ABSTRACT

AN OBSERVATIONAL STUDY OF MOTHERS'
COMMANDS AND THEIR YOUNG
CHILDREN'S RESPONSES

BY

Margaret M. Bailey

Thehpurpose of the present study was to observe
and document a three-step sequence of interaction between
mothers and young children. The sequence began with a
mother's command and included the child's compliance or
noncompliance and the mother's response to that behavior.
The subjects were 15 mothers with their firstborn children
who were 13% to 14% months of age at the beginning of the
study. Each mother-child pair was observed in their own
home six times, for 40 minutes each observation, over a
two to three week period.

Most mothers gave commands fairly frequently, one
every two minutes on the average. Of the commands, most
were verbal prescriptions with one or two open slots or
possibilities for noncompliance. The mothers usually did
not repeat their commands. The mothers were individually
consistent and different from one another in their rates of

issuing commands. They were also individually consistent

Margaret M. Bailey

in their use of the two types of commands, prescriptions
and proscriptions.

Mothers' issuance of commands correlated signifi-
cantly with several characteristics of the children. Their
total commands and particularly their prescriptions were
more frequent with children who had higher levels of
activity. Mothers' proscriptions were more frequent with
children who had lower levels of persistence and with
those who understood less complex sentences. Mothers'
total commands were also more frequent in homes where there
was less space available to the child. Mothers' prescrip-
tions were more frequent in homes where there were more
forbidden objects within the child's reach.

IJ These correlations with mothers' commands were
interpreted to indicate that the mothers in large part
reacted to their children. They responded with different
degrees of "commanding-mess" to their children's tempera-
ment and language comprehension, and in part, to the space
available. However, the mothers' characteristics, par-
ticularly a hypothesized variable labeled "directiveness"
and the mothers' willingness to accommodate to their Chil-
dren, also may have had an impact on their use of commands.

In response to allof their mothers' commands,
most children complied about half of the time, with small
but significant consistent individual differences.

Children's compliance correlated with only two variables,

Margaret M. Bailey

negatively with mothers' warmth and positively with
mothers' use of punishment following noncompliance. A
microanalysis of the results which examined children's
compliance as a consequence of their mothers' responses to

their previous behavior did not prove significant. Thesem‘

 

results indicate that children's compliance is affected

/ ‘“

"V by more global rather than minute aspects of the inter-

action. For example, most mothers punished their children's

~”noncompliance only one-quarter of the time, on the average.
However, individual mothers had differing low rates of
punishment which they showed consistently. Punishment is
seen as a potent enough response that the consistent,
occasional use of it "encouraged" compliance.

The negative correlation between mothers' warmth

and children's compliance was interpreted as indicating

\
2 3“

that warm mothers confused the contingencies for compliance.
The warmer mothers provided affection and praise regardless
of their children's responses to commands.
/:1 These findings generally support the thesis that
~- ..... .- ”##1##!”
children do not naturally comply but must have a motive to.
do so. Further, the children did not exhibit a trait of

compliance which spanned both the observations and their

behavior in the language comprehension session. These

ha—u——-——~o..._.

results were considered in light of the relevant literature

on mothers' commands and children's compliance.

AN OBSERVATIONAL STUDY OF MOTHERS'
COMMANDS AND THEIR YOUNG

CHILDREN'S RESPONSES

BY

Margaret M. Bailey

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Psychology

1977

Q; 1 o“; a: ““I 0

AC KNOWLE DGEMENTS

There are many people who contributed to this work.
I especially thank the mothers who participated in the pre-
testing and in the study itself. The Michigan Early
Childhood Center allowed me to pretest parts of the method
by observing children at their day care center. I also
thank the members of my committee and in particular
Dr. Ellen Strommen who served as chairperson. She was
instrumental in my initial formulation of the problem and
helped in each step of the process. Each of the other
committee members, Dr. Helen Benedict, Dr. Charles Hanley,
Dr. Martha Karson and Dr. Albert Rabin, contributed in
unique and important ways. Dr. John Paynter helped
immeasurably in my attempts to understand Rousseau and
provided useful comments on portions of an earlier draft.
Ms. Sheri Traison proved to be a reliable and thoughtful
assistant. Finally, the Department of Psychology provided
funds which allowed me to compensate partially the mothers

who participated in the study. Thank you all.

ii

TABLE OF CONTENTS

LIST OF TABLES. . . . . . . .

LIST OF APPENDICES. . . . . .

INTRODUCTION 0 O O O O O O O O

Ethological Point of View .
Psychoanalytic Point of View.
Rousseau's Point of View.

Child's Temperament as a Variable
Empirical Questions of Interest .

Methodological Issues .

METHOD. . . . . . . . . . . .

Subjects. . . . . . . .
General Procedure . . .
Observational Method. .
Mothers' Speech . . . .

Sequence Following a Command.
Additional Behaviors Recorded
Child Commanding the Mother. . . . . . .

Warmth . . . . . . .
Hostility. O I 0 O 0

Ratings of the Environment.

Rating the Child's Temperament. . . . . .
Rating the Child's Language Comprehension

Data Reduction. . . . .
Mothers' Commands. .

Children's Compliance.
Mothers' Responses to Children's
Compliance and Noncompliance . . . . .
Children's Temperament .
Children's Language Comprehension. . . .

Mothers' Responses to Children's Commands

on Occurrence

Mothers' Warmth and Hostility. . . . . .

Environment. . . . .

Statistical Analysis .

iii

0

Page

89

RESULTS 0 C O O O O O O O O O O O O O O O O O O O O O

Mothers' Commands.
Consistency .
Frequency . .
Repetitions .

Form. C O 0

Complexity. .
Content . . .
Reflections . .
Mothers' Commands Correlated with Other

Variables . . . . . . .
Sex of the Child. . . .
Temperament of the Child. . . . . . .
Children's Language Comprehension . .
Environment . . . . . . . . . . . . .
Summary: Mothers' Commands and Their
Correlates. . . . . . . . . . . . . . .
Children's Compliance and Noncompliance. .
Consistency . . . . . . . . . . . . . .
Frequency . . . . . . . . . . . .
Children's Compliance Correlated with Other
Variables . . . . . . . . . . . . . . .
Sex of the Child. . . . . . . . . . .
Temperament of the Child. . . . . . .
Children's Language Comprehension . .
Mothers' Warmth and Hostility . . . .
Mothers' Compliance to Children's Commands
Mothers' Issuance of Commands . . . . . .
Mothers' Responses to Children's Compliance
and Noncompliance . . . . . . . . . . . . .
Mothers' Responses Correlated with Children's
Compliance. . . . . . . . . . . . . . . . .
Microanalysis of Mothers' Responses . . . .

DISCUSSION . C C C O C O O O O O O O O O I O O O O 0

LIST OF REFERENCES 0 O O O O O O O O O O O O O O O 0

iv

Page
42

43
43
44
44
45
46

48

49
49
49
52
54

56
57
57
58

59
S9
59
59
60
62
62

66
68

71

121

LIST OF TABLES

Table Page
1 Mothers' Issuance of Commands. . . . . . . . . . 44
2 Mothers' Repetitions of Commands . . . . . . . . 4S

3 Mothers' Issuance of Commands: Prescriptions
and Proscriptions Separated. . . . . . . . . . 47

4 Mothers' Issuance of Reflections . . . . . . . . 48
5 Children's Temperament Scores. . . . . . . . . . 50

6 Children's Temperament and Language Compre-
hension Correlated with Mothers' Commands. . . 51

7 Distribution of Children's Language Compre-
hension Scores . . . . . . . . . . . . . . . . 52

8 Children's Environments. . . . . . . . . . . . . 54

9 Environment Variables Correlated with Mothers'
comand S O O O O O O O O O O O O O O O O O O O 5 5

10 Children's Compliance Ratios to all Commands . . 58

11 Mothers' Total Expressions of Warmth and
Hostility. O O O O C O O O O I O O O C C O O 0 6o

12 Mothers' Responses to Children's Compliance
and Noncompliance. . . . . . . . . . . . . . . 64

13 Mothers' Particular Responses as a Percentage
of All Responses . . . . . . . . . . . . . . . 65

A Potential Subjects' Reasons for Not
PartiCipating. O O O O C C O O O O O O O O O O 92

B Children's Scores on Carey Infant Temperament
Study. 0 O O C O O O O O O O O O O O O O O O O 108

Table

C Mothers' Use of Commands: Individual Data. . .

D Mothers' Use of Prescriptions and Proscrip-
tions: Individual Data. . . . . . . . . . .

E Children's Individual Scores on Temperament
variables. 0 O O O O C O O I O O O O O O O O

F Children's Compliance Ratios: Individual Data.

vi

Page

111

114

118
120

LIST OF APPENDICES

Appendix

A

Phone Interview and Number of Potential

Subjects Contacted . . . . . . .
Sample Observation Recording Form.

Reliability of Measures. . . . . .

Code for Categories of the Mothers'

Glossary of Terms. . . . . . . . .

Code for the Mothers' Responses to Children's

Compliance or Noncompliance. . .

Speech .

Checklist for Rating the Environment . . . .

Carey Infant Temperament Survey:
and critique O O O O O O O O O I

Mothers' Use of Commands: Individual Data

Results

Mothers' Use of Prescriptions and Proscrip-

tions: Individual Data. . . . .

Children's Individual Scores on Temperament

Variables. . . . . . . . . . . .

Children's Compliance Ratios: Individual

Data 0 O O O O O O O O O O O I 0

vii

Page

90
93
95
98
100

103
105

107

110

113

117

119

INTRODUCTION

This study is concerned with many variables which
may affect young children's compliance. As defined here,
"compliance" is obeying a parent's command (prescription
or prosciption). This behavior, in simplified form, is
one part of a three-step interaction. The interaction
begins with the parent giving a command, follows with the
child's response to the command (compliance or non-
compliance) and concludes with the adult's response to the
child's behavior.

This study is based on observations of this three-
step interaction between mothers and their young children
at home. It is an attempt to delineate the relative
importance of several variables which have an impact on
the child's compliance. Several theories are considered,
including an ethological model emphasizing that children
are naturally inclined to comply, and the psychoanalytic
model, emphasizing that compliance is a defense against
loss of the parents' love. An additional model proposes
that the child's temperament will be an important modifier
of this interaction. These theories, each attempting

to account for whether or not the child complies, are

explained in detail below. This review includes and is
expanded by Rousseau's position that the primary adults
in the environment determine first whether or not to
command compliance. If the adult chooses to issue a
command, then the child may or may not comply. Both
aspects of the interaction, the frequency of commands
from the adult, and the child's response once the command
is given, are of interest in this study.

Many authors have observed that children's com-
pliance is often important in parent-child relationships
before the child is two years old. (See for example
Tulkin and Kagan, 1972; MacFarlane et al., 1954; and
Spock, 1968.) To examine this behavior at an early phase,
the children studied are therefore young, approxi-
mately one year old. The variables which affect their
behavior may not be the same as those which affect older
children. Children this young differ from older children
particularly in their speech and language skills. Further,
with children this young, compliance is not a moral or
immoral act. In the theories described below, compliance
or noncompliance is, however, often viewed as a precursor

to moral development.

Ethological Point of View
Stayton, Hogan and Ainsworth (1971) recently
considered the role of mothers' attitudes in the compliant

behavior of year-old infants. They observed a sample of

25 mother-infant pairs at home. The mothers were rated
by two judges on three attitude scales: sensitivity-
insensitivity, acceptance-rejection, and cooperation-
interference. The mothers' behaviors which were coded
were frequency of verbal commands which were "comprehen-
sible" to the child, frequency of physical intervention,
and the extent of floor freedom allowed the child. The
primary infant variable was "compliance" to mother's
commands. The results of their study indicate in general
that the ratings of the mothers' attitudes showed the
highest (and only statistically significant) correlations
with the children's compliance. The mothers' attitudes
themselves exhibited extremely high intercorrelations,
from .87 to .91, indicating that they reflect a fairly
unitary scale. The authors interpret their findings as
indicating that children are naturally inclined to obey.
Further, this "disposition toward obediance emerges in a
responsive, accommodating social environment without
extensive training, discipline, or other massive attempts
to shape the infant's course of development."

There are several obvious problems with this study
and the authors' interpretations of their findings.
Basically, it is difficult to ascertain exactly what
attitude was scaled, and what behaviors underlay that
attitude. The judges were apparently rating only one

attitude, as evidenced by the high intercorrelations

among the three scales. However, they did not describe
specific behaviors which led them to rate as they did.
The codings of the mothers' behaviors which were included
are too global to specify the attitudes. The nature of
the commands is not specified. Apparently they included
both pre- and proscriptions. Most importantly, there is
no separation of commands which require new activity
from those which are merely reflections of the child's
ongoing activities. A reflection occurs if the child is
already engaged in an act, such as dropping a ball into
the mother's lap, and the mother says, "Give me the ball."
This type of command is very different in character from
saying "Give me the ball" when the child is playing with
blocks. In effect, while this study is very suggestive,
it is not possible to know which of the mothers' attitudes
were scaled or exactly how compliant the children were.
Stayton, Hogan and Ainsworth (1971) had their
observers include commands which they judged "comprehen—
sible" to the children. The authors apparently recognized
that the child's comprehension of the command is a neces-
sary requirement for the child's compliance. Their
attempt to include comprehension as a variable is admir-
able, although their method was unsatisfactory. (Currently
there are no standardized measures of young children's
language comprehension. This problem is addressed more

fully under "Methodological Issues.")

Despite the problems with this study, the findings
are worth considering. The authors' interpretation
reflects a growing point of view in the deve10pmental
literature. Hogan (1973) fleshes out the scheme of moral
development which includes the Stayton, Hogan and
Ainsworth (1971) interpretation of their findings. He
suggests that "socialization" is one dimension of morality.
The "disposition to comply" makes socialization and
morality possible without a fundamental struggle between
the child and society. The child is naturally disposed
to comply and needs no further motive.

