12 SCHIZOPHRENIA BULLETIN

the contributions of the familyto the development of competence in children

Diana Baumrind*

The ideal home or school in the late 1940's and1950's was organized around unlimited acceptance ofthe child's current needs for gratification, rather thanaround preparation for adult life. The child was to begranted maximum freedom of choice and self-expressionin both settings. Spock's 1946 edition of Baby and ChildCare advocated such infant-care practices and the exten-sion into early childhood of lenient disciplinary prac-tices. Yet the avalanche of studies on the effects ofinfant-care practices did not support the supposedharmful effects of such restraints on the child asscheduled feeding, early weaning, and early toilettraining. Indeed, Spock's emphasis altered in the 1957edition. Comparing the changes in child-rearing practicesfrom 1940 to 1955, he stated that "Since then a greatchange in attitude has occurred and nowadays thereseems to be more chance of a conscientious parent'sgetting into trouble with permissiveness than strictness"(p. 2). In his Redbook columns (beginning in 1964) andhis new book entitled Raising Children in a DifficultTime (1974), Spock speaks out more affirmatively forthe reinstitution of parental controls and for theinculcation by the parent of ideals and standards. Thevigorous introduction of permissive and child-centeredattitudes into educational philosophy began at least 40years ago (Coriat 1926 and Naumberg 1928) as anoutgrowth of the psychoanalytic theory of psychosexualdevelopment. The view that the effects of adult author-ity on the child are inhibiting, neurotogenic, andethically indefensible was promoted by such articulatespokesmen as Goodman (1964), Maslow (1954), A. S.

•Requests for reprints should be addressed to the author atthe Institute of Human Development, University of Californiaat Berkeley, 1203 Tolman HaJI, Berkeley, Calif. 94720.

Neill (1964), and Rogers (1960) in the fields ofeducation and child rearing.

The practices favored by American parents forinfluencing the actions and character of their offspringhave varied from time to time in keeping with thepredominant view of the child as a refractory savage, asmall adult, or an angelic bundle from heaven. Theseconvictions have, for the most part, been based onhumanistic or religious values rather than upon scientificfindings. Research findings have helped to debunkcertain clinically derived notions about the obligatoryneurotogenic effects of one or another of the commonchild-rearing practices, notions characterized perhapsmore by creative flair and inner certitude than bydemonstrable validity.

My program of research was initiated to investigatethe effects of actual patterns of child rearing currentlypracticed by American parents. For the past 15 years Ihave supervised a program of research that has as itsmajor objective the identification of effects of alterna-tive patterns of parental authority on the developmentof instrumental competence in children and adolescents.

The term "instrumental competence" derives fromParsons' (1951) distinction between instrumental andexpressive functions. Parsons designates as instrumentalthose functions:

. . . oriented to the achievement of a goal which is ananticipated future state of affairs, the attainment ofwhich is felt to promise gratification; a state ofaffairs which will not come about without theintervention of the actor in the course of events.Such instrumental or goal-orientation introduces anelement of discipline, the renunciation of certainimmediately potential gratifications, including thatto be derived from passively "letting things slide"

ISSUE NO. 14, FALL 1975 13

and awaiting the outcome. Such immediate gratifica-tions are renounced in the interest of the prospec-tively larger gains to be derived from the attainmentof the goal, an attainment which is felt to becontingent on fulfillment of certain conditions atintermediate stages of the process, [pp. 48-49]

By expressive functions, Parsons means activitieswhere "the primary orientation is not to the attainmentof a goal anticipated for the future, but the organizationof the 'flow' of gratifications... and of course thewarding off of threatened deprivations" (p. 49). Parsonsregards expressive functions (such as receptivity, nur-turance, and empathy) as traditionally feminine andinstrumental functions (such as assertiveness, ambition,self-discipline, and objectivity) as traditionally mascu-line. As Parsons points out, traditionally masculinequalities—whether one is male or female—further successin competitive achievements; those most able to surviveand flourish with the least dependence upon others arethose who perform instrumental functions in the mostcompetent manner.

In choosing to study the antecedents of instrumentalcompetence, I have made the assumption that thequalities basic to instrumental competence are and willcontinue to be of benefit to the individual and thesociety. I believe that the abilities both to accommodateto social mores and laws and to take self-assertive andautonomous action in opposition to those mores andlaws when their legitimacy cannot be defended areessential to healthful, successful functioning of personsof both sexes in any society. Parents differ in the valuethey place on instrumental competence; these differ-ences are in large part a function of political philosophy.The modern intellectual parent has the difficult task ofreconciling opposing values. While the conservativeparent can teach the child to conform but not to dissent,and the radical parent can teach his child to dissent butnot to conform, the modern intellectual parent mustteach the child to discriminate between legitimate andillegitimate authority and to conform to the former anddissent from the latter. This must be done in a societywith ill-defined and conflicting values. Because theyquestion the legitimacy of the state's authority, manyliberals also question the legitimacy of their ownauthority. The liberal will ask whether there are anycircumstances in which one individual has the right toconstrain the activity of another or to shape another'svalues—even in the parent-child relationship. Parents'political attitudes, particularly those concerning the

relation between freedom and control inevitably affecttheir philosophy of child rearing.

