series, are sufficiently promising to warrant an invest-
ment in more powerful recording equipment.
Behavior Classes
1
Baseline I Rules
Rules +
Ignore
Rules +
Ignore +
Praise I Baseline II
Rules +
Ignore +
Praise II
Inappropriate Behavior
2
46.8 39.8 68.5 20.5 37.6 15.1
Gross Motor 13.9 11.3 32.7 5.9 15.5 4.1
Object Noise 3.5 1.4 1.3 0.5 1.9 0.8
Disturbing Other's Property 3.3 1.8 1.9 0.7 0.7 0.3
Turning Around 21.6 9.9 11.4 9.1 12.8 7.6
Verbalizations 12 16.8 21.8 6.5 8 3.5
Other Inappropriate Behavior 10.9 7.8 16.5 3.9 7.8 2.6
Mouthing Objects 5.5 2.9 3.5 0.7 0.2 0.1
1 Contact occurred less than 1% of the time and is not tabulated here.
2 The sum of the separate problem behaviors will exceed that for Inappropriate Behavior, since the latter
measure does not reflect the possibility that more than one class of problem behavior may occur in an interval.
Experimental Conditions
Table 4: Percentage of intervals in which behaviors occur: averages for two children in classroom A by
experimental conditions.
C:\rsm\y520\readings\madsen68\madsen68.fm 12
Teacher Reactions
42 Teacher A. Initially, Mrs. A generally maintained control through scolding and loud critical comments. There
were frequent periods of chaos, which she handled by various threats.
43 When praise was finally added to the program, Mrs. A had these reactions: I was amazed at the difference the
procedure made in the atmosphere of the classroom and even my own personal feelings. I realized that in prais-
ing the well-behaved children and ignoring the bad, I was finding myself looking for the good in the children. It
was indeed rewarding to see the good rather than always criticizing. . . . I became convinced that a positive
approach to discipline was the answer.
44 Teacher B. During Baseline Mrs. B was dispensing a great deal of praise and approval to her classroom, but it was
not always contingent on Appropriate Behavior. Her timing was wrong and inconsistencies were apparent. For
example, on one occasion two children were fighting with scissors. The instigator was placed under a table away
from the rest of the class and left there for 3 min. After 3 min Mrs. B took the child in her arms and brought her
back to the group even though she was still emitting occasional loud screams. Mrs. B would also ignore behavior
for a period of time and then would revert to responding to Inappropriate Behavior with a negative comment;
she occasionally gave Approval for Inappropriate Behavior. The training given in seminar and discussions with
the experimenter led to an effective use of contingencies. Teacher B was also able to use this training to provide
instructions and training for her aide to eliminate problems which arose in the final phase of study when the aide
was continuing to respond to Disruptive Behaviors.
Changes in the Children
45 Cliff showed little change until Mrs. A started praising Appropriate Behavior, except to get worse during the
Ignore phase. He was often doing no academic work, talking to peers, and just fiddling away his time. It took
considerable effort by Mrs. A to catch Cliff showing praiseworthy behavior. As the use of praise continued, Cliff
worked harder on his assigned tasks, learned to ignore other children [begin page 150] who were misbehaving,
and would raise his hand to get teachers attention. He participated more in class discussions. He was moved up
to the fastest arithmetic group.
46 Frank showed little change in his hyperactive and inattentive behavIors until praise was introduced. Frank
responded rapidly to praise. After just two days in the praise phase, Frank was observed to clean his desk qui-
etly and quickly after completing a handwriting assignment. He was able to finish a task and study on his own
until the teacher initiated a new activity. He began to ask for extra assignments and volunteered to do things to
help his teacher. He had learned to sit quietly (when appropriate), to listen, and to raise his hand to participate in
class discussion, the latter occurring quite frequently.
47 Stan slowly improved after contingent praise was instituted, but some of the gains made by Mrs. B were in part
undone by the teacher aide. The aide was described as playing policeman and it took special efforts by the teacher
to get her to follow the program. Mrs. B summarized the changes in Stan as follows: Stan has changed from a
sullen, morose, muttering, angry individual into a boy whose smile seems to cover his whole face. He became
very responsive to teacher praise and learned to follow classroom rules, to pay attention to teacher-directed activ-
ities for long periods of time, and to interact with his peers in a more friendly way.
Implications
48 This replication and refinement of an earlier study by Becker, et al. (1967) adds further confidence to the asser-
tion that teachers can be taught systematic procedures and can use them to gain more effective behaviors from
their students. Unless teachers are effective in getting children ready to learn, their technical teaching skills are
likely to be wasted. Knowledge of differential social reinforcement procedures, as well as other behavioral princi-
ples, can greatly enhance teachers enjoyment of the profession and their contribution to effective development of
the students.
49 The reader should note that while we formally recorded the behavior of a few target children, teacher and
observer comments indicated dramatic changes in the whole atmosphere of the classroom and in the teachers
C:\rsm\y520\readings\madsen68\madsen68.fm 13
enjoyment of their classes.
REFERENCES
Allen, K. E., Hart, B. M., Buell, J. S., Harris, F. R., and Wolf, M. M. Effects of social reinforcement on isolate behavior of
a nursery school child. In L. P. Ullmann and L. Krasner (Eds.), Case studies in behavior modification. New York:
Holt, Rinehart, & Winston, 1965. pp. 307-312.
Becker, W. C. Madsen, C. H., Jr., Arnold, Carole R., and Thomas, D. R. The contingent use of teacher attention and
praise in reducing classroom behavior problems. Journal of Special Education, 1967, 1, 287-307.
Bijou, S. W. and Baer, D. M. Some methodological contributions from a functional analysis of child development. In L.
P. Lipsitt and C. S. Spiker (Eds.), Advances in child development and behavior. New York: Academic Press, 1963,
pp. 197-231.
Hart, Betty M., Reynolds, Nancy J., Baer, DOnald M., Brawler, Eleanor R., and Harris, Florence R. Effect of contingent
and non-contingent social reinforcement on the cooperative play of a preschool child. Journal of Applied
Behavior Analysis, 1968, 1, 73-76.
Thomas, D. R., Becker, W. C., and Armstrong, Marianne. Production and elimination of disruptive classroom behavior
by systematically varying teachers behavior. Journal of Applied Behavior Analysis, 1968, 1, 35-45.
Madsen, C. H., Jr., Becker, W. C., Thomas, D. R., Koser, Linda, and Plager, Elaine. An analysis of the reinforcing func-
tion of Sit Down Commands. In Parker, R. K. (Ed.), Readings in educational psychology. Boston: Allyn and
Bacon (in press).
Received 25 February 1968.