Lytton (1973) also correlated children's com-
pliance behavior with ratings of mothers' "warmth"

(among other variables). The subjects were 30 sets of
mothers and sons, with the children two and one-half-
year-old boys. The sample included twins and triplets.
Lytton did not find a significant correlation between
mothers' "warmth" and the boys' compliance. Further, his
ratings of "warmth" did not correlate significantly with
two behavioral measures, frequency of expressed affection
and frequency of noncontingent positive contacts initiated
by the mother. That is, Lytton attempted unsuccessfully
to determine what raters were responding to when rating
"warmth." Compliance was also unrelated to either of

these behavioral measures.

Lytton used a "compliance ratio," defined as the
ratio of all instances of compliance to instances of
compliance plus non-compliance. He found that this
ratio was significantly related to two other behavior
counts. Compliance was positively related to the percent
of time the child was active and negatively related to
the frequency of the mother's negative contact with the
child. (This negative contact was not in immediate
response to the child's action, i.e., it was noncontingent
displeasure, hostility, or restriction of the child.) In
terms of behaviors, Lytton found children more likely
to be compliant if they were active and if their mothers
did not frequently initiate "negative" contact with them.

Lytton used other observer ratings, as opposed
to behavioral counts, of certain parental and child
characteristics. Typically, these ratings did not
correlate with the behavioral counts of the (supposedly)
same characteristics. For example, the rating of the
child's compliance and the compliance ratio (behavior
count) correlated nonsignificantly (r = .33). As noted,
mother's "warmth" was not related to theoretically similar
behaviors. These findings are important methodologically
because they indicate the questionable validity of ratings
of characteristics. (Lytton does not agree with this
interpretation. Indeed, he finds useful those ratings
which are intercorrelated, even when the ratings are not

related to behavior.)

At any rate, Lytton's method and findings are
interesting. They specifically counter Stayton, Hogan
and Ainsworth's (1971) findings and suggest that behavioral
characteristics of the child (activity) as well as of
the mother (negative contacts) are important.

A comment about Lytton's sample is in order. The
children were only one sex, were older than those studied
by Stayton, Hogan and Ainsworth (1971) and included
twins and triplets. The findings are interesting but
may not be immediately generalized to younger, single

children.

Psychoanalytic Point of View

Other authors have interpreted the phenomenon of
young children's compliance or noncompliance in radically
different ways. The traditional psychoanalytic position,
as expressed by Freud (1923/1960, 1933) and Fenichel
(1945) is in agreement with Stayton, Hogan and Ainsworth
(1971) in two regards. First, authors from both
positions agree that compliance in this age group is a
possibility. (However, Fenichel and Freud refer only
to compliance to mothers' prohibitions.) They further
agree that this behavior has implications for later moral
development but is not in itself a "moral" or "immoral"
act.

Aside from these points, the two schools have very

different assumptions about young children's compliance.

Basically, Freud (1923/1960, 1933) holds that compliance
is a defensive behavior used to avoid the loss of parental
love. The child may also comply to avoid physical
punishment but this motive is secondary. In support of
this argument, Fenichel (1945) points out that typically
children this age only obey in the presence of the parent.
That is, the psychoanalytic school holds that the child
complies to maintain love and prevent physical punishment.

Freud (1933) states that parental behaviors which
will foster compliance include "proofs of affection" and
"threats of punishment." Apparently parents' statements
and behaviors which link their affection for the child
with the child's compliance will promote compliance as
a defense. These statements and behaviors would probably
include: affection as a reward for compliance; withdrawal
of love and threats of same used as a punishment for
noncompliance; and secondarily, physical punishment or
threats of same for noncompliance.

Recent empirical studies on withdrawal of love
as a punishment technique with older children indicate
that it may be important in impulse control. Impulse
control and resistance to temptation, usually measured
in the parent's absence, function like an internalized
form of overt compliance. These studies are therefore
tangentially related to the major topic at hand. Two

studies (Hoffman, 1963, observing nursery school children;

and Hoffman and Saltzstein, 1967, studying 7th grade
children) found a negative correlation between parents'
use of withdrawal of love and the child's expression of
overt hostility away from home. These results do not
agree with those found by Sears, Rau and Alpert (1965).
These authors correlated parent's use of withdrawal of
love with their four- and five-year old children's
aggression at home and at nursery school. Withdrawal of
love (in this case, physical isolation) appeared unrelated
to girls' aggression in either setting. For boys, use

of isolation was negatively correlated with aggression

at home and positively correlated with aggression at
nursery school. (This study appears to be the only one
directly relating withdrawal of love to overt compliance.)
Finally, Burton, Maccoby and Allinsmith (1961) found that
mothers' use of withdrawal of love correlated positively
with four-year-old boys' resistance to temptation. The
relationship did not hold for girls.

To summarize, Freud's postulation that parents'
use of withdrawal of love techniques will lead to com-
pliance in the child has not been studied with very young
children. The results with nursery school aged children
are equivocal.

Freud's postulation that physical punishment may
lead to increased compliance does not receive support

from a direct empirical study of children slightly older

10

than those in this study. Minton, Kagan and Levine
(1971) observed 90 children who were 27 months of age,

in their own homes for a total of five hours of observa-
tion per child. They found that the children's rate of
violations of established rules was positively correlated
with parents' use of physical punishment and negatively
correlated with explaining the reprimands which followed
the child's violations. These correlations were signifi-
cant for both sexes. (The form of punishment was not
related to the child's compliance to mothers' prescrip-
tions, i.e., to commands that the child perform a desired
act.)

Sears, Maccoby and Levin (1957) reported that
approximately 50 percent of the 379 mothers of kinder-
garteners whom they interviewed felt that physical
punishment was "effective." (There was no uniform
definition of "effective.") The mothers were also divided
by the interviewer's rating of them as "warm" or "cold."
Of those mothers who were rated as "warm" and who reported
using physical punishment frequently, 66 percent felt
that punishment was effective. This finding contrasts
with relatively cold mothers who also punished frequently,
of whom 43 percent felt physical punishment was effective.
"Warmth" did not appear to be a modifier of effectiveness
when physical punishment was rarely used. These results

suggest very tentatively that mothers' warmth may be a

ll

variable in the effectiveness of physical punishment.
Again, the results are only suggestive because the children
were older than those of concern here.

It is possible that Freud was referring to fairly
severe physical punishment when he proposed that it
might lead to compliance. Parents who employ extremes
of punishment are not likely to participate in studies

such as those reported above.

Rousseau's Point of View

 

Rousseau's treatise on education, the Emile
(1762/1972) restates the question about compliance in an
interesting way. Instead of asking what variables affect
whether the child complies, Rousseau asks whether com—
pliance is a necessary issue. Up to this point, we have
considered the continuum ranging from child's noncompliance
to compliance. Rousseau stretches the perspective to
include the continuum, adult's making an issue of com-
pliance to not making an issue of it. If an adult makes
an issue of compliance (i.e., utters a command), the

child may then comply or not comply. Schematically:

ADULT compliance not an issue-+ + compliance an issue

CHILD overt compliance ++ overt noncompliance

According to Rousseau, if the young child
complies, it is out of self-interest. The child may be

either fearful of punishment or desirous of rewards.

12

This view of the child is similar to Freud's, in that
both authors emphasize that adults must provide the

child with a powerful reason to comply. That is, both
authors disagree with the thesis of a "disposition to
comply." The major difference between Rousseau and Freud
is that Rousseau does not consider loss of love as a basic,
natural motive. In his view, the child's motives are
fear of punishment or desire for concrete rewards, such
as food. (This difference is rooted in their views of
child development. Regarding young children, however,
they explain behavior in much the same terms.)

Rousseau's understanding of young children's
compliance is not an endorsement of noncompliance. In
fact, Rousseau holds that habits of compliance or non-
compliance each harbor dangers for the child's later
deveIOpment. If the parent makes an issue of the child's
compliance, the primary danger is that the child will
become rebellious. When an infant or young child exper-
iences an adult's "will" thwarting their own will, the
first reaction is anger. Noncompliance or rebelliousness
is an expression of this anger. Ultimately, such an
attitude prevents genuine happiness.

Habits of compliance are equally dangerous.

They are the behavioral manifestations of the child's
scheming and maneuvering from an early age. Rousseau

holds that a person with such an attitude is also unhappy.

13

According to Rousseau's developmental theory, a
sense of "will," or a sense of self and one's own ability
to produce certain changes, emerges in infancy. With this
understanding of causality, the child can also understand
that other persons can effect changes. As explained
above, should the child experience another's "will"
thwarting her/his own desires, the result is rebellious-
ness. Overt compliance to another's will is possible,
but is not the first reaction. The developmental ground-
work for habits of noncompliance or compliance is laid
early in life. Rousseau is not specific about age, but
clearly states that the preverbal child is capable of
these reactions.

Given that both compliance and noncompliance are
undesirable habits in the young child, Rousseau proposes
a scheme of child-rearing in which compliance is not an
issue during infancy and early childhood. In general,
the adult achieves this by immediately meeting the
child's "natural" desires, and by never meeting "capricious"
desires. "Natural" desires include food, clothing and
shelter. "Capricious" desires are of two kinds. They
include the desire to be served as well as the desire for
particular objects which the child does not "need." In
fact, these two forms of capricious desires are usually
linked, in that children rely on an adult to provide

them with the objects which they don't actually need.

14

Rousseau points out two ways to limit capricious
desires. The basic way is to plan the child's environment
so that the child's fantasies are not taunted. The adult
is to provide the child with space to explore, without
encountering forbidden objects. Secondly, the adult must
be rigidly consistent (a "wall of brass") about which of
the child's desires she/he meets. In this way, the child
will not be tempted to manipulate the adult. The adult
is to immediately meet the child's "natural" desires,
but should the child desire a "capricious" object, the
adult is to respond firmly and consistently "no."

These two methods, structuring the environment and
responding consistently, should produce an environment
where the issue of compliance rarely arises.

Rousseau's theory receives modest empirical
support from Crandall et a1. (1958) in a study of
children older than those of direct concern in this study.
The subjects in this study were aged three to eight years,
and were observed at nursery schools as well as at home
with their mothers. For the youngest children (three to
five years) compliance was significantly negatively
correlated with attempts to direct the behaviors of adults,
and with aggressive behaviors aimed at adults. Young
children's compliance was not correlated with their
frequency of help-seeking or approval-seeking from adults.

The significant variables suggest that noncompliance is

15

associated with the "battle of wills," expressed in the
children's attempts to control adults and their hostility
toward adults. Crandall et a1. (1958) interpret their
results as indicating that compliance toward adults is

one result of a "general dominance-submission picture."

Child's Temperament as a Variable

In the discussion of Lytton's (1973) work earlier,
the importance of the child's activity level was noted.
However, none of the major theories considers character-
istics of the child as key variables in compliance or
noncompliance. While Thomas, Chess and Birch (1968)
did not consider compliance as a dependent variable,
their studies of temperament have implications here.

Thomas et a1. (1960) report the methodology of
their longitudinal study of infant temperament. Basically,
the authors relied on interviews with the parents of 130
children, using a loose interview schedule related to the
details of daily living for the child. There were nine
categories or temperamental variables: activity, per-
sistence, rhythmicity, approach, adaptability, intensity,
threshold, quality and distractibility. Each statement
made by the parents about the child was scored if relevant
to any or more than one of the variables. Therefore, the
number of scored items varied among parents and even

across interviews with the same parent.

16

Thomas, Chess and Birch (1968) report on the 42
children who comprise the clinical subsample of the
original sample. These 42 children developed a "signifi-
cant degree of behavioral disturbance," as judged by
interviews, play sessions and parental report. All of
these children were at least two years old when the
diagnoses was made. The 42 children had symptoms which
were generally divided into two groups, active and passive.
Of the children with active symptoms, 16 had noncompliance
included as part of the presenting complaint. (No
children with passive symptoms had noncompliance included
in their descriptions.) The authors then posed the
statistical question of whether the development of
active symptoms was related to early temperament. They
found that high activity level and increased persistence
in the first year of life were related to the deve10pment
of these active behavior disorders, including noncom-
pliance. These results are only suggestive but they do
indicate that temperament may be an important factor in
noncompliance.

Each of the points of view discussed above
addresses either maternal variables or child variables
as central to children's compliance or non-compliance.
Escalona (1968) points out that this approach is the
usual one in developmental theory, that is, to relate

either environmental or organismic variables to

17

developmental outcomes. She suggests as an alternative
approach that the interaction between these types of
variables is the key to outcome. That is, the child's
experience is the product of the interaction of both
types of variables. In terms of young children's com-
pliance, the key is the match between the environment
provided by the mother and the child's temperament.

As illustrated, children's compliant behavior has
been interpreted in many ways. However, much of the basic
research on this issue has not been completed, especially
with young children. As noted repeatedly, many of the
studies cited above used children significantly older
than one year. We do not yet have a detailed analysis
of mothers' actual behaviors, as different from ratings
of their character, or an analysis of their commands,
as separate from their reflections of their children's
acts. Lacking these, we do not have an accurate measure
of the child's compliance. Indeed, "compliance" may not
be a trait in young children, but a behavior exhibited
only in fairly particular circumstances.

The most promising method appears to be a micro-
analysis, that is, a detailed analysis within pairs.

The aim of this study is to use such a method to observe
and analyze the behavior of mothers and year—old children.
Stayton, Hogan and Ainsworth (1971) represent an attempt
in this direction, with further refinement of the method

attempted here.

18

Empirical Questions of Interest

The empirical questions of interest include two
continua. First, there is the question of the degree to
which compliance is an issue. Rousseau hypothesizes
that the adult can minimize demands for compliance by
carefully structuring the child's environment. To test
this idea, the frequency of commands (as different from
reflections) indicates the degree to which compliance is
an important issue in a particular mother—child pair. If
Rousseau is accurate, minimal demands will be associated
with those mothers who provide space for the child to
explore, with a limited number of forbidden objects which
the child can touch, and who consistently refuse to
respond to the child's commands for service or for
capricious objects. Although not considered by Rousseau,
it seems plausible that the child's temperament (as
studied by Thomas, Chess and Birch, 1968) affects the
existence of this continuum.