Studies of Child-Care Practices

In a pilot study (Baumrind 1967), three patterns ofchild rearing emerged, each associated with an identifi-able type of child behavior. These patterns, which weredesignated authoritative, authoritarian, and permissive,differed primarily in the role that authority played inthe parents' thinking and practices. Immediately follow-ing this pilot study, a larger study (Baumrind and Black1967) of linear relations between parent attitudes andchild characteristics was conducted with a separatesample. Several years later a large longitudinal study wasinitiated to investigate more intensively the kinds ofrelationships revealed in the earlier studies. The pre-school phase of this longitudinal study was detailed in amonograph entitled Current Patterns of Parental Author-ity (Baumrind 1971a). The methodological features ofthis investigation were intended to overcome some ofthe shortcomings of previous research on socializationeffects. Most previous research had relied on retrospec-tive reports and upon ratings with an inadequate database.

Subjects

Subjects in all three studies were homogeneous insocioeconomic background and education; they camefrom white, middle class, well-educated, and urbanefamilies. The children who were selected for the coresample all attended nursery schools in or near Berkeley,Calif. Their average age was 4-3/4 years, with no childyounger than 3-3/4 years. Their average IQ was 125. Allwere white. (The data on 16 black families werereported separately; see Baumrind 1972a.) The childrenfrom the core sample continue to reside in Berkeley, anarea in the forefront of social change. The reader shouldbear in mind the population sampled; this sample,relative to the general population, is nontraditional,child centered, and rational. Thus, when I speak of"rejecting" parents, I am talking about parents who areless child centered than the rest but not actuallyneglectful, abusive, or brutal.

Methodology

The methodological features that characterize ourprogram of research are as follows:

14 SCHIZOPHRENIA BULLETIN

• In-depth contact with a large number of familieswho could be classified as normal (but whose childrenshow a range of competencies and deficiencies thatsuggest "low risk" and "high risk" personality types).

• Use of both molar ratings and molecular indices.The molar ratings are carefully developed, combinedinto reliable clusters, and based on at least 25 hours ofobservation across a range of situational contexts. Themolecular measures are related systematically to themolar measures of family communication, guided bytheory rather than being merely descriptive, and focusedon mctacommunication among several family members.

• Use of a longitudinal study format in which thedata for each time period are analyzed completelybefore proceeding to the next period of data collectionsso that specific hypotheses suggested by findings fromearlier time periods can guide the successive stages ofdata collection and in which the multivariate analyses ofthe numerous variables focus on configural as well aslinear relations among the parent and child variables.

Trait Ratings

Trait ratings are used extensively to assess disposi-tional tendencies in children. By trait, I mean a relativelyenduring characteristic of an individual, manifest withina broad range of circumstances, which can be used todistinguish him from another individual. The constructof a dispositional trait is used to account for continuityand stability within the changing personality (Emmerich1964); it suggests a generalized structure that renderssocial stimuli functionally equivalent for the individual,initiating and guiding his adaptive and stylistic behaviorat various stages of his development.

For the subject's unique disposition to becomemanifest, the observer must study him carefully in manysituations, and must ask him about the meaning of theobserved activity. For example, suppose the observer'stask is to rate children on the trait aggression. One childis aggressive in the sense that he bullies weaker childrenbut is not aggressive in the sense that he resistsdomination from stronger children. A second childresists domination by other children and is aggressiveonly in the presence of bullies and intruders capable ofinterfering with his ongoing activities. Toward weakerchildren, he may be protective. A third child is aggressivein most circumstances; he will go after what he wants

whether the other is weaker or stronger, friendly orunfriendly, adult or child. Were each child to beobserved in a variety of situations by a highly discrimi-nating observer who asked the child what he felt hisbehavior meant, the child's behavior might well bereliably predicted as aggressive or bullying or resistant todomination. In an experimental situation or during abrief period of observation, the way these variouschildren discriminate among stimuli might not benoticed-although each child's behavior is probably quitepredictable to his agemates, who are attuned to theircontext by personal needs. It is by the patterning ofthese traits that the uniqueness of each child is captured.

Multiple Stimuli and Behavior Settings

No single research method can assess all aspects ofthe empirical reality under investigation, and eachmethod has its characteristic strengths and weaknesses.Within the limits of a given budget, however, an attemptcan be made to assess numerous aspects of the subject'spsychological reality. The characteristics of the subjectsinteract with those of the behavior setting. Individualobservers and investigators screen selectively on the basisof personal preference, training, life history, and theo-retical orientations. Hence, a variety of behavior settingsare necessary to ensure that no important feature of thesubject's psychological world is neglected. The methodsused include self-report, interview, observation (bothstructured and field), and standardized tests. Table 1summarizes the sample sizes and the kinds of measure-ments that were made for each of the three studies ofpreschool children and the followup of the longitudinalstudy. In addition to the cross-sectional studies withpreschool children, we are presently conducting alongitudinal study using families from the last and mostextensive of the earlier studies supplemented by anadditional 60 families. These children are now 8 to 9years old and will be studied through secondary school.

Longitudinal Design

A longitudinal design permits a search extendingthroughout the development of the child for the earlyfamily patterns contributing to competence. The use of

ISSUE NO. 14, FALL 1975 15

Table 1. Data sources and sample sizes.