‘ The continuum more frequently studied is the
degree to which the child complies with the adult's
commands. Stayton, Hogan and Ainsworth (1971) would
predict that the most important variables in compliance
would be those maternal behaviors which indicate the
mother's willingness to respond to the child and cooperate
with the child. That is, they would predict increased

compliance with mothers who comply with infant requests

19

and who do not interfere with ongoing behaviors. The
psychoanalytic school would predict greater compliance
toward mothers who use affection as a reward and with-
drawal of love as a punishment, related to compliance or
noncompliance. Secondarily, the occurrence of physical
punishment would promote compliance. Rousseau (1762/1972)
and Crandall (1958) would predict greater noncompliance
among children who make requests of their mothers, where
little floor space is available and where there are many
"capricious" objects to desire. They would also predict
noncompliance among children who express hostility

toward their mothers. Rousseau would predict compliance
toward parents who use physical punishment and/or who
promise concrete rewards. Thomas, Chess, and Birch (1973)
would argue that infant temperament will be an important
modifier of the occurrence of compliance.

The empirical question is which variables are
most salient in relation to the compliance which occurs.
The one clear disagreement in these hypotheses is between
Stayton et a1. (1971) and Rousseau (1762/1972) on the
outcome of mother's responsiveness to the child's commands.
Testing Escalona's point of view requires studying the
individual mother-child dyads in more detail. An evalua-
tion of child temperament as matched with environment

provided by the mother must be correlated with compliance.

20

An alternative way of stating these questions is
to separate those which may be studied within each mother-
child pair from those which require a comparison among
pairs for evaluation. The basic data on frequency of
the mother's commands and the degree to which the child
complies are available and informative as a within-pair
analysis. Specifically, an analysis within pairs helps
resolve the question of which factors affect whether or
not a particular child complies. The other continuum
of interest, behaviorally defined as the frequency
of the mothers's commands, also requires an analysis
within pairs as well as among several pairs. An analysis
among pairs also makes possible a comparison of the
environments provided for the children, and of the
children's temperament. In effect, both forms of analysis
reveal useful information. Because this research is in
a preliminary stage, a thorough analysis within pairs

must be included.

Methodological Issues
Important methodological questions are raised by
the attempt to study young children's compliance. In
general, these questions may be separated into those
which involve the general observational procedures and

those which involve particular variables and their measure-

ment.

21

In this study the general procedure was to observe
mother-child pairs on six occasions in their own homes
during relatively unstructured times for approximately
40 minutes per observation. Further, the observer was
to follow the mother, so that if she left the child,
the observer also left.

The decision to observe in the home was based
primarily on the experimenter's desire to see the usual
transactions between mother and child. There is empirical
evidence that young children's behavior is altered in
relevant ways by the laboratory situation. Brookhart
and Hock (1976) recently reported that 10- and lZ-month-
old children exhibited more proximity-avoiding behavior
of both mother and a stranger in the home than in the
laboratory. The authors interpret the children's behavior
as a gesture of independence in the home. This interpre-
tation is supported by Wenar's (1972) findings that one-
year-old children observed in the home typically Spent
more time involved with the physical environment than
socializing with their mothers. The findings are relevant
because the question of compliance is more likely to
arise as the child engages in these more "independent"
behaviors away from the mother.

Brookhart and Hock (1976) also report that as the
observations and stranger presence progressed over three

sessions, the children's contact-maintaining behavior

22

lessened in the home but not in the laboratory. That

is, the children appeared to "adjust" to stranger presence
in the home. These findings suggest that observing in

the home should be preceded by visits which allow the
child to become accustomed to the observed. In line with
this finding, the first two of the six visits were con-
sidered "preliminary" sessions.

The data used in the final analysis were gathered
from four sessions per pair. Pretesting of the observa-
tion method indicated that there is a great deal of
variability within mother-child pairs. For example,
during one session Mother A gave commands at the rate of
one every three minutes. During the next session, she
gave commands at the rate of one every ten minutes.
Gathering data on four different occasions seemed neces-
sary to explore and record the variability within pairs.

Additional procedural questions were answered
pragmatically. The mother was the parent observed because
she is usually more available, and is usually the primary
caretaker of children this age. Also, the data are then
comparable to the other studies which have observed the
mother and child. The time per observation was limited
by the tolerance of all parties, especially the mothers.
The observer followed the mother instead of the child
because pretesting indicated that some children became

alarmed when left with a stranger.

23

Most of the variables and their measurements are
straightforward, and are presented in the "Method"
section. Problems are raised by the questions of how to
determine what constitutes a command from the mother, how
to determine whether the child could understand the
command, and how to select a sample.

The key methodological question is specifying
what constitutes a command. The most direct answer to
this question is to code as commands what appear to be
commands in intent. Searle (1975) points out that
speakers often make indirect requests or commands because
they assume the context will clarify their point, and
out of politeness. Thus, a person saying, "Can you pass
the salt?" is most often not asking whether the hearer is
able to pass the salt, but is asking that the bearer do
so. Holzman (1974) observing the mothers of four children
aged 15-27 months, reports that all mothers used both
direct and indirect commands. Methodologically, Searle's
(1975) and Holzman's (1974) results emphasize the
importance of including as commands both those that are
directly stated ("Pass the salt") and those that are
indirectly stated ("Can you pass the salt?").

An additional issue about the nature of commands
raised earlier is the importance of separating commands
from reflections. In the observational procedure used

here, this differentiation was accomplished by noting the

24

context at the time the command statement was spoken.
Using an earlier example, if the mother says "Give me

the ball,“ as the child is dropping the ball into her
lap, there are no possibilities for noncompliance on the
child's part. If, instead, the child is playing with
blocks, there are three "open slots" or possibilities for
noncompliance: the child must use the appropriate

object (ball), take it to the appropriate recipient (the
mother) and perform the act (give). By noting the context,
the observer could record which slots were "open," that
is, slots where the child had a choice, and could also
analyze the child's response in the same detail. The
total sequence includes the mother's command, the child's
response, and the mother's response to the child's
behavior.

As noted in the discussion of Stayton, Hogan and
Ainsworth's (1971) research, an additional issue is what
speech the child understands. The Stayton, Hogan and
Ainsworth (1971) solution to this problem was to decide
while coding whether the command was intelligible, and to
omit those that were not. The method used here included
all commands which the mother intended as such, whether
the experimenter believed them intelligible or not. This
design is preferable, given that the aim of the study was

to delineate important variables in the usual, total

interaction.

25

Benedict (1976, 1977) is developing a measure of
young children's language comprehension. In this study,
an additional experimenter administered her measure to
the children in the sample. These independent results
provide a base with which to compare the children's
compliance. That is, use of Benedict's measure allows
for a correlation of compliance with comprehension.

The second important issue about particular
variables involves how to select the sample. The problem
is how to account for important sample variables, and
still keep the sample small enough to allow for the
detailed analysis that seems necessary. The research
reviewed below indicates that two sample variables may
have an impact on the behaviors of interest here, the
sex of the child and the socioeconomic status of the
child's family.

Goldberg and Lewis (1969) and Messer and Lewis
(1972) found sex differences in the play behaviors of year-
old children. In general, the differences involved a
higher frequency of attachment behaviors (proximity-
seeking and touching) by the girls. Also of interest is
one aspect of play behavior, namely, that the boys more
often played with "nontoys," such as doorknobs. Either
or both of these behaviors may lead boys to more encounters
with their mothers over the question of compliance.

Certainly, sex differences in this area are worth observing.

26

Socioeconomic status may also have an impact on
mothers' speech and on the environments they provide.
Tulkin and Kagan (1972) observed 30 middle class and 26
working class mothers with their ten-month old daughters.
All subjects were Caucasians. They report that the middle
class mothers exhibited every type of verbal behavior,
including prohibitions, more frequently than the working
class mothers. Perhaps middle class mothers assume their
children (or at least their daughters) comprehend more
than working class mothers assume about their daughters.
In turn, middle class mothers may make more frequent
genuine commands of the children. An alternative explana-
tion is that the middle class mothers may have been using
reflections of their children's actions as a language
teaching device.

Tulkin and Kagan (1972) also reported that the
working class children in their sample had fewer oppor-
tunities to explore and fewer objects to manipulate.
Holzman (1974) did not find differences in physical
stimulation available to children in her sample according
to social class. The sample was small, including two
children from each of two social classes. These con-
flicting results leave open the question of the impact
of social class on physical environment.

Clearly there are questions about the impact of

socioeconomic status on variables of interest in this study.

27

However, an analysis according to both sex and socio-
economic would require a very large sample. Sex of the
child remains the more interesting and compelling
variable to study. Therefore, the sample in this study
was selected to provide an equal number of boys and girls.
The families were as similar as possible, in terms of

socioeconomic status.

METHOD

Subjects

The subjects were 15 mother-child pairs, located
primarily through birth records. One of the first sub-
jects also provided names of two other mothers who par-
ticipated. The children, seven girls and eight boys,
were 13% to 14% months of age at the first observation.
All had walked for at least two weeks. (The children
all had at least minimal abilities to separate from their
mothers and engage the environment.) They were the only
children in their families and lived with both parents,
with their mothers as their primary caretakers.

The families were selected to be as similar in
socioeconomic status as possible. In fact all were lower
middle to upper middle class. In every family the husband-
father was a white collar worker with some post-high
school education. All of the mothers had also graduated

from high school; some had college and advanced degrees.

General Procedure
The primary observer first contacted potential
subjects by phone. If the person met the criteria and

was willing to participate, they set up the observation

28

29

appointments. At the first observation session the
observer gave the mother a written sheet with all the
information from the phone interview. (The phone inter-
view and the number of calls necessary to obtain the
complete sample are detailed in Appendix A.)

The children and their mothers were observed in
their own homes six times, for 40 minutes each visit.
All observations were conducted during a time which the
mother identified as "free play," i.e., a non-structured
time of day. The first two visits served to accustom
the mother and the child to the observer. The behavioral
data were gathered from the final four visits. A second
observer went on two visits, at least one from the final
four. During one observation, she rated the child's tem-
perament (levels of activity and persistence) and the
environmental variables, as explained in detail below.
During her other observation she and the primary observer
alternated roles in terms of which data each recorded.

Each mother-child pair was observed over a two
to three week period to lessen deveIOpmental changes
within the observation period and to minimize the subject
drop-out rate. (The actual range of days from first
through sixth visit was 11 to 18 days.) At the end of
the sixth observation the primary observer interviewed
each mother. The interview included validation of the

ratings of the environment and administration of the

30

activity and persistence scales from the Carey Infant
Temperament Survey. Finally the observer gave each
mother $5.00 for her help.

After the six observations a third experimenter
visited each pair one time to administer Benedict's
measure of language comprehension. She conducted this
visit within one month after the basic data collection

was completed.

Observational Method

The behavioral data were gathered through a con-
tinuous observational record. The primary focus was the
mother's speech. Each statement and the person to whom
it was addressed were coded according to the speaker's
intent. Commands and the ensuing sequence of interaction
were recorded in detail and analyzed as explained below.
Additional variables were recorded whenever they occurred
as indicated on the sample recording sheet, Appendix B.

The primary and secondary observers pretested and
obtained a measure of the reliability for this observa-

tional record. These results are in Appendix C.

Mothers' Speech
The observers coded the mothers' speech according
to the intent of the speaker. They also indicated the per-
son to whom the statement was addressed. The five types

of statements which related to compliance and noncompliance

31

were commands, withdrawal of love, praise, threats and
promises. All of these statements were recorded verbatim.
Nonverbal commands were also included in this recording and
analysis. The speech categories and codes are in Appendix
D. Appendix E is a glossary of operational definitions of
mothers' speech categories and other behaviors referred to
throughout this report.

A command began a sequence of recorded interaction.
First the observer noted the context of the command and
specified the command's "open slots" or possibilities for
noncompliance. The observer also recorded all repetitions

of each command, analyzed in the same way.

Sequence Following a Command

The child could comply or not with each open slot
of a command. The child's response for each open slot was
therefore recorded. The next step in the interaction was
the mother's response to the child's response to the com-
mand. The mother's possible reactions to compliance
included praise, concrete reward, reflecting the act, issu-
ing another command, doublebinding the child or not respond-
ing. In "reflecting" the act, the mother simply stated the
child's compliant act, as in "you brought me the towel."
A "doublebind" occurred when the mother promised the child
some form of reward for compliance. However, after the
child complied the mother did not carry through on her

promise. The mother's possible reactions to noncompliance

32

included no reaction, performing the act for the child,
physically directing the child's movements to force
compliance, physical punishment, physical isolation of the
child such as putting the child in a corner, and withdrawal
of love statements. A "withdrawal of love" statement was
one in which the mother indicated that she not only dis—
approved of the child's act but was withholding affection

as a result.

The observers' codes for these and other behaviors

are in Appendix F.
Additional Behaviors Recorded on

Occurrence

Child Commanding the
Mother

These sequences included the child's command and
the mother's response. The operational definitions of
the child's commands included any instances of the child
crying, pointing and vocalizing, or using words to
indicate desires. These behaviors and the mother's
response of compliance or noncompliance were recorded.
Mother's compliance was defined as the mother behaviorally
meeting the request. Mother's noncompliance included
overt statements of noncompliance, attempts to distract
the child from the commands, commanding the child in
response to the child's command, speaking as if they
would complybut not carrying through, and finally, simply

not responding.

33

Warmth

Mothers' expressions of warmth were defined as
instances of giving the child unconditional praise or
physical affection. Physical affection was defined as
the mother stroking, patting, kissing or hugging the
child, not in response to the child's compliance or non—
compliance. Unconditional praise was defined as a
positive evaluative statement, again not in response to

the child's compliance or noncompliance.

Hostility

The mothers' unconditional negative statements
or acts comprised this category. They were unconditional
in the sense that they were not in response to the

child's compliance or noncompliance.

Ratings of the Environment

The primary and secondary observers each
rated the environment on different days. They
recorded two aspects, the space available to the
child and within that space, the number of toys within
reach and the number and names of forbidden objects
within reach. To record the amount of space
the observers identified each room available

to the child as small (less than 100 square feet),

34

medium (100-200 square feet) or large (more than 200
square feet). Restrictions within a room were noted
according to size, such as "4 x 4 playpen." The observers
counted and recorded toys and forbidden objects with the
aid of the checklist, Appendix G. (See Appendix C for

the results of reliability testing of these ratings.)

Rating the Child's Temperament

The two temperamental variables of interest were
activity level, defined as the degree of movement exhibited
by the child, and persistence level, defined as the degree
to which the child continued an activity as opposed to
being distracted. There were three ratings of temperament.
Each observer rated each child on these variables, using
a five-point scale from low to high. (In pretesting, the
interrater reliability for these ratings was .90.
Appendix C includes the details of this pretest.) In
addition each mother answered the questions from the Carey
Infant Temperament Survey (Carey, 1970) scales on activity
and persistence. The primary observer posed these

questions at the end of the sixth observation.

Rating_the Child's Language Comprehension
A third experimenter administered to each child
Benedict's measure of language comprehension. This
measure was designed to determine the degree of grammatical

complexity which children are able to understand. Briefly,

35

the experimenter had mothers present commands ranging
from simple to complex in terms of their number of

semantic units. ("Semantic units" are equivalent to
"open slots" as defined in this study. See Benedict,

1977, for a fuller explanation of this method.)

Data Reduction

 

This section describes the methods and formulas
which were used to summarize the raw data. These tech-
niques are explained first for the major segments of the
interaction (mothers' commands, children's compliance or
noncompliance, and mothers' responses to the children's
behaviors) and then for the secondary variables which are
theoretically related to them. The data which were
summarized and used in the analysis were from the final

four observations for each mother-child pair.

Mothers' Commands

The raw data were the numbers of commands given
by the mothers during each relevant session. To provide

a measure of the frequency of commands the formula was:

(1) number of commands
40 (minutes of observation).
Because the mothers varied in talkativeness, a second
measure of commands was necessary. This measure controlled

for the mother's total use of speech through the formula:

36

(2) number of commands ‘ .

total units of speech, including commands x 100.
This formula yielded the percent of a mother's speech
constituted by commands. In formula (2) "total units of
speech" was a count of all separate utterances addressed
to the child. This count included the forms of speech
specified in Appendix D. The data were expressed for
each mother during each session using these formulas.
These figures were then averaged for each mother over her
sessions.

Commands varied according to their form (verbal
or nonverbal), their degrees of complexity (number of
open slots) and their content (prescriptive or proscrip-
tive). For form of commands, the data simply consisted
of the percent of total commands which were verbal and
the percent which were nonverbal. The mothers' use of
the various degrees of complexity of the commands was
similarly expressed as percent of total commands.

The breakdown for mothers' use of commands in
terms of content was like that for mothers' issuance of
all commands combined. That is, formulas (l) and (2)
were used to describe prescriptions and proscriptions
separately for each mother, each session. The results

for each mother were then averaged across her sessions.

37

Children's Compliance
The raw data were the children's responses to
their mothers' commands. Compliance ratios were computed

using the formula:

(3) instances of compliance
instances of compliance plus noncompliance.

This ratio was computed for each child for each session
and then the results were averaged to yield a mean com-
pliance ratio for each child. The same formula applied

to the child's compliance to prescription and proscrip-
tions when they were considered separately.

Mothers' Responses to

Children's Compliance
and Noncompliance

Appendix E includes a list of the particular
categories of response. The initial reduction was simply
the percent of a mother's total responses constituted by
each category of response, for each session. The data
were then averaged for each mother across her sessions.
In later stages of the data analysis particular response
categories were grouped into various sets by adding their
percentages. (Those sets are fully explained in the

Results section.)

Children's Temperament

There were three measures of each of two tempera-

mental variables, activity level and level of persistence.

38

The primary measures were the ratings by each observer.
The interrater reliability for activity was .80 and for
persistence was .66. In none of the cases of disagree-
ment were the observers' scores more than one point apart
on the five-point scales. To resolve each case of dis-
agreement the observers reexamined the observational
notes regarding the child's general activities during the
two sessions of the last four in which neither observer
was rating temperament. These notes described the child's
actions according to 30-second intervals. _It was there-
fore possible to track the child's activities and the
lengths of time the child engaged in each. The observers
used these additional notes as an independent basis for
a final score. As a modest check on this procedure the
observers used the same technique for all children for
whom there was not disagreement. Their scores from these
notes agreed with their scores from the actual observations.
The third measure of children's temperament was
the Carey Infant Temperament Survey. During administra-
tion of these questions the observers noted many problems,
with the instrument and decided not to use it in the
data analysis. Appendix H includes the results from the

Carey and a discussion of the difficulties with this scale.

Children's Language
Comprehension

The examiner using the language comprehension

measure assigned each child a score or "stage of

39

comprehension" ranging from one to four. The score indi-
cated the most complex command (in terms of semantic units
or open slots) with which the child complied and presumably
understood, during the testing situation. A score of 1A
indicated a precomprehension stage where the child did not
comply with any commands, and a score of 4 meant the child
could understand relations between words, such as preposi-
tions. The children were ranked according to these results.

Analogous results were available from the observa-
tion sessions. Using the children's responses to their
mothers' commands the observers noted the greatest degree
of complexity of command with which each child complied at
least twice.

Mothers' Responses to
Children's Commands

 

The observers counted children's commands as well
as their mothers' commands. Mothers' responses to these
commands were expressed as compliance ratios computed

according to formula (3).

Mothers' Warmth and Hostility

Each mother's warmth was a count of her total num-
ber of instances of giving her child unconditional praise
or physical affection. These totals of expressed warmth

resulted in a rank-ordering of the mothers from least to

most “warm."

40

Hostility was a count of the mother's total
instances of unconditional negative statements or physical
acts directed to the child. These data were also used to

rank-order the mothers.

Environment

There were two indices of the environment. One
was space available to the child, the raw data for which
were the observer's counts of the rooms available and their
relative sizes. These data were used to rank-order the
children from those with least to most space available.

The second index was the percent of all objects
within the child's reach which the mother forbade the
child to have. The formula for this index was:

(4) total forbidden objects
total of all objects within the space

These data were also used to rank-order the children in

terms of environments.

Statistical Analysis

Most of the comparisons among variables were cor—
relational. Those correlations which involved two ratios
were computed using Pearson's product-moment correlation.
Spearman's rho applied where the final form of at least
one variable was a rank ordering. A point biserial cor-
relation was used where one measure was dichotomous, such

as the sex of the child. (See Bruning and Kintz, 1968).

41

This research was preliminary work in the study
of many of the variables. Therefore significance levels
of .10 and less were considered noteworthy. These less
stringent alpha levels allowed exploration and discussion
of findings which may merit further investigation.

Except where otherwise stated, all tests of

significance were two-tailed.

RESULTS

The organization of this section follows the
sequence of the events which were of interest for this
study. The results begin with data on the mothers'
commands, then include data on the children's compliance
and noncompliance and finally, data on the mothers'
responses to those behaviors. Major comparisons of
interest are those generated by the hypotheses, such as
the correlation between environment variables and mothers'
commands, and that between mothers' responsiveness to
their children and their children's compliance. Before
asking those questions, other descriptive information
regarding each variable must be established. For each of
the major variables the initial question is whether the
participants' responses were internally consistent.

If they were consistent it is reasonable to compare
"average" responses of each subject. This procedure
means that the descriptive data are needed for each major
variable and then for each secondary variable which is
theoretically related to it. Actual comparisons between
variables follow the presentation of the relevant

descriptive data.

42

43

Mothers' Commands

Consistency

The first question of interest is how the mothers
did in fact use commands. There are two measures reflect-
ing mothers' use of commands: the average number of
commands per minute or "frequency" and the percent of
speech constituted by commands, or simply "percent of
speech." The mothers' individual consistency across
sessions is considered separately for each of these
indices of commands. The measure of consistency is the
split-half reliability coefficient, corrected by the
Spearman-Brown formula to take into account the total
number of sessions. (That is, the data for each mother
from odd-numbered sessions were combined and correlated
with data from the even-numbered sessions. This procedure
answers the question "Were the mothers consistent across
sessions?") The resulting coefficient for frequency of
commands is r = .92 (p <.001) and for commands as a
percent of speech r = .67 (p <.01). It is therefore
reasonable to compare mothers' mean scores on these
measures.

Appendix I includes the session-by-session raw

scores on both measures of each mother's use of commands,

as well as their mean scores.

44

Frequency
The descriptive summary data showing means and
variability of both frequency and percent of speech are

presented in Table 1.

Table 1.--Mothers' Issuance of Commands.

 

Measure of Commands Mean Median Range S.D.

 

Average Number of
Commands per minute .50 .46 .18-l.20 .25
(frequency)

Percent of speech
constituted by 18 18 7-25 6
commands

 

The data indicate that a mother with the mean score would
have given commands at the rate of one every two minutes
on the average. Commands would have constituted 18

percent of her total speech directed to the child.

The two measures of commands correlate with each

other at r .57 (p‘<.01). While the measures tap
different aspects of the mothers' degrees of "commanding-
ness" mothers who frequently issued commands also tended

to have commands as a large percentage of their speech.

Repetitions

 

The findings thus far indicate that the mothers
gave separate commands at the rate of one every two minutes

on the average. Another question of interest is the

45

degree to which mothers repeated their commands. These
mothers typically repeated one-third of their commands.
If they did repeat a command they usually did so only
once, though occasionally they repeated the command twice
or more. The descriptive statistics on the mothers' use

of repetitions are included in Table 2.

Table 2.--Mothers' Repetitions of Commands.

 

 

 

 

Percent Repeated
, l 2 3
Mean Median S.D. Time Times (or
more)
Percent of all
commands
which were 32 32 10 17 8 7
repeated
Form

There are three more specific analyses of the
nature of the commands: their form (verbal or nonverbal),
their relative complexity (the number of open slots in
them) and their content (prescriptive or proscriptive).
Nonverbal commands were very rare. Only four mothers
used them, and those four mothers used a grand total of
only 11 such commands. They constituted less than 1 per-
cent of the total commands used by all mothers. There is
not enough variation in the form of commands used to

merit further study of this variable.

46

Complexity

In addition to the form of their commands, the
mothers were also fairly consistent as a group in the usual
complexity of their commands. Of all commands, 89 percent
had one or two open slots. Eleven percent had three Open
slots and less than 1 percent had four open slots. Four
mothers used commands with one open slot more frequently,
one mother used commands with one and two open slots with
equal and greatest frequency, and the remaining ten
mothers used commands with two Open slots most frequently.
As with the form of commands, the mothers' use of them
in terms of complexity is not variable enough to make

their correlation with other variables meaningful.

Content

The final way to describe the data on mothers'
commands is to consider commands in terms of content,
that is, whether they were prescriptive or proscriptive.
Every mother used more prescriptions than proscriptions.
The mothers were also individually consistent across
sessions in terms of their use of each type of command.
The reliability coefficients confirming this consistency
as well as the descriptive summary statistics are presented
in Table 3. An average mother would have issued pres-
criptions every two and a half minutes and proscriptions
every ten minutes, on the average. Prescriptions would

have constituted 14 percent of her speech and proscriptions

47

Table 3.--Mothers' Issuance of Commands: Prescriptions
and Proscriptions Separated.

 

 

 

 

 

Measure of . Reliability
Commands Mean Median Range S.D. Coefficient
Prescriptions

Average number _

per minute .41 .33 .10 .93 .21 .88***
Percent of _ **
speech 15 14 4 24 5 .72*
Proscriptions

Average number _ ***
per minute .10 .09 .02 .28 .06 .79
Percent of _ ***
speech 3 3 l 6 2 .69
**

*
p‘<.01

3 percent on the average. In effect, the distributions
of mothers' prescriptions and proscriptions were quite
different. These findings confirm the importance of
distinguishing between the two types of commands. Con-
sequently, they are examined separately in all subsequent
analyses.

Another way to compare these types of commands is
to correlate mothers' use of the two types with each
other. In terms of the frequency of the commands,
mothers' issuance of prescriptions and proscriptions are
significantly correlated (r = .62, p <.01). In terms of
the percent of speech constituted by each type, mothers'

issuance of them is not significantly correlated (r = .22,

48

ns). While the two types of commands are different by
definition and in the mothers' practice, mothers who
frequently used one type also frequently used the other
type. Appendix J includes the measures of each mother's

use of prescriptions and of proscriptions.

Reflections

Before discussing the correlations among commands
and other variables a note on reflections is in order
because commands were specifically differentiated from
reflections by definition. Commands were defined accord-
ing to the intent of the mothers' words and the context
when they spoke. The theoretical separation of commands
from reflective statements, which might have the same
words but different intent and/or content, is supported
by the findings on reflections. Mothers were consistent
across their own sessions in their use of reflections. At
the same time their expression of reflections was dis-
tributed differently, as can be seen by comparing the

results in Table 4 with those in Table 1.

Table 4.--Mothers' Issuance of Reflections.

 

 

Measure of . Reliability
Reflections Mean Median Range S'D' Coefficient

Averagf Number .37 .25 .l6-.95 .21 .82***
per minute

Percent of 15 12 8-32 6 .80***

speech

***p < . 01

49

As final evidence that commands and reflections are
indeed different categories of speech, mothers' expression
of reflections is not significantly correlated with their
expressions of commands (r = .28, ns).
Mothers' Commands Correlated with
Other Variables

With the establishment of the fact that mothers
were reasonably consistent in their use of commands, the
next questions of interest are the correlations of the
command measures--total commands and prescriptions and
proscriptions separately-~with other mother and child
variables. When necessary the descriptive data regarding
the secondary variable preceeds the correlations with
commands. The first measures correlated with mothers' use
of commands are the relatively enduring characteristics

of the child.

Sex of the Child
The sex of the child does not correlate signifi-
cantly with either measure of mothers' total commands or

with prescriptions or proscriptions considered separately.

Temperament of the Child

The descriptive data on the temperament variables
of activity and persistence are shown in Table 5. The
children were more variable in terms of their levels of

persistence than in their levels of activity. For both

50

Table 5.--Children's Temperament Scores.

 

 

Temperament Variable Mean Median Range S.D.
Activity level 3.00 3.00 2-4 .53
Persistence level 3.01 3.00 1-5 1.31

 

measures the average score was three on a scale of five.
The correlation between children's scores on the two
measures of temperament is r = .41, which is statistically
nonsignificant but suggests that further study is neces-
sary before conclusions can be made regarding the relation-
ship between them. The correlation is not high enough

to argue that the two measures are simply alternate man-
ifestations of a single variable. The children's individ-
ual scores are in Appendix K.