A. Socialization practices associated with dimensions of competencein preschool boys and girls.1

(45 girls, 50 boys)

Procedure Involving parent Involving child

Self-reportInterview

ObservationStructuredField

Standardized tests

NoneChild-rearing attitudes and practices

NoneOne home visit

None

NoneNone

None3 months of observation at nursery school

by psychologistStanford-Binet Intelligence Test

B. Child care practices anteceding three patterns of preschool behavior.2

(16 girls, 16 boys)

Procedure

Self-reportInterview

ObservationStructured

Field

Standardized tests

Involving parent

NoneChild-rearing attitudes and practices

Mother teaches child with cuisenaire rodsMother plays with childTwo home visits

None

Involving child

NoneNone

3 puzzles of graded difficulty given bypsychologist

14 weeks of observation at nursery schoolby psychologist

Stanford-Binet Intelligence Test

C. Current patterns of parental authority.(60 girls, 74 boys)

Procedure

Serf-reportInterview

ObservationStructuredField

Standardized tests

Involving parent

Parent attitude inquiryChild-rearing practices interview

NoneTwo home visitsNone

Involving child

NoneNone

Observation of intelligence testing sessionApproximately 36 school visitsStanford-Binet Intelligence Test

1 Baumrind and Black (1967).'Baumrind (1967).'Baumrind (1971a).

16 SCHIZOPHRENIA BULLETIN

Table 1. Data sources and sample sizes. (Continued)

D. Current study: Longitudinal followup of past sample plus new subjects.4

(46 girls, 58 boys—longitudinal; 34 girls, 22 boys—new)

Procedure Involving parent Involving child

Serf-report

Interview

ObservationStructured

FieldStandardized tests

Ethical judgment and parental attitudesquestionnaire

Child-rearing practices and attitudeinterview

Ethical judgment interview

Parent teaches Piaget tasksParent plays with childFamily ethical discussionTwo home visitsWitkin Rod and Frame TestDraw Yourself Test

Internal vs. external locus of control andsocial desirability questionnaire

Child identity interview

Child moral interviewValue of Life film and interview

Observation of intelligence testing sessionCard buildingFree playThree school visitsWitkin Rod and Frame TestDraw Yourself TestCreativity testsTest of role-taking abilityStanford-Binet Intelligence TestWISC Performance Scales

4 Total of 160 subjects.

such a design obviates the need for retrospective reports,which are known to lack the validity of data based onsuccessive observations taking place at crucial periodsthroughout the subject's childhood and adolescence. In alongitudinal design, pathological and contrast groups canbe identified and studied throughout childhood andadolescence. Some contrast groups that require a longi-tudinal design in a study of the etiology of schizophreniaare:

• A group of children identified as "high risk" atthe latency period who do not develop severe pathologyin adolescence;

• A group of families in which the parents sharewith the pathological group certain characteristics hypo-thesized to be pathogenic (e.g., the stereotypic domi-nant, overprotective mother and weak father) but wheresevere pathology is absent in the child in latency andadolescence; and

• A group of children unusually competent through-

out childhood and adolescence included to determinewhether there is a constellation of family behaviors thatis entirely absent for this group (e.g., neglect) butfrequently present in families of pathological groups.

There are several other advantages of a longitudinaldesign worthy of mention. A longitudinal design permitsthe investigator to pinpoint a crucial period of develop-ment in which one or another form of family interactionis particularly pathogenic (e.g., "double binding" byparents when the child is developing logical operations).Also, direction of cause-effect relationships can bedetermined more readily with repeated observations overtime. Which situational influences are the concomitantsor effects of a pathology and which are their antecedentscannot be determined with a cross-sectional design.Given a certain set of correlations across time, itbecomes possible to cast alternate explanations concern-ing antecedent-consequent relations (concerning, let ussay, the relationship of maternal overpossessiveness to

ISSUE NO. 14, FALL 1975 17

withdrawal in the child) as formal hypotheses and toaccept the hypothesis that is most tenable after the timedependencies over several observational periods havebeen considered.

In each of the studies to be discussed, trainedobservers visited the homes twice from before dinner tothe child's bedtime; written records were made of theobservations, and each parent was interviewed sepa-rately, with the interview tape-recorded. Using this database, each parent was rated on numerous scales ofspecific child-rearing techniques that permitted assess-ment of the degree to which they enforced rules, madedemands on the child, used reason when giving direc-tives, and demonstrated intimate communication andwarmth. Over a period of 3 to 5 months, one of a teamof observers recorded numerous interpersonal episodesfor each child as he or she took part in routine nurseryschool activities. The observer also watched as the childwas administered the Stanford-Binet Intelligence Test.From these multiple observations, the observer rated thechild on more than 70 scales using the Q-sort method.The Q-sort is a personality inventory in which the judgesorts a series of items into several piles ranging fromextremely characteristic to extremely uncharacteristic ofthe subject. Generally, a forced distribution is used(Block 1961). In these studies, the observer was requiredto sort the items into a rectangular distribution consist-ing of nine piles. Ratings were "bl ind" (i.e., uninflu-enced by knowledge of home factors, since differentobservers visited the home). The interrater reliabilities ofthese ratings averaged 0.70.