Table 6 details the correlations among temperament
variables and mothers' use of commands. Children's
activity levels correlate significantly with both their
mothers' frequency of total commands and with their
mothers' frequency of prescriptions. Correlations of
activity level and proscriptions are nonsignificant.

The children's levels of persistence do not
correlate significantly with either measure of their
mothers' total commands or of their prescriptions.
However, children's persistence does correlate negatively

with their mothers' proscriptions. For frequency of

51

Table 6.--Children's Temperament and Language Comprehen-
sion Correlated with Mothers' Commands.

 

Measure of Temperament Variable Language
Commands Activity Persistence Comprehension

 

Total Commands

 

Average number

 

 

per minute .46* -.22 .31

Percent of speech .28 -.30 -.03
Prescriptions

Average number

per minute .46* -.18 .43

Percent of speech .36 0 .05
Proscriptions

Average number

per minute .06 -.48* -.52*

Percent of speech .10 -.44 -.46*

 

Note: All correlations are Spearman's rho.

*
p‘<.10

proscriptions this correlation is significant. This
negative relationship indicates that children who were
more persistent than average had mothers who gave
relatively fewer commands.

In sum, activity level correlates positively
with mothers' issuance of all commands and particularly
prescriptions. The children's levels of persistence
correlate negatively with their mothers' issuance of

proscriptions.

52

Children's Language
Comprehension

 

The children's scores on Benedict's measure of

language comprehension are presented in Table 7.

Table 7.--Distribution of Children's Language Comprehension

 

 

Scores.
, ' Results from
Scores Benedict 5 Measure observations
1A 11 0
113 4 0
5 5
3 9
4 2 1

 

1The examiner using Benedict's measure reported
that four children were remarkably noncompliant during
the testing situation, one with a score of 1A, two with
scores of 1B and one with a score of 3.

The number of the score indicates the maximum number of
open slots of a command with which the child complied dur-
ing the test situation. The child who scored 1A did not
comply with even the simplest command, and the children who
scored 4 understood relations between words. Comparison
data are available from the observation sessions. The
observers assigned each child a score representing the
maximum number of open slots with which the child complied
at least twice. These findings are also presented in

Table 7.

53

The correlation between these two measures is
low (rho = .43, ns). This apparent lack of comparability
stems from two sources. Nine of the children complied
with more complex commands during the observations than
during the testing situation. At the same time the
children's scores from the observations are limited by
the commands their mothers' "naturally" gave. In two
cases where children received scores of two and in all
cases where children received scores of three their
mothers did not use more complex commands. Therefore
while many of the children's rankings on the two measures
are similar several children had no opportunity to main-
tain their ranking in the "natural" situation. Benedict's
measure provides a more thorough test of children's
responses to all relevant degrees of complexity in
commands. The results from that measure are therefore
used to rank the children in terms of their comprehension.

Table 6 includes the correlations between the
children's language comprehension and each measure of
the mothers' use of commands. Of significance is the
finding that the children's comprehension is negatively
correlated with both measures of their mothers' issuance
of proscriptions. In general, children who understood

only less complex commands had mothers who issued more

proscriptions.

54

Environment

 

Rousseau (1762/1972) hypothesizes that mothers
can minimize their use of commands by offering the child
adequate space to explore and by eliminating "capricious"
objects from within that space. The correlations between
mothers' commands and environment variables were designed
to test these hypotheses.

There are two measures of the children's environ-
ment, the space available to them and within that space,

the percent of objects within their reach which were

forbidden to them. The descriptive data on these variables

are presented in Table 8.

Table 8.--Chi1dren's Environments.

 

Environment Variable Mean Median Range S.D.

 

Space available
(in square feet) 662 600 300-1200 221

Percent of all
objects which were
forbidden 26 24 ll— 41 ll

 

In terms of space available to the children, both
ends of the continuum were houses. The child with the
least space lived in a small house in which the mother
had blocked off most of the rooms, while the child with
the most space had access to most rooms in a fairly large

house. Table 9 also indicates that every home included

55

Table 9.--Environment Variables Correlated with Mothers'
Commands.

 

Environment Variable
% of Objects

 

 

 

 

Measure of Commands Space which were
Available Forbidden

All Commands

Average number per

minute -.12 .19

Percent of speech -.66*** .41
Prescriptions

Average number per

minute -.05 .34

Percent of speech —.41 .48*
Proscriptions

Average number per

minute -.23 .01

Percent of speech -.33 .17

 

Note: All correlations computed using Spearman's

 

rho.
*
p<<.10

*‘k‘k

p‘<.01

some forbidden objects within the child's reach. No
mother had "baby-proofed" the child's space to include

less than 11 percent of such objects.

The correlations between these variables and the
mothers' issuance of commands are presented in Table 9.
Of significance is the finding that less space is associ-
ated with more frequent issuance of prescriptions but not

proscriptions.

56
Summary: Mothers' Commands and Their
Correlates

The mothers were individually consistent and
different from one another in their issuance of all
commands and of prescriptions and proscriptions con-
sidered separately. They were very similar as a group
in terms of two aspects of their commands, the form and
the complexity, and in terms of their lack of use of
repetitions. Most mothers issued commands fairly fre-
quently, one every two minutes on the average. Most of
these commands were verbal prescriptions with one or two
open slots and were not repeated.

Mothers' issuance of commands correlates signifi-
cantly with several characteristics of their children.
Mothers' total commands and particularly their prescrip-
tions were more frequent with children who had higher
levels of activity. Mothers' proscriptions were more
frequent with children who had lower levels of persistence.
The converse associations were also true. Mother's
proscriptions were also more frequent with children who
understood less complex sentences. As for environmental
variables, mothers' total commands were more frequent
in homes where there was less Space available to the
child. Mothers' prescriptions were more frequent in
homes where there were more forbidden objects within

the child's reach.

57

Children's Compliance and Noncompliance
The second step in the major interaction sequence
is the children's compliance and noncompliance to their
mother's commands. As in the discussion of commands, the
descriptive data on these behaviors are presented first,
beginning with the findings on children's individual con-

sistency across sessions.

Consistency

All of the children varied across their sessions
in terms of their compliance to commands. To determine
whether they showed reasonable individual consistency
of response, a split-half reliability coefficient was
again computed. This coefficient is r = .49 (p<:.10).
This finding indicates that the children were sufficiently
consistent to merit further examination of their scores,
but they were at the lower limit of consistency defined
as acceptable in this study.

Compliance may also be considered in terms of the
content of the relevant command, as prescriptive or pro-
scriptive. The reliability coefficients for compliance
to commands separated in this way are r = .40 for com-
pliance to prescriptions and r = .08 for compliance to
proscriptions (both nonsignificant). The children do not
appear as consistent when their responses for the day are

divided up according to the content of the commands. They

58

were particularly inconsistent regarding their reactions

to proscriptions.

Frequengy

Since only compliance to all commands showed
acceptable internal consistency, as specified for this
study, only this measure is used to study the relationships
between children's compliance and other variables. The
children's mean scores on this measure are largely but
not entirely the product of their reactions to prescrip-
tions, because prescriptions were more frequent than pro-
scriptions and because the children were more consistent
in reacting to prescriptions. The descriptive data on

the children's compliance ratios are provided in Table 10.

Table 10.--Chi1dren's Compliance Ratios to all Commands.

 

Mean Median Range S.D.

 

Compliance ratios .53 .48 .42-.64 .08

 

Most of the children complied about half of the
time. The distribution is limited, with a constricted
range and a small standard deviation. The children did
not differ radically in compliance ratios although as
shown above they were somewhat consistent in the differ-

ences they exhibited. Appendix L includes each child's

59

compliance ratios for each session as well as each

child's average ratio.

Children's Compliance Correlated With
Other Variables

In the following sections children's compliance
is correlated with other variables, beginning with the

relatively enduring characteristics of the children.

Sex of the Child

The correlation between children's compliance and

the sex of the child is rpb = .34 (ns).

Temperament of the Child

The children's compliance does not correlate
significantly with either their activity levels (rho = .01)
or with their levels of persistence (rho = .28).

Children's Language
ComprehensiOn

This variable also correlates nonsignificantly
with children's compliance (rho = .26).

The Stayton, Hogan and Ainsworth (1971) thesis
is that mothers who are "warmer" and "more responsive" to
their children will have more compliant children. Rousseau
(1762/1972) argues that mothers' responsiveness will be
associated with less compliance in children. These

hypotheses were tested by correlating children's compliance

60

with their mothers' warmth, hostility and responsiveness

to their children's commands.

Mothers' Warmth and Hostility

The descriptive data regarding these behaviors
are provided in Table 11. Mothers expressed warmth both
physically and verbally. All hostility was verbal in
nature. Because the distributions of these behaviors are
skewed the medians are better indicators than the means
of the "usual" total of each behavior. The "average"
mother expressed a total of 8.5 warm behaviors and no
hostile ones over four observations. These two behaviors
correlate with rho = -.15, indicating that there was no
relationship between them. That is, mothers who were very
warm were not necessarily less hostile. There were a
number of mothers who were not expressive in terms of

warmth gr hostility.

Table ll.--Mothers' Total Expressions of Warmth and
Hostility.

 

Mothers' Behaviors Mean Median Range S.D.

Warmth (total
behaviors) 17.73 8.5 4-67 6.57

Hostility
(total behaviors) 1.33 0 0-5 1.33

 

61

As noted earlier, mothers were rank—ordered
according to their "warmth" totals. The children's com-
pliance correlates with their mothers' warmth with rho =
-.74 (p‘<.01). The strength of this correlation is
derived from the "warmest" mother who had a child of
less than average compliance, and four mothers who were
not warm but who had highly compliant children.

The children's compliance correlates nonsignifi-
cantly with their mothers' hostility with rho = -.l4.
While the children's compliance is negatively related to
their mothers' warmth, it is not related to their mothers'
expression of hostility.

Mothers' Compliance to
Children's Commands

This behavior is the measure in this study of a
mother's "responsiveness." Children's commands were
always prescriptive, i.e., the children always requested
that their mothers g9 something. Mothers were not
individually consistent in their compliance with children's
commands. The corrected split-half reliability coefficient
is r = .35 (ns). Therefore it is not meaningful to cor-
relate children's compliance with the averages of this
behavior of the mothers. The mothers' inconsistency also
makes it unlikely that this behavior is significantly

related to the children's compliance, which was relatively

consistent.

62
Mothers' Issuance of
Commands

The children's compliance or noncompliance is
the second step of the interaction which begins with a
command from the mother. To provide a sense of how
these components relate to each other, the mothers'
issuance of commands was correlated with their children's
compliance to them. The frequency of mothers' issuance
of commands correlates with compliance with r = -.43,
while the percent of speech constituted by commands
correlates with compliance with r = -.38. Neither result
is statistically significant although both suggest the
possibility of a negative relationship between a mother's
degree of "commanding-ness" and her child's compliance.

Up to this point in the analysis the one factor
clearly associated with the children's compliance is their
mothers' lack of expression of warmth. Rousseau's (1762/
1972) and Freud's (1923/1960, 1933) theories both emphasize
the role of the mothers' response 39 compliance and non-
compliance in determining compliance. The results con-
cerning this third step of the interaction are described
next and correlated with children's compliance.

Mothers' Responses to Children's
Compliance and Noncompliance

This section of the paper addresses the third and

final step of the major interaction sequence, the mothers'

responses. This section begins with a description of

63

these responses including measures of the mothers'
individual consistency in using them. The statistical
tests of the relationships between these responses and
the children's compliance follow the description of the
behaviors.

The mothers responded in a variety of ways to
their children's compliance and noncompliance. Because
the hypotheses relate these responses to the child's
behavior, a useful categorization of them is in terms of
the child's likely View of them. In line with this idea
the mothers' responses were grouped into those which
rewarded and those which punished the child's behavior.
"No response" constituted a third category. This cate-
gorization also follows Rousseau's (1762/1972) hypothesis
that children will comply when they are fairly certain of
reward or when they fear punishment if they don't comply.

Table 12 presents the responses mothers used
following their children's compliance and noncompliance.
In each subcategory the reSponses are listed in order of
their frequency of use. For example, if a child complied
and the mother chose to reward the child she would most
likely have used praise. These lists simply indicate
which behavior was most likely to occur, over all mothers.
As such they provide a very general picture of the inter-

action. Before comparing among the mothers in any greater

64

Table 12.--Mothers' Responses to Children's Compliance
and Noncompliance.

 

Mother' Responses

 

Children's . No
Behavior Reward Punishment Response
Compliance Praise Issued second No overt
command reaction
Reflection Doublebind

Concrete reward

Noncompliance Performed act Physically No overt
herself forced child reaction
to comply

Verbally Issued second
accepted command
noncompliance

Distracted Physical
child punishment

Withdrawal of
love statement

 

Note: Responses are listed within the sub-
categories in order of their frequency of use.

detail a measure of their individual consistency over
sessions is necessary.

The mothers were consistent to varying degrees in
terms of their use of the response categories listed in
Table 12. The split-half reliability coefficients, given
in Table 13, indicate that mothers' most consistent respon-
ses to compliance were not responding and punishment.

The mothers responded consistently to noncompliance with
punishment. These results indicate that while mothers

engaged in the other behaviors, they did not do so con-

sistently.

65

Ho . v 9:;

 

 

mo. v as;

mo. II II II II wmcommmm oz

axmm. mo Ne Iva mm em ucmezmflcsm

mm. II II In nu pumzmm
moansameoocoz ou mmmcommmm

«««o>. ha om.|ea vm Hm wmcommmm oz

«tame. no mm.lo mo mo pcmenmficdm

mm. II II II II pumzmm
mUCMHHmEou ou mmmcommmm

ucmfloflmmmou .o.m mmcmm cmepmz cmmz

muwHHQMAHmm mmwuommumo oncommmm

mmmcommmm mo mocmsvmum

 

.mmmcommmm Had no mmmusmoumm 8 mm noncommmm “manoeuumm .mumnuozll.ma manna

66

Descriptive data for those categories in which
mothers were consistent are also provided in Table 13.
The mean scores indicate that in response to compliance
the mothers' lack of nny response usually constituted
more than half of their total responses. Mothers did not
punish compliance frequently, with the mean score at
less than 10 percent of a mother's total responses. While
the remaining instances of compliance were rewarded, the
mothers did not consistently exhibit this behavior.

In terms of responses to noncompliance, all
mothers used punishing responses occasionally, with a
mean at 27 percent of their total responses. The range
of this response category indicates that no mother
punished her child for noncompliance more than 42 percent
of the time on the average.

Mothers' Responses Correlated with
ChildrenTs Compliance

 

These three response categories in which the
mothers were consistent correlate to varying degrees with
the children's compliance. Mother's punishment of non-
compliance correlates significantly with children's
compliance with r = .57 (p‘<.05). Neither punishment nor
no response to compliance correlates significantly with
children's compliance (with r = -.01 and r = .21, respec-

tively).

67

In sum, children's compliance to their mothers'
commands across all their sessions correlates signifi-
cantly with just one of their mothers' responses, punish-
ment of noncompliance.

Freud's (1923/1960, 1933) hypotheses deal with
more specific responses than the response categories
used above. Specifically, Freud hypothesizes that
mothers who use their affection as a reward for compliance
and their withdrawal of love as punishment for noncom-
pliance will have more compliant children. A secondary
hypothesis is that mothers who use physical punishment
and threats of it will have more compliant children.

None of these particular responses occurred frequently.
No mother used affection as a reward for compliance in an
overt way. For example, no mother hugged or patted her
child following compliance. One mother used physical
punishment four times and threats of it twice following
noncompliance. These sequences always began with pro-
scriptions. Her child had a mean compliance ratio of
.47, which was slightly less than the mean for all the
children.

Five mothers (including the mother who used
physical punishment) occasionally made withdrawal of love
statements following their children's noncompliance to
the prescriptions and proscriptions. In four sessions,.

two mothers each made one such statement, two made two,

 

68

and one issued four withdrawal of love statements. The
mothers were divided into those who used withdrawal of
love statements and those who did not. This categoriza-
tion correlates with their children's compliance with

r = .03 (ns).

In sum, none of the behaviors which Freud hypothe-
sizes as determinants of compliance occurred frequently
or consistently, nor were these behaviors significantly
related to children's compliance in this study. Up to
this point in the analysis mothers' responses to their
children's compliance and noncompliance are associated
with compliance in just one way. There is a positive
correlation between mothers' punishment of noncompliance
and the children's compliance.

Microanalysis of Mothers'
Responses

 

The previous analyses have all been relatively
global in orientation, as they have relied on the children's
mean compliance ratios across all sessions. It is
possible that the children's responses were affected by
more contemporary events. Specifically, perhaps the
children's compliance or noncompliance to one command was
a product of the preceding command sequence between
mother and child. If the mother responded to compliance
with reward or to noncompliance with punishment, perhaps

the child was more likely to comply with the next command.

69

To test this hypothesis the sequence of mothers' responses
and children's compliance was examined using a contingency

table as illustrated in Figure l.

Mothers' Responses

Children's behavior On Off

 

Compliance

 

Noncompliance

Figure l.--Contingency Table for Microanalysis.

The mothers' responses were "on" if they rewarded
compliance or punished noncompliance. Their responses
were "off" if they rewarded noncompliance, punished
compliance or did not respond to either behavior. The
table was compiled for each child, using the mother's
response to the previous command sequence and the child's
compliance or noncompliance to the current one.

Two percentages were then computed for each child,
the percent of behaviors following a mother's "on"
responses which were compliance behaviors, and the percent
of behaviors following a mother's "off" responses which
were compliance behaviors. A t-test for related means was
used to compare these percentages among all the children.
This microanalysis answers the question, "Was the children's

compliance to one command related to their mothers' response

70

to their behavior on the preceding command?" Briefly
stated, the answer is "no," because the t-test was not
significant.

The consideration of the impact on children's
compliance of their mothers' responses to compliance
and noncompliance has resulted in one significant finding.
There is a positive association between mothers' punish-
ment of noncompliance and their children's compliance.
This global finding holds while a more microanalysis of
children's behaviors considered in light of the preceding

command sequence did not produce significant results.

DISCUSSION

Before interpreting these results, a summary of
the usual pattern of mother-child interaction is in order.
The mothers differed significantly and consistently in
their rates of giving commands. However once they stated
commands, the commands were almost always verbal prescrip-
tions of fairly simply construction (one or two open
slots). The children obeyed the commands about half of
the time, with small but consistent individual differences
around this mean compliance ratio. The mothers responded
in a variety of ways to their children's compliance and
noncompliance. They were inconsistent in terms of their
use of several of these types of responses. However,
they did respond consistently and differed from each other
in their uses of not responding to compliance, punishing
compliance and punishing noncompliance. These results
are considered below in greater detail and in light of the
various theoretical points of view.

Rousseau is the major theorist who addresses
variables which may affect the mothers' commands. He
argues that children need free space to explore and that

that space needs to be free of forbidden objects. If

71

72

these conditions are not met the adult will have to
intervene and command the child. A critical test of

these proposed associations was not provided by this
study. Every home included several forbidden objects
within the child's reach and every mother commanded her
child, with most mothers having done so rather frequently.
Neither the environment nor the result (eliminating
commands) which Rousseau proposes was found in the
mother-child pairs studied.

Among the mothers observed, Rousseau's thesis
received some support. The space available to the child
was significantly negatively correlated with the percentage
of the mothers' speech constituted by commands (but not
with the absolute frequency of commands). In effect,
with less space available to the child, the mother's
total speech was more "commanding." This finding may be
partially the result of the child not having enough
space, so that the child was more likely to encounter
forbidden objects and elicit proscriptions from the mother.
However, this variable, space available to the child, was
inversely correlated with the mothers' issuance of 311
commands, not proscriptions alone. Therefore the variable
contributes in some other way to the rest of the commands,
that is, to the mothers' use of prescriptions.

In small spaces the mothers and children were in

close proximity whether they wished to be or not. Perhaps

73

the mothers became unhappy and therefore more controlling
in response to enforced closeness. An observation in

two of the smallest homes supports the idea that they

may not have enjoyed the closeness. The mothers in these
two small homes were the only two who relied on child
”gates" to prevent their children from entering certain
areas. Further, both mothers typically closed off the
kitchens when they were in them and the children were
outside of them. These observations suggest that the
mothers may have had minimum space requirements not met by
all the homes. To account for the correlation between
space and commands one must further assume that with

too little space, mothers become unhappy and more command-
ing.

Mothers whose speech contained relatively large
percentages of prescriptions also tended to be those who
left many forbidden objects within their children's reach.
They did not issue more pnnscriptions so there is no
indication that their children actually grabbed or
threatened the objects in some way. There are at least
two possible interpretations of this finding. One inter-
pretation is that those mothers who left more forbidden
objects within their children's reach had learned to
redirect their children away from them by commanding pre-
scriptively, as a kind of camouflage. This suggestion

seems unlikely. If a mother were to manage such a command

74

she would have to speak and act with no indication that
the child had to stop its current actions or the command
would have been coded as proscriptive. While it is
possible that a certain set of mothers were quite skilled
in this type of intervention, it did not strike the
observers that way. Both observers felt that separating
prescriptions from proscriptions was a simple, direct
task. If a mother offered a distraction, it was usually
of the form, "No, you can't have that. Here, play with
this." Such statements make clear the proscriptive nature
of the initial command. Along the same line of reasoning,
one might argue that mothers who left many forbidden
objects within their children's reach frequently spoke
prescriptively simply to keep their children busy and
occupied away from such objects. While no definite test
of this hypothesis is possible, both observers felt
strongly that most mothers reacted to their children much
more often than they engaged in playful commands. That
is, even with prescriptions most mothers seemed to respond
to the child's ongoing play or activity.

A more plausible interpretation is that a third
factor led to this correlation between the number of
prescriptions and the number of forbidden objects within
reach. This additional factor might be labeled the
mothers' overall "directiveness." That is, mothers who

were generally more directive issued more commands and

75

also left more forbidden objects within the children's
reach. They may have expected the children to adjust to
an adult environment. This study does not provide a
direct test of this hypothesis. The narrative account
given by one mother who gave very £33 commands indicated
she was also very low in general directiveness. When
speaking about her son's transition from crawling to
walking, she said, "We had to learn to accommodate to
him." She pointed to the living room, in which the space
from the floor to about 36 inches high was barren of

all objects except toys. (No mother who gave many
commands spontaneously provided an account of insisting
that the child accommodate to adults.) This interpreta-
tion fits Rousseau's model rather closely, in suggesting
that adults have some control over their "commanding-ness"
by careful attention to the child's environment. Of
course, the mother's level of general "directiveness“
must be low enough that she is willing to make such
accommodations. ‘

In sum, the‘Correlations between environmental
variables and the mothers' use of commands generally
support Rousseau's understanding although a complete
test of his theory was not provided.

Other variables assumed to be fairly enduring
characteristics of the child were also associated with

the mothers' use of commands. These variables were not

76

specifically tied to a theory of mothers' use of commands.
The children's temperaments, activity levels and persis-
tence levels correlated in different ways with their
mothers' commands. There was a positive correlation
between the children's activity levels and their mothers'
use of all commands, and particularly prescriptions.

The correlation may be interpreted in two ways. Mothers
who had more active children may have found it necessary
to direct their activities more, so that total commands
and prescriptions increased. The other possibility is
that children who had more commanding mothers had to be
more active to keep up with the commands. This latter
interpretation is unlikely. Children who were more active
were both more physically active during play and also
simply moved more. Again a narrative account may help
clarify the issue. The typical behaviors which made two
of the children seem particularly active were not of a
sort that the mother commanded. For example, one of these
two, a boy who received a score of four on activity level,
spent a great deal of time running through the house
dragging a string, while his cat chased him. While his
mother allowed this activity she did not command it.

It seems more likely that in the case of the correlation
between activity level and commands the mothers were

responding to a characteristic of the child, and found it

77

necessary to be more directive with those children who
were more active.

The other association between temperament and
mothers' commands was a negative correlation between
mothers' use of proscriptions and children's persistence
levels. The more persistent children were typically
more self—reliant in the sense that they could amuse
themselves for longer periods of time. They may have been
less likely to involve themselves in activities which
would elicit a proscription simply because their atten-
tion wandered less. The other possible interpretation is
that mothers who commanded their children less frequently
allowed their children to develop greater persistence.
This second interpretation is less likely because the
correlation holds only for proscriptions, not for pre-
scriptions or for all commands. Mothers spoke proscrip-
tively in response to a behavior or anticipated behavior
of the child. Therefore the direction of the correlation
is relatively clear, and again indicates that the mothers
responded to the children's temperaments.

The final correlation with mothers' use of
commands is a negative one with children's language
comprehension. Specifically, children who understood
less complex sentences had mothers who gave proscriptions
more frequently and for whom proscriptions constituted a

larger percentage of their speech. The most likely

78

interpretation of this finding is that children who under-
stood less simply did not comprehend general rules like
"Never touch the plants." Such a command, entirely out

of context, has three open slots. The mean comprehension
score on Benedict's measure was only two. Therefore, many
children probably did not understand such rules and there-
fore broke them more often, eliciting proscriptions from
their mothers.

These interpretations of variables which affect
mothers' commands suggest that in large part the mother is
a reactor. She responds to the child's temperament and
language comprehension, and in part, to the space avail-
able to them. Howeveri the mother does more than respond.
The likely interpretation of the correlation between
forbidden objects and mothers' commands is that a general
characteristic of the mother which might be labeled ‘
”directiveness" played a part in mothers' commands and
in the number of forbidden objects they left within the
child's reach. Therefore the mothers certainly did more
than react. Children's levels of persistence and language
comprehension were negatively associated with mothers'
issuance of proscriptions. Presumably if the mothers had
left nn forbidden objects within reach, they would not

have required proscriptions regardless of their children's

persistence level or language comprehension.

79

In terms of Rousseau's theses, these findings
indicate that his hypotheses are generally accurate but
not complete. Providing adequate space and eliminating
forbidden objects appear very important in limiting mothers'
use of commands. However, the child's temperament,
particularly activity level, appears to have an impact
on commands which would not be controlled simply by
monitoring the environment. It would apparently take a
Special adult temperament to avoid issuing more commands
in response to a very active child for example. The
issue of the impact of adult temperament and Specifically,
whether mothers differ on a variable labeled "directive-
ness," would be useful to study in any further research
in this area.

The children's compliance or noncompliance was
the next step in the interaction. The children in this
study were somewhat consistent across sessions in their
compliance to all the commands their mothers gave.

They were not consistent in responding to prescriptions
and proscriptions considered separately. Children's
compliance correlated with only two variables, negatively
with mothers' warmth and positively with mothers' use of
punishment following noncompliance. A microanalysis of
the results which examined children's compliance as a
consequence of their mothers' responses to their previous

behavior did not prove significant. These results indicate

80

then that children's compliance is affected by more
global rather than minute aspects of the interaction.
For example, no mother punished noncompliance even half
the time on the average, and the mean among mothers was
use of punishment approximately one-quarter of the time.
Although the mothers did not use punishment frequently
they did so consistently. Apparently punishment for
noncompliance is a potent enough response that the con—
Sistent, occasional use of it "encouraged" compliance.

The strong negative correlation between mothers'
warmth and children's compliance is more complicated to
explain. The warmest mothers frequently praised or gave
physical affection to their children when the child's
compliance was not an issue. In behavioral terms, these
women frequently entered into their children's activities
to insert praise or affection for particular acts. They
often did so as they observed and commented on the
children's ongoing play. The positive correlation
between compliance and punishment for noncompliance suggests
that children may be more compliant if their mothers make
the contingencies for their behaviors clear. Warm mothers
may have confused the observable contingencies by pro-
viding praise and affection regardless of the child's
responses to commands.

This model suggests that to insure compliance a

' ""’.""‘\.

mother would have to reward compliance and punish

81

noncompliance consistently while refraining from praising

any other of the child's acts. The results of this

 

study indicate that mothers could use punishment fairly
infrequently if they used it consistently. However, this
study does not provide a test of the impact of consistent
praise for compliance as the mothers Simply did not
employ it often or consistently. These findings are
considered below in light of the various theoretical
points of view.