In the preschool period, factor-analytic studies gen-erally yield two fundamental dimensions: 1) behaviorthat is socialized and responsible at one end and dis-obedient and unfriendly at the other and 2) behaviorthat is independent and autonomous at one end anddependent and suggestible at the other. At the earlyprimary school period, a three-dimensional model withtwo of the dimensions somewhat correlated best des-cribes the data. The first dimension remains fairly wellintact but the second becomes differentiated into two,one of which measures social independence and domi-nance and the second, intellectual creativity and pur-posiveness. At later ages we expect these dimensions toemerge as three independent factors; that is, socialindependence and intellectual creativity should becomemore differentiated as the individual matures and thusshould no longer be significantly correlated.

In this presentation, I will first summarize oursubstantive results for three separate studies with pre-school children. Then I will present certain preliminaryfindings from the core longitudinal sample in which thechildren average 8-3/4 years. Next I will present certainanalyses comparing high and low risk types which wereundertaken especially for this paper. I will conclude witha brief presentation of hypotheses concerning the familyetiology of schizophrenia.

Parent-Child Relations Duringthe Preschool Period

Socialization Practices Associatedwith Dimensions of Competence

The objective of this study was to identify parentdimensions and variables associated with instrumentalcompetence in preschool children. A correlational studycan postulate the general effect on the child of atheoretically interesting parent dimension. If no effect isobserved for a parent variable thought to have auniversal linear effect, or if an effect is seen but thedirection is opposite to that which would be postulatedas universal by a currently popular child-rearing belief ortheory, the validity of that belief or theory mayproperly be questioned.

In this study, subjects were 95 sets of parents andtheir preschool children. Data were analyzed separatelyfor boys and girls. Behavioral and interview data onparents were derived 1) from home visits, with the datain the form of discrete Home Visit Sequence Analysis(HVSA) variables computed from a molecular analysis oftranscripts of the evening home visit and 2) frominterviews, with parent interview dimensions arrived atthrough cluster-analytic techniques. Data on childrenwere derived from prolonged observations in nurseryschool and structured settings and consist of Q-sortratings. A child-behavior model (similar in structure tomodels presented by Schaefer (1961) and Becker andKrug (1964) was developed and its relation to theseparent measures was assessed. The names given theQ-sort clusters in the four-cluster solution described inBaumrind and Black (1967, p. 297) were disaffiliative-affillative, resistive-cooperative, independent-dependent,^nd assertive-withdrawn.

18 SCHIZOPHRENIA BULLETIN

Table 2. Correlation coefficients between parent variables and child-behavior clusters(four-cluster solution).1

Part A. Girls

Parent variables Correlations with child behavior clusters

Disaffiliative-

aff iliative

Resistive-

cooperative

Independent-

dependent

Assertive-

withdrawn

Father-interview clusters

Warmth

Consistent discipline

Strictness concerning orderliness . . .

Punitiveness

Mother-interview clustersWarmth

Consistent disciplineMaturity demands

PunitivenessSocialization demands

HVSA variablesPositive outcomeAccepts power conflict with child . .

Independence training

Respects child's decision

Uses reason to obtain compliance . .Encourages verbal give-and-take . . .

Satisfies child

Uses coercive power

Takes initiative in control sequences

18350910

1010202612

222107052230061205

-.22-.25.27.25

.04

.08

.24-.12.05

-.04-.32.17.01

-.18-.29-.04-.11.17

-.13.05.25.34

.03

.01

.15

.09

.29

-.09-.32.15.17

-.05-.06-.19-.28.28

.02

.28

.08

.08

.24

.12

.09-.05.38

.16

.04

.32

.12

.30

.21-.20-.42.19

1 Baumrind and Black (1967).1 Significant at p <. .05 for r > .30 for Independent tests.

Table 2 summarizes the relationships between child-rearing dimensions and dimensions measuring compe-tence in boys and girls.

Maturity Demands

The effects of maturity demands on girls weremeasured by father-interview clusters designated strict-ness concerning orderliness and mother-interview clus-

ters designated maturity demands and socializationdemands. For boys, parent-interview clusters were thesame, except that a mother cluster equivalent tosocialization demands for girls did not appear. TheHVSA variable, independence training, was the behavi-oral measure of maturity demands for both sexes.Maternal socialization demands and independence train-ing were both associated with assertive behavior in girls.The maternal maturity demands cluster was associatedwith independence and assertiveness in boys. For both

ISSUE NO. 14, FALL 1975 19

Table 2. Correlation coefficients between parent variables and child-behavior clusters(four-Cluster Solution). (Continued)

Parent variables

Part B. Boys

Correlations with child

Disaffiliative-aff iliative

Resistive-cooperative

behavior dusters

Independent-dependent

3

Assertive-withdrawn

Father-interview clustersWarmthConsistent disciplineStrictness concerning orderliness . . .Punitiveness

Mother-interview clustersWarmthConsistent disciplineMaturity demandsPunitivenessRestrictrvenessEncourages independent contacts . .

HVSA variablesPositive outcomeAccepts power conflict with child . .Independence training controlRespects child's decisionsUses reason to obtain compliance . .Encourages verbal give-and-take . . .Satisfies childUses coercive powerTakes initiative in control sequences

- . 0 8- . 2 5- . 2 1

.24

.07- . 1 0- . 1 7

.14

.06

.03

- . 1 8- . 0 7- . 0 2- . 3 2

.12- . 0 5- . 1 7

.00- . 0 5

.12

.06- . 1 3- . 0 2

.17- . 0 5

.20- . 0 7- . 0 9

.21

- . 0 6.05

- . 1 1- . 2 8

.26

.03- . 1 8- . 0 9- . 1 1

.18

.42- . 1 1

.02

.12

.04

.34

.10- . 2 9

.42

.02

.20- . 1 1

.14

.36

.13

.01- . 2 7- . 0 9

.11

.30- . 0 9

.01

.13

.05

.40- . 1 3- . 0 8

.15

.02

.22

.01

.40

.16

.16

.22- . 0 3

.03

'Significant atp < .05 for r > .28.

boys and girls, then, significant mother-child effectsappeared between measures of maturity demands andassertive behavior in the child.

Firm Control

The effects of firm control were measured for bothboys and girls by father- and mother-interview clustersdesignated consistent discipline and HVSA variablesdesignated positive outcome, accepts power conflict, and

takes initiative in control sequences. Paternal consistentdiscipline was associated in girls with affiliative behaviortoward peers. The HVSA variable designated acceptsconflict with child (which measures the instances inwhich the parent confronts the child and precipitatesconflict) was associated in girls with cooperative be-havior with adults but also with dependent behavior.Paternal consistent discipline was associated in boys withindependent and assertive behavior. The HVSA variable,accepts conflict with child, was associated positively

20 SCHIZOPHRENIA BULLETIN

(though insignificantly) with independent behavior inboys rather than associated negatively as with girls.

Rational Methods of Discipline

The effects of rational methods of discipline weremeasured by HVSA variables designated respects child'sdecision, uses reason to obtain compliance, encouragesverbal give-and-take, and uses coercive power (negative).Assertive behavior in girls was associated positively withuses reason to obtain compliance and negatively withuses coercive power. Girls whose parents encourageverbal give-and-take were affiliative. For boys, respectschild's decision was associated with affiliative, coopera-tive, and assertive behavior, while uses reason to obtaincompliance was associated with independent behavior.

Parental Warmth

The effects of parental warmth for both boys andgirls were measured by father and mother interviewclusters designated warmth and by the HVSA variabledesignated satisfies child. Warmth, as measured by thesevariables, was not a significant predictor of childbehavior for either sex in this study.

Parental Punitiveness

The effects of parental punitiveness for both sexeswere measured by father- and mother-interview clustersdesignated punitiveness. Punitiveness in the father corre-lated negatively with measures of consistency andpositively with low self-confidence, suggesting that thepunitive father was weak rather than strong. Punitivenessin the father was associated positively with independentbehavior in girls but not for boys.

Maternal Restrictiveness

The effects of maternal restrictiveness for boys weremeasured by two interview clusters for mothers of boysdesignated restrictiveness and encourages independentcontacts. Both were related significantly to indepen-dence in the son, the former negatively and the latterpositively.

Summary

In summary, warmth was not linearly related toindices of either social responsibility or independence inboys or girls. The belief that unconditional acceptance

promotes competence—defined either as self-assertive-ness and independence or as affiliativeness and coopera-tiveness—was not supported by these data. In fact,parental practices that were stimulating and to someextent tension producing (e.g., maternal socializationand maturity demands and paternal punitiveness forgirls) were associated in the young child with assertive-ness. Firm paternal discipline was associated for girlswith socially responsible behavior and for boys withindependence and self-assertiveness. Consistent disciplineand high socialization demands were not characteristicsof restrictive or punitive parents. In fact, the oppositewas true. Restrictive, nonrational discipline (in contrastto consistent discipline and high maturity demands) wasassociated with withdrawn, dependent, and disaffiliativebehavior in both boys and girls. The use of reason toobtain compliance was significantly associated in theparent with a complex of other parent variables similarto those that define authoritative discipline (i.e., highmaturity demands, encouragement of independent con-tact, and lack of punitiveness), and this may contributeto its associations in the child with independent andsocially responsible behavior.

Typological Approach to the Studyof Parent-Child Effects

In the studies that follow, a typological approach todata reduction was used to reveal the relations betweencontrasting configurations of parent variables and childbehaviors. It will be seen that the relationship of aparticular parent behavior to child behavior dependsupon the total configuration of variables. For example,the relation of high parental control to social responsi-bility and independence in the child depends upon theextent to which the parent also encourages individualityand independence; thus, a distinction is made betweenthe effects on the child of authoritarian control and theeffects of authoritative control, which will be discussedshortly.