Stayton, Hogan and Ainsworth (1971) argue that
children will be more compliant if their mothers are warm
and responsive to them. These authors use an "ethological"
framework and suggest that children are naturally com-
pliant and will express this behavior if only the maternal
environment will allow and support it. This point of
view receives no support from this study. Indeed, mothers'
warmth here was inversely correlated with their children's
compliance while mothers' hostility was not significantly
associated with compliance. The mothers were not consis-
tent in their responsiveness to children's commands, so
that a correlation between this variable and compliance
would not be meaningful. This lack of consistency also
makes it unlikely that there is a significant relationship
between responsiveness and children's compliance, which

was a fairly consistent behavior.

82

The occurrence of such different findings between
this study and Stayton, Hogan and Ainsworth (1971) needs
explanation. The most plausible reason for the differ-
ent findings is that these studies defined the character-
istics differently. In this study the raters observed
and recorded specific behaviors hypothesized to be the
behavioral outcomes of warmth, hostility and responsive-
ness. Stayton, Hogan and Ainsworth (1971) used observers'
rankings which were not tied to particular behaviors,
making it difficult to interpret their findings. What-
ever characteristics they tapped with their rankings are
apparently not equal to overt displays of affection or
hostility or mothers' compliance with children's commands.

Rousseau (1762/1972) and Freud (1923/1960, 1933)
posit Similar theses regarding very young children.

Both argue that young children are motivated by the direct
results of their compliance or noncompliance. Specifically,
Rousseau (1762/1972) argues that children will comply

only if they are promised concrete reward or punished

for noncompliance. This position is supported partially

by the results of this study. These children were more
compliant if their mothers used punishment following
noncompliance. This finding was significant despite the
infrequent but consistent use of such punishment. These
mothers did not use reward following compliance consis-

tently enough to allow a correlation of that technique

83

with compliance. Their inconsistency also makes it
unlikely that the children's fairly consistent compliance
ratios were associated with that response from the
mothers.

This study does not allow for an adequate test
of the Freudian hypotheses. These mothers typically did
not use the techniques which Freud argues will engender
compliance, i.e., physical punishment, threats of physical
punishment, withdrawal of love following noncompliance
and "proofs of affection" as a reward for compliance.
Mothers who did use any of these techniques did so rarely
and inconsistently.

However, the negative correlation between mothers'
warmth and children's compliance is suggestive. This
finding could be interpreted as indirect support for the
Freudian model in the following way. Freud posits that
if children fear withdrawal of love for noncompliance they
will comply. Therefore, if mothers provide praise and
affection liberally the child will not fear the loss of
love so readily and will be less compliant. This inter-
pretation is plausible and does not contradict the earlier
interpretation of the finding. It does make Freud's
understanding extremely concrete by equating "love" with
”overt praise and affection." It iS not clear from
Freud's brief statements on the subject whether he

intended this reduction of the notion "withdrawal of love."

84

This study indirectly supports the Freudian
hypothesis without providing a clear test of it. Further,
the correlation between punishment for noncompliance and
children's compliance indicates that Freud's hypothesized
variables are an incomplete list at best. They did not
occur consistently enough in this study to account for
compliance.

As noted, this study did not provide a complete
test of either Freud's (1923/1960, 1933) or Rousseau's
(1762/1972) point of View because the variables these
authors posit as important contributors to compliance
were not observed frequently and/or consistently. How-
ever, these children were fairly compliant. The theoretical
issue is, then, that both Freud's and Rousseau's des-
criptions of variables leading to compliance are more
clean-cut, perhaps more idealized, than what one actually
observes between mothers and children at home.

Children's compliance appears to be a function of
mothers' lack of expressed warmth and mothers' use of
punishment as a technique if their children did not
comply. The view of children which emerges from these
findings is that they will not comply without a fairly
potent motive. However, despite the consistent individual
differences among the children, their group mean for
compliance was approximately 50 percent. That is, mothers'

warmth and use of punishment affected deviations around

85

that mean. The questions of why the children were that
compliant and so Similarly compliant arises.

These mothers were Similar in at least one way
which appears important to children's compliance. All of
the mothers punished noncompliance at least occasionally,
and relatively consistently. Perhaps occasional punish-
ment was sufficient to insure nearly 50 percent compliance,
and compliance increased with more frequent use of it.
This explanation would account for the degree of compliance
as well as the similarities among children. Mothers'
warmth does not apply as easily to the explanation because
these mothers were quite variable as a group in their
exhibition of warmth.

The effectiveness of occasional punishment of
course calls to mind learning theory and in particular the
concept of intermittent reinforcement. Rousseau's (1762/
1972) point of view is in agreement with learning theory
in emphasizing the importance of the consequences of a
child's acts. In this one aspect of the results, both
points of view apply; Rousseau's theory is more comprehen-
sive as he addresses both the mothers' commands as well as
the children's compliance. This viewpoint and terminology
have therefore been used throughout.

The more interesting theoretical question is
whether the children's compliance was a product of a

natural inclination to comply. That is, is the Stayton,

86

Hogan and Ainsworth (1971) thesis correct although they
appear to have wrongly understood the correlation between
compliance and warmth? There are interesting data avail-
able which indicate that many of these children were not
naturally compliant. The language comprehension measure
was also based on children's responses to commands,
written by the examiner and administered by the mother.
One interesting Sidelight to the results of this measure
is that many children were uncooperative and not compliant
during these sessions. The examiner reported several
stories including one about the girl who had the highest
compliance ratio during the four final observation sessions
of all the children. During the administration of the com-
prehension measure this child would not comply with a Single
command. She Simply sat in her rocking chair. AS the
examiner prepared to leave the child's mother asked the
child to help pick up the toys. The child did so, clearly
demonstrating her ability to select the toy her mother
named and place it in the bag which the examiner had brought.
This narrative account indicates that this child, and others
who responded similarly, were 223 compliant when the Sit-
uation changed. That is, they did not exhibit a "trait"
of compliance which carried over to their mothers' commands
when the commands were written by someone else and admin-
istered out of the usual context of interaction.

This account of results from the language compre-

hension measure also indicates that there may be additional

87

variables which occur in most fairly natural Situations
and foster compliance. That is, there may be several vari-
ables in addition to mothers' warmth and use of punishment
which contribute to children's compliance. In their usual
interactions, the mothers may provide the children with
subtle rewards which were absent during the language com-
prehension measure, such as smiles and physical approaches,
without actual touching. In addition, most commands given
during the observations sessions were prescriptions and
many occurred during the course of play. The children may
have understood that if they complied, then the play would
continue.

The results of this study indicate that these chil-
dren did not "naturally" comply. When they did comply it
was for a number of reasons, some of which were documented
here and some identified only speculatively. This study
for example does not provide tests of the impact of the
consistent use of rewards following compliance. This study
also did not tap a variety of subtle interaction cues to
which the children may have responded.

The strength of the study is that it does provide
a look at the actual sequence of events which begins with
a command from the mother. As such, the study provides
interesting documentation of many of the daily events
between mothers and young children. A serious limitation

to generalizing these results is of course the sample Size

88

and characteristics. The observations need replication
with a larger group and one more varied in socioeconomic

status before any wide generalizations are appropriate.

I “tain‘mm I” min-‘13“

, 3'

APPENDICES

89

APPENDIX A

Phone Interview and Number of Potential

Subjects Contacted

90

Q‘-

 

The primary observer conducted the phone interviews.
Initially, She introduced herself and gave the following

statement.

I'm from MSU and am doing research on young children's
typical responses and reactions. I got your name

from the county birth records because you apparently
have a child about the right age to participate.

In general, I'm observing young children and their
mothers in their own homes. Are you interested in
'hearing more about this?

If the woman indicated further interest the observer
stated the other qualifications for participation, that
the child be the only child, be able to walk, and that
the mother be the primary caretaker of the child. If
the mother met these qualifications the observer explained
in detail the number and length of visits, the role of the
second observer, and the visit by the language comprehen-
sion examiner. The observer also acknowledged to the
mother that these statements were somewhat ambiguous
about exactly what was being observed, and promised a full
report at the end of the study. The observer offered
the mother individual results regarding her child's
temperament and language comprehension, and $5.00 for
her help. If the mother was still interested, they
worked out a schedule for the six observations.

The first subject contacted was a friend of a

friend, who also gave the names of two other mothers who

participated. The other 12 subjects' names were obtained

91

. f

Fir—_u _________ ——_""
Peg.
.

 

92

through the birth records. The observer called 51 mothers

through birth records, of whom 14 (27 percent) agreed to

participate. Of these 14, two dropped out before the

first observations without giving any reason. No mother

dropped out once the observations began.

the reasons that the other 37 mothers gave for not par-

ticipating.

Table A.--Potential Subjects' Reasons for Not Participating.

Table A details

 

Reason Given

Sex of the Child

 

Boy Girl
Had other children 8 8
Mother worked and not the primary
caretaker 3 7
Child had died 1 0
Outright "no" with no reason 4 5
Total who did not participate 16 21

 

._.; z
‘73!»

IT'—_ —— —————— —————

 

APPENDIX B

Sample Observation Recording Form

93

 

94

Susaaumo:
Havamscm

coauumuua
Hwosmxnm

mcaouooou

MOM mam>umucfl :om poumoflpcfl

HOSuO

.osomm

. - .flgo .02

“meme. 4%. “and.
. .7 . . m . . m.UHHSU

.u xflccmddm Ca @600 wow

 

macaw cwdo

Euom mcapuooom coflum>uomno oHdEmm

m xHDZQQQG

m

.:\= 6 >2 pmumofiocfl mmB wocwuusooo nomm~

.o xflocmdmd cfl mooo momH

mpfluo
umcuo
.Qflomm
.nbo uo< macauauomom Acummdm >us>auo< Boom
mpcmEEoo CH m . . m.umnu02 Hmuocmo

APPENDIX C

Reliability of Measures

95

WW

Main Observational Record

The two observers pretested the main observational
record by observing two mother-child pairs. After pre-
liminary work on operational definitions, they agreed
completely on the occurrence of commands, the number of
open slots in each and the ensuing sequence of interaction
in two separate sessions. They reached a .90 level of
reliability regarding the occurrence of reflections. They
were also in total agreement during the final two pre-
testing observation sessions about instances of the
child commanding the mother and the mother's responses,
and about displays of physical affection. Neither observer
identified behavior considered "unconditionally hostile"

during the pretesting sessions.

Children's Temperament
The two observers also pretested the ratings of
children's temperament at the Michigan Early Childhood
Center. After three hours of preliminary rating and
hassling with each other, they achieved a .89 interrater
reliability for both indices of temperament, activity

level and persistence level.

Ratings of the Environment
The two observers pretested the ratings of the

environment at the homes of the two mother-child pairs

96

 

 

97

used to pretest the main observational record and at a

day care home associated with the Michigan Early Childhood

Center. The raters always agreed about the amount of

space available to the children. They also reached

near-perfect agreement on the count of forbidden objects

within the child's reach, disagreeing on only one object fit

within the three homes.

. APPENDIX D

Code for Categories of the Mothers' Speech

98

 

5
d
d.
L

'1
‘ i“_~_--Q..

 

Speech related to the child's compliance or non-

compliance was coded and recorded verbatim.

Other speech was coded.

Q-
UH-
UP-
Pl-

R-

The

command (later separated into pre- and pro-

scriptions)

threat

praise

promises

withdrawal of love statements. I

question

unconditional hostility

unconditonal praise or affection

playful statements ("peek-a-boo") and animal
or toy noises

rhetorical statements, where the mother
appears to be talking to herself. For example,
"I seem to have lost my glasses."

‘ “at“,

 

person to whom the statement was addressed

was indicated by a subletter, if other than the child.

no designation: addressed to child

E:
O:

addressed to experimenter
addressed to another person

Example: QE was a question addressed to the

experimenter.

99

APPENDIX E

Glossary of Terms

100

 

Mother's Speech

Command - statement Spoken with intent of directing the
child either to initiate new activity (prescription)
or to stop ongoing activity (proscription)

Threat - indicated physical punishment would occur,
depending on child's actions

Praise - positive evaluative statement considered
"unconditional" if not in response to compliance

Promise - indicated desirable action would follow
child's act

Withdrawal of love - indicated mother's disapproval of
child's act 3nd that mother was withholding affec-
tion as a result

Hostility - negative evaluative statement considered
"unconditional" if not in response to noncompliance

Rhetorical statements - mother appeared to be talking

to herself
Play - game statements ("peek-a-boo") and animal noises

Reflection - description of child's ongoing activity

Mothers' Responses to
Compliance

Praise (as above)
Reflection (as above)
Issuing another command - a command which required

different actions
Doublebind - promised reward but did not deliver

No response - no overt response within two minutes or
before next command

Mothers' Responses to
NoncompIiance

Performed act for the child

Physically directed child's movements--forced compliance

Physical punishment - direct negative contact such as
Slap or spanking

Isolation - physically confined child, away from mother

Withdrawal of love (as above)

No response (as above)

Children's Commands - child cried, pointed and vocalized
or used words to indicate desires

101

102

Mothers' Warmth

 

Unconditional praise (as above)
Physical affection - direct positive contact not in
response to compliance or noncompliance

Hostiligy

 

Unconditional hostile statement (as above)
Physical hostility - direct negative contact not in
response to compliance or noncompliance

.lmgwmm man. man

APPENDIX F

Code for the Mothers' Responses to
Children's Compliance or

Noncompliance

103

..
M -

rot-r

. .L'. {xv-.151

 

IP‘

D - directing the child's physical movements I
(forced compliance)

N - no visible reaction

P - performing the act for the child ;

 

PP- physical punishment
R - concretely rewarding the child

I - physical isolating the child

Repetitions of the command were indicated in a
separate column. Threats and withdrawal of love state-

ments were coded in the column for "mother's Speech."

104

 

APPENDIX G

Checklist for Rating the Environment

105

 

ll)” '1 L.
1:

 

For each room to which the child had access, the

observer indicated the following:

§inn_(small, medium or large) according to approximate
square footage)
Further restriction on Size or accessibility, such
as playpen or child "gate."
The observers also indicated the number of objects

within each category which the child could reach.

uncovered light plugs

electric cords

objects on tables (specify)
objects on low Shelves (Specify)

other probably forbidden objects (Specify)

 

Number of toys to which the child had direct access

106

 

I

[7.1
i

 

 

APPENDIX H

Carey Infant Temperament Survey:

Results and Critique

107

 

Each mother answered the Carey questions regarding

activity level and level of persistence. There were

five questions for each of these variables. The observers
scored the results using Carey's suggested method which
weights the responses and results in scores between zero
and two for each variable. A score of zero indicated

high or increased level, one indicated an average level
and two a low level for the child on that particular

variable. The group results are presented in Table B.