First Study

Methods. Subjects were 32 children (aged 3 and 4)selected from among all those enrolled at the ChildStudy Center, Institute of Human Development, Uni-versity of California at Berkeley during the fall semesterof 1961 (see Baumrind 1967). The selection proceduresbegan with the assessment of all the children in the Child

ISSUE NO. 14, FALL 1975 21

Study Center on five dimensions intended to measureinstrumental competence at ages 3 and 4: self-control,approach-avoidance tendency, self-reliance, vitality(from buoyant to dysphoric), and peer affiliation. Eachdimension was illustrated and given concrete meaningfor the nursery school teachers by reference to relevanttime sample categories and by instances of actualobserved behavior. After being observed for 14 weeks,the children were ranked on each dimension by boththeir nursery school teacher and the observing psycholo-gist. The 52 children who received one of the fivehighest or the five lowest rankings on at least two of thefive dimensions were observed further, individually, in alaboratory setting where they were given some standard-ized tasks. For example, one task included three puzzlesgraded in difficulty so that each child experienced easysuccess, probable success, and certain failure. Theirresponses to success and failure were observed and ratedon the five dimensions by the testing psychologist andby the observing psychologist. In order for a child toremain in the study, the observing and the testingpsychologists' ratings of the child in the two settings hadto concur. Using these multiple-assessment procedures,three contrasting groups of children were selected, eachwith a clear-cut, stable pattern of interpersonal at-tributes.

The three groups of children were selected in orderto test a set of hypotheses concerning the interactingeffects on child behavior patterns of parental control,parental maturity demands, parent-child communica-tion, and parental nurturance. Pattern 1 contained allchildren (six girls, seven boys) who were ranked high onvitality (from buoyant to dysphoric), self-reliance, ap-proach-avoidance tendency, and self-control. Pattern 2contained all children (seven girls, four boys) whoranked low on peer affiliation and vitality and did notrank high on approach-avoidance tendency. Pattern 3contained all children (three girls, five boys) who rankedlow on self-reliance, self-control, and approach-avoidance tendency.

Each of these children was observed interacting withthe mother in a 2-hour structured teaching sessionfollowed by a play experience in the laboratory setting(see Baumrind 1967a, pp. 68-70, for description of thestructured observation). Also, two home visits weremade to each family by a psychologist who had notpreviously rated the child's behavior. These home visits,which were structured identically for each family, took

place during a period from shortly before the dinnerhour until just after the child's bedtime. This 2- to3-hour period is commonly known to produce instancesof parent-child divergence and was selected for observa-tion in order to elicit a wide range of critical controlinteractions under maximum stress. Each mother andfather was interviewed separately, and the interviewswere tape recorded. Families were rated on a 24-itemparent behavior rating scale (PBRS), based in part on theFels Parent Behavior Rating Scales (Baldwin, Kalhorn,and Breese 1949). These observations and interviewswere used to rate each parent's early activities on fourdimensions of child-rearing practices selected for theirtheoretical importance as predictors of competence inpreschool children.

At the time this study began, considerable concernabout the negative effects of strict discipline on childrenwas being expressed, particularly in the clinical andeducational literature (Becker et al. 1962, Goodman1964, Kagan and Moss 1962, Maslow 1954, Neil I 1964,and Rogers 1960). In the study reported here, parentalcontrol was defined so that punitiveness and arbitrari-ness were not components. Thus, the effects on the childof strict, punitive discipline could be distinguished fromthe effects of discipline that, though equally strict, wasneither punitive nor arbitrary.

Results. Pattern 1 children (designated mature) hadparents who, in comparison with all other parentsstudied, were both controlling and warm and were ratedon all measures as communicating more clearly withtheir children. Despite parental readiness to use rein-forcement, the homes of pattern 1 children, based onthe observations, lacked discord and disciplinary fric-tion. According to the interview, these parents usedcorporal punishment rather than such tactics as ridicule,frightening the child, or withdrawal of love; however,they generally used positive reinforcement rather thannegative reinforcement to obtain compliance. Pattern 1children, subjected to consistent parental pressure formature and obedient behavior, were both sociallyresponsible and assertive. This combination of highcontrol and positive encouragement of the child'sindependent strivings was called authoritative parentalcontrol.

Pattern 2 children (designated dysphoric and dis-affiliated) had parents who were rated lower on use ofrational methods of control and were less nurturant andsympathetic with their children. They were not quite so

22 SCHIZOPHRENIA BULLETIN

controlling as parents of pattern 1 children but, accord-ing to interview data, these parents (especially mothers)were inclined to give respect for parental authority andreligious belief as reasons for their demands. Unlikeparents of pattern 1 children, they did not attempt toconvince the child through use of reason to obey adirective; nor did they encourage the child to expresshimself when he disagreed. The attitudes they expressedto the child were less sympathetic and approving andthey more often admitted to frightening the child. Theseparents, who were detached, controlling, and somewhatcool relative to other parents studied, were call authori-tarian.

Pattern 3 children (designated immature) hadparents who, in comparison to parents of pattern 1children, behaved in a markedly less controlling mannerand were not so well organized or so effective in runningtheir households. According to the interviews, they didnot feel in control of their child's behavior or feel theirinfluence on the child to be high. Compared to otherparents, they made very few demands on their childrenfor mature behavior. Fathers were especially lax. Parentsengaged in less independence training, although theygranted the child's demands for independence; accordingto the ratings, they also babied their children more. Bycomparison with mothers of pattern 1 (mature) children,mothers of pattern 3 (immature) children used with-drawal of love and ridicule—rather than overt power,physical punishment, or reason—to obtain compliance.The major difference between parents of pattern 2(dysphoric and disaffiliated children) and pattern 3(immature children) is that the former were morecontrolling and less warm, while the latter, designated aspermissive, were less controlling and somewhat warmer(although not so warm as parents of pattern 1 children).The child-rearing styles characterizing the parents ofchildren in each of the three patterns are summarized intable 3.

Second Study

While the methods and variables used in this studyare similar to those in study 1 described above, theydiffer in several important respects: 1) Parent-childrelationships for boys and girls are studied separately. 2)Study 2 is longitudinal; the children were first observedin nursery school and are now being studied at ages 8 to9. 3) The sample is more varied; the children wereselected from 13 different nursery schools, including

private and public cooperatives as well as University ofCalifornia facilities. 4) Patterns of parental authorityadditional to those observed in study 1 have beenidentified.

Subjects in study 2 were chosen from childrenenrolled in the fall of 1967 and in the spring of 1968 in1 of the 13 nursery schools. Out of this group a finalsample was selected that comprised 134 white childrenand their families and 16 black children and theirfamilies.

After the usual period of observation, the observersorted the 72-item Q-sort into a rectangular distributionconsisting of nine piles. Eight items were put into eachpile. The items were constructed to measure eightconstructs: high vs. low stress tolerance, high vs. lowself-confidence, achievement vs. nonachievement orien-tation, approach-oriented vs. avoidant behavior, auto-nomous vs. suggestible behavior, rebellious vs.dependable behavior with adults, destructive vs. con-structive behavior with peers, and alienated vs. trustingapproach to adults and peers. After the children's scoreson the items were intercorrelated, seven empiricalclusters of items appeared:

1. Hostile-friendly. This cluster consists of suchtraits as understands other children's position, nurturanttoward other children, bullies other children, insulting,and selfish.

2. Resistive-cooperative. This cluster measurescompliance with adult authority and includes such itemsas obedient, facilitates nursery school routine, tries toevade adult authority, and provocative with adults.

3. Domineering-tractable. This cluster means abra-siveness with adults and children in terms of traits suchas nonintrusive, hits only in self-defense, does notquestion adult authority, and timid with other children.

4. Dominant-submissive. This cluster measures peerleadership and contains items such as suggestible, plansactivities for other children, and individualistic.

5. Purposive-aimless. This cluster measures self-directiveness and contains such items as spectator,interesting, arresting child, vacillates and oscillates, anddisoriented in his environment.

6. Achievement oriented-not achievement oriented.This cluster measures self-directed cognitive effort andcontains such items as does not persevere when he en-counters frustration, likes to learn new skills, sets him-self goals that expand his abilities, and gives his best towork and play.

ISSUE NO. 14, FALL 1975 23

Table 3. Parental child-rearing dimensions characterizing three patterns of character developmentin preschool children.

Pattern of characterdevelopment (child)

Child-rearing dimension (parents)

Control Maturity demands Communication Nurtu ranee

Mature-competentDysphoric-disaffiliativeI m mature-dependent

431

431

422

423

Note: 4 = High; 3 = Medium-high; 2 = Medium-low; 1 = Low.

7. Independent-suggestible. This cluster of theo-retical interest overlaps with clusters 3, 4, and 5 aboveand contains such items as individualistic, stereotyped inhis thinking, and suggestible.

Data about family interaction were described interms of 15 hypothetical constructs, rather than by the4 used in the first study. Fifty parent behavior ratingscales (PBR) were devised to assess the observed andreported behavior of the mother and father separately;25 additional scales measured the joint influence of theparents.

Each scale is quite specific in its referents, generallyreferring to behavior that could be observed in the homevisit and discussed during the interview. For example,item 78, rated separately for the mother and father, is a4-point scale and reads:

Disciplines Harshly

1. Discipline or correction is administered in a harshor frightening manner.

2. Discipline or correction is administered in anonsupportive manner.

3. Discipline or correction is administered in asupportive manner.

4. Discipline or correction is administered with loveand concern.

The joint scales rate behavior that cannot easily bedistinguished on the basis of parent responsibility. Forexample, item 66, also a 4-point scale, reads:

Fixed Bedtime Hour

1. There is a fixed bedtime ritual for the child fromwhich almost no deviation is permitted.

2. Child's bedtime hour is fixed within a narrowlimit such as 45 minutes, although the hour may

be changed to accommodate special circum-stances.

3. Child has considerable latitude in determining hisown bedtime but goes to bed before 8:30 p.m.

4. Child pretty much sets his own bedtime andoften goes to bed after 8:30 p.m.

From analysis of the data, five empirical clusters ofitems emerged for mothers, six for fathers, and five forjoint behavior:

Firm enforcement. For each parent an empiricalcluster emerged that included such items as cannot becoerced by child, forces confrontation when childdisobeys, and willingly exercises power to obtain obedi-ence.

Encourages independence and individuality. Thisempirical cluster, again common to both parents, con-tained items such as gives reasons with directives, defineschild's individuality clearly, listens to critical comments,and engages in meaningful verbal interaction.

Passive-acceptant. The items in this cluster thatemerged for each parent measured parents' reluctance toexpress anger and avoidance of open confrontation. Theparent with high scores on this cluster tended to bemild-mannered and unaggressive. Low scores on thiscluster did not signify that the parent was punitive orrejecting.

Rejecting. High scores on this cluster (unlike pas-sive-acceptant) emerging for each parent signified thatthe parent was cool, inaccessible, and unresponsive andtended to discipline harshly.

Father clusters: 1) Promotes nonconformity and 2)authoritarianism. For the father, two additional clustersemerged that did not have counterparts for mothers. Thefirst, designated promotes nonconformity, was com-posed almost entirely of items originally designed to

24 SCHIZOPHRENIA BULLETIN

measure an emphasis on encouraging individuality andeven rebellion in the child. The second, designatedauthoritarianism, was defined primarily by items measur-ing the father's attempts to promote respect for estab-lished authority and the parental role.

Mother cluster: Self<onfident, secure, potent paren-tal behavior. Items falling into this cluster were designedto assess flexibility and clarity of the mother's views andher confidence in herself as a parent. A similar clusterdid not emerge for fathers.

Five joint PBR clusters. An additional cluster anal-ysis was performed on the 25 items describing theparents' joint conduct. The resulting clusters of itemswere very close to the original grouping in terms of thehypothetical constructs. These five joint clusters weredesignated expect participation in household chores,enrichment of child's environment, directive, discourageemotional dependence, and discourage infantile be-havior.

Four types of families were defined in terms ofparental authority on the basis of their patterns of scoreson the parent behavior rating clusters. Fifty-four familiesof white boys and 48 families of white girls wereassigned to patterns. Because of the stringent ratingcriteria, not all families in the study could be assigned.These patterns of parental authority were definedtheoretically to produce contrasting groups of familiescorresponding to more refined definitions of the pat-terns (authoritative, authoritarian, permissive) describedin study 1. We focused on four theoretically interestingpatterns, two high on control and two low on control.Those high on control are designated authoritarian andauthoritative; those low on control are designatedpermissive and nonconforming.1

Type 1: The authoritarian parent. The charac-teristics of parents of this type are summarized below:

The Authoritarian parent values obedience as avirtue and favors punitive, forceful measures to curbself-will at points where the child's actions or beliefsconflict with what he or she thinks is properconduct. He or she believes in keeping the child inhis place, in restricting his autonomy, and inassigning household responsibilities in order to incul-cate respect for work. This parent regards thepreservation of order and traditional structure as ahighly valued end in itself. He or she does not

'This particular typology is not empirical, as were theparental clusters derived from the longitudinal study and de-scribed just above.

encourage verbal give and take, believing that thechild should accept the parent's word for what isright. [Baumrind 1972b, p. 179]

Parents were assigned to the authoritarian pattern onthe basis of having scores high on firm enforcement, lowin passive-acceptant, low in encourages independenceand individuality, and, for the father, either a very lowscore on promotes nonconformity or a very high scoreon authoritarianism.

Type 2: The authoritative parent. Parents of thistype can be described as follows:

The Authoritative parent attempts to direct thechild's activities in a rational, issue-oriented manner.He or she encourages verbal give and take, shareswith the child the reasoning behind a policy, andsolicits objections when the child refuses to con-form. Both autonomous self-will and disciplinedconformity are valued by the Authoritative parent.He or she exerts firm control at points of parent-child divergence but does not hem the child in withrestrictions. The parent enforces his or her ownperspective as an adult but recognizes the child'sindividual interest and special ways. The Authorita-tive parent affirms the child's present qualities butalso sets standards for future conduct. He or she usesreason, power, and shaping by regime and reinforce-ment to achieve objectives and does not base his orher decisions on group consensus or the individualchild's desires. [Baumrind 1972b, pp. 177-178]

Parents assigned to the Authoritative pattern scoredhigh in firm enforcement and low in passive-acceptant,but, in contrast to authoritarian parents, they alsoscored high in encourages independence and indi-viduality. Parents who met these criteria actually turnedout to be more directive and demanding than authoritar-ian parents.

Type 3: The permissive parent. The qualities of thepermissive parent are summarized below:

The Permissive prototype of adult control requiresof the parent that he or she behave in an affirmative,acceptant, and benign manner toward the child'simpulses and actions and that the parent be pre-sented to the child as a resource for him to use as hewishes, but not as an active agent responsible forshaping and altering his ongoing and future behavior.The immediate aim of the ideologically awarePermissive parent is to give the child as muchfreedom as is consistent with the child's physicalsurvival. Freedom to the Permissive parent meansabsence of restraint. [ Baumrind 1972b, p. 179]

As in study 1, there was no group of parents

ISSUE NO. 14, FALL 1975 25

corresponding exactly to the prototypic permissiveparent. Many noncontrolling, passive-acceptant parentswere also cool or uninvolved. The criteria for assigningparents to this authority pattern were low scores on firmenforcement, high scores on passive-acceptant, lowscores on expect participation in household chores, andlow scores on directive.

Type 4: The noncon form ing parent. The criteria forassignment to the nonconforming pattern (which wasoriginally conceived of as a variant of the permissivepattern) w&re either that both parents scored very highon encourages independence and individuality or thatthe