Table B.-—Children's Scores on Carey Infant Temperament

 

 

Study.
Mean Median Range S.D.
Activity Level .76 .80 0 - 1.6 .42
Persistence Level .87 1.00 .2 - 1.4 .36

 

These results were not used in the main analysis
because of several problems that became apparent during
administration of the Carey. The behavioral problem was
that many of the women did not answer all the questions,
resulting in scores based on only three or four questions.
The women usually did not answer questions because of
difficulties in wording. For example, many of the ques-

tions ask in effect "15 your child very active, average or

108

 

 

109

not very active in (a Specified Situation)." Many women
quite reasonably said that they had no basis for judgment
as this child was their first. Others answered such
questions but did so with qualifications such as "Oh well,
he seems very active to me." That is, the problems with
the wording of the questions led to some women not
answering and some women obviously "guessing" to supply
an answer.

Another problem which became apparent is that the

Carey is designed for younger children, making some ques-
tions inappropriate. For example, a question from the

persistence scale asks in effect, "Can your child lie in
the play pen or crib and amuse her/himself for as long as

half an hour?" None of the children could do so, probably

 

because the norms for activity and persistence change
with age. As a result the most persistent child observed
could amuse himself for 15 minutes but never half an hour.
In sum, the observer found that the Carey includes
many items which are poorly worded or Simply inappropriate
for this age child. One result of these problems is
that many women did not answer all the questions. There-
fore the Carey scores are often based on three or four
items. AS a result the scores from the Carey were not

used in the data analysis.

Mothers'

APPENDIX I

Use of Commands:

110

Individual Data

 

V"

The following data are each mother's use of
commands in each session and as an average across sessions.
Each mother's data includes both her frequency of giving
commands, i.e., the average number of commands per minute,
and the percentage of her Speech constituted by commands.

Data are included for the final four observations.

Table C.--Mothers' Use of Commands: Individual Data.

 

 

 

 

Sessions

Mother 3 4 5 6 Average
Gl

Frequency .55 .45 .63 .93 .64

% of speech 19 23 22 35 25
G2

Frequency .10 .30 .45 .18 .26

% of Speech 07 13 13 10 11
G3

Frequency 1.38 1.48 .80 1.13 1.20

% of Speech 23 36 22 18 25
G4

Frequency .78 .50 .70 .40 .60

% of Speech 22 18 18 15 18
G5

Frequency .45 .28 .28 .30 .33

% of Speech 16 09 12 07 11
G6

Frequency .58 .65 .40 .58 .55

% of speech 15 20 14 19 17
G7

Frequency .55 .58 .35 .23 .43

% of Speech 31 18 14 09 18

111

112

Table C.--Continued.

 

 

Sessions

Mother 3 4 5 6 Average
Bl

Frequency .95 .55 .38 .68 .64
% of Speech 27 l7 17 21 21
B2

Frequency .73 .93 .53 .48 .68
% of Speech 19 33 17 20 22
B3

Frequency .25 .45 .43 .70 .46
% of speech 13 20 26 18 19
B4

Frequency .28 .15 .53 .58 .39
% of speech 08 04 ll 15 10
B5

Frequency .20 .05 .38 .03 .17
% of speech 24 29 15 33 25
B6

Frequency .20 .25 .20 .08 .18
% of speech 10 08 05 03 07
B7

Frequency .45 .70 .30 .40 .46
% of Speech 19 21 21 25 22
B81

Frequency .58 .95 .30 .20 .51
% of speech 19 l7 l3 13 16

 

 

1The sex of the child is identified by G (girl)
and B (boy). The numerical identification of the mother-
child pair is consistent across appendices.

APPENDIX J

Mothers' Use of Prescriptions and

Proscriptions: Individual Data

113

E

”Li-.1}. .Tiifi'
1
m - .

J

 

The data in Table D are each mother's use of pre-
scriptions and proscriptions in each session and as an
average across sessions. Each mother's data includes
both her frequency of giving commands, i.e., her average

number per minute, and the percentage of her speech

constituted by commands.

Table D.--Mothers' Use of Prescriptions and Proscriptions:
Individual Data.

 

Sessions

 

Mother Commands Variable Average
3 4 5 6
G1 Prescriptions
Frequency .17 .16 .15 .83 .20
% 50 33 43 88 54
Proscriptions
Frequency .05 .13 .20 .05 .11
% 02 06 07 02 04
G2 Prescriptions
Frequency .10 .25 .40 .18 .23
% 08 ll 12 10 10
' Proscriptions
Frequency 0 .05 .05 0 .03
% 0 02 01 0 01
G3 Prescriptions
Frequency 1.05 1.30 .50 .85 .93
% 18 31 l4 13 19
Proscriptions
Frequency .33 .20 .30 .28 .28
% 06 05 08 O4 06
G4 Prescriptions
Frequency .78 .53 .68 .35 .59
% 22 17 18 14 18
Proscriptions
Frequency 0 .03 .03 .05 .03
% 0 01 01 02 01

114

 

Table D.--Continued.

115

 

 

Mother Commands Variable §§§§i2§§ Average
3 4 5 6
G5 Prescriptions
Frequency .43 .25 .20 .30 .30
% 15 08 09 07 10
Proscriptions
Frequency .03 .03 .03 0 .02
% 01 01 03 O 01
G6 Prescriptions
Frequency .43 .45 .33 .50 .43
% 11 14 11 l6 l3
Proscriptions
Frequency .15 .20 .08 .08 .13
% 04 06 03 03 04
G7 Prescriptions
Frequency .53 .48 .33 .23 .39
% 3O 15 13 09 17
Proscriptions
Frequency .03 .10 .03 O .04
% 01 03 01 0 01
B1 Prescriptions
Frequency .83 .43 .33 .50 .52
% 23 13 15 16 17
Proscriptions
Frequency .13 .13 .05 .18 .12
% 04 04 02 06 04
32 Prescriptions
Frequency .13 .75 .48 .32 .57
% 19 27 l6 14 19
Proscriptions
Frequency 0 .18 .05 .15 .09
% 0 06 02 06 04
B3 Prescriptions
Frequency .20 .28 .30 .55 .33
% 11 13 l8 14 14
Proscriptions
Frequency .05 .18 .13 .15 .13
% 03 08 08 04 06

 

"W ~.

Table D.—-Continued.

116

 

 

Mother Commands Variable 3 §E§§12£§ 6 Average
B4 Prescriptions
Frequency .23 .13 .43 .43 .31
% 07 04 09 ll 08
Proscriptions
Frequency .05 .08 .10 .15 .10
% 02 01 02 04 02
B5 Prescriptions
Frequency .18 .05 .30 .03 .14
% 21 29 12 33 24
Proscriptions
Frequency .03 0 .08 0 .03
% 03 0 03 0 02
B6 Prescriptions
Frequency .13 .13 .10 .04 .10
% 06 04 03 02 04
Proscriptions
Frequency .08 .13 .10 .03 .09
% 04 04 03 01 03
B7 Prescriptions
Frequency .40 .30 .20 .35 .31
% 17 09 14 22 16
Proscriptions
Frequency .05 .40 .10 .05 .15
% 02 12 07 03 06
B8 Prescriptions
Frequency .45 .85 .25 .15 .43
% 15 15 ll 10 13
Proscriptions
Frequency .13 .10 .05 .05 .08
% 04 02 02 03 03

 

 

APPENDIX K

Children's Individual Scores on

Temperament Variables

117

k
!

trmm.fma
.‘

'l'v
'- '

 

Each child's scores on the two temperament vari-
ables, activity and persistence, are presented in Table E.
The scores are based on observers' ratings using a five-
point scale with a score of one as "low" and five as

"high.”

Table E.--Chi1dren's Individual Scores on Temperament

 

 

Variables.
Child Activity Level Persistence Level
G1 3 3
G2 3 4
G3 3 3
G4 3 4
G5 3 3
G6 3 4
G7 3 3
Bl 4 2
B2 4 3
B3 2 1
B4 3 3
B5 3 5
B6 3 4
B7 2 1
B8 3 4

 

118

APPENDIX L

Children's Compliance Ratios: Individual Data

119

Each child's compliance ratio for each of the
final four observations and the average of those sessions
are presented in Table F. The compliance ratio indicates

the percentage of all commands with which the child

complied.

Table F.--Children's Compliance Ratios: Individual Data.

 

 

Child Session Average Across
3 4 5 6 Sessions
G1 .48 .39 .28 .60 .44
G2 .75 .40 .51 .57 .63
G3 .46 .57 .41 .45 .47
G4 .61 .50 .64 .81 .64
G5 .66 .45 .64 .67 .61
G6 .44 .62 .50 .61 .54
G7 .46 .78 .57 .56 .59
B1 .53 .36 .64 .26 .85
B2 .45 .54 .43 .42 .46
83 ' .70 .45 .71 .47 .58
B4 .64 .67 .72 .35 .60
B5 .50 .50 .31 1.00 .58
B6 .75 .33 .50 .33 .48
B7 .45 .43 .42 .56 .47
B8 .44 .42 .58 .25 .42

 

120

LIST OF REFERENCES

121

LIST OF REFERENCES

Benedict, H.E. "A Clinical Method for Assessing Language
Comprehension in Very Young Children." Manuscript,
1977.

. "Language Comprehension in 10 to 16 Month-Old
Infants." Doctoral dissertation, Yale University,
1976.

Brookhart, J. and Hock, E. "The Effects of Experimental
Context and Experimental Background on Infant's
Behavior Toward Their Mother and a Stranger."
Child Development, 1976, 41, 333-340.

Bruning, J.L. and Kintz, B.L. Computational Handbook of
Statistics. Glenview, Illinois: Scott, Foresman
and Company, 1968.

 

Burton, R.V., Maccoby, E.E., and Allinsmith, W. "Antece-
dents of Resistance to Temptation in Four-Year-Old
Children." Child Development, 1961, 32, 689—710.

Carey, W.B. "A Simplified Method for Measuring Infant
Temperament." Journal of Pediatrics, 1970, 11,
188-194.

Crandall, V., Orleans, 8., Preston, A., and Rabson, A.
"The Development of Social Compliance in Young
Children." Child Development, 1958, 29, 429-443.

Escalona, S.K. The Roots of Individuality. Chicago:
Aldine, 1968.

Fenichel, O. The ngchoanalytic Theory of Neurosis. New
York: Norton, 1945.

Freud, S. [The Ego and the Id.] (J. Strachey, Trans.)
New York: Norton, 1960. (Originally published,

1923.)

. [New Introductory Lecture on Psychoanalysis.]
(W.J.H. Sprott and J. Strachey, Trans.) New York:
Norton, 1933.

122

123

Goldberg, S. and Lewis, M. "Play Behavior in the Year-
old Infant: Early Sex Differences." Child
DevelOpment, 1969, 40, 31-32. A

Hoffman, M.L. "Parent Discipline and the Child's Consid-
eration for Others." Child Development, 1963, 31,
573-588.

Hoffman, M.L. and Saltzstein, H.D. "Parent Discipline and
the Child's Moral Development." Journal of Person-
ality and Social Psychology, 1967, 5, 45-57.

Hogan, R. "Moral Conduct and Character." Psychological
Bulletin, 1973, lg,'217-232.

Holzman, M. "The Verbal Environment Provided by Mothers
for Their Very Young Children." Merrill-Palmer
Quarterly, 1974, 22, 31-42.

 

Lytton, H. Determinants of Compliance in Two-Year Old
Boys: Comparative Findings Via Three Approaches.
JSAS Catalog of Selected Documents in Psychology,
1973, 3, 121. (Ms. No. 475).

McFarlane, J.W., Allen, L., and Honzik, M.P. A Develop-
mental Study of the Behavior Problems of Normal
Children Between Twenty-One Months and’Fourteen
Years. Berkeley: University of California
Press, 1954.

 

Messer, S., and Lewis, M. "Social Class and Sex Differences
in the Attachment and Play Behavior of the Year-
Old Infant." Merrill-Palmer Quarterly, 1972, 18,
295-306.

Minton, C., Kagan, J., and Levine, J.A. "Maternal Control
and Obediance in the Two—Year-Old." Child Develop-

Rousseau, J. Emile. (B. Foxley, Trans.) London: J.M.
Dent and Sons Ltd., 1972. (Originally published,
1762.)

Searle, J.R. Speech Acts and Recent Linguistics. Annals
of the N.Y. Academy of Sciences, 1975, 263, 27-38.

Sears, R.R., Maccoby, E.E., and Levin, H. Patterns of
Child Reari_g. New York: Harper and Row, 1957.

 

Sears, R.R., Rau, L., and Alpert, R. Identification and
Child Rearing. Stanford: Stanford University

Press, 1965.

 

124

Spock, B. Baby and Child Care. New York: Pocket Books,
1968.

 

Stayton, D., Hogan, R., and Ainsworth, N. "Infant Obediance
and Maternal Behavior: The Origins of Socializa-
tion Reconsidered." Child Development, 1971, 42,
1057-1069.

Thomas, A., Chess, 8., Birch, H.G., Hertzig, M.E., and
Korn, S. Behavioral Individuality in Early Child-
hood. New York: New York University Press, 1960.

Thomas A., Chess, 8., and Birch, H.G. Temperament and
Behavior Disorders in Children. New York: New
York University Press, 1968.

Tulkin, S.R., and Kagan, J. "Mother-Child Interaction in
the First Year of Life." Child Development, 1972,

_4_3_, 31-41.

Wenar, C. "Executive Competence and Spontaneous Social
Behavior in One-Year-Olds." Child Development,
1972, 43, 256-260.

 

RES

rang I
lH‘llWlll
E55i4

|\
ES

TE UN V. L

MICHIGAN STa
\ll‘l‘llHIWIIW“
31293

I

E3

 

1'11 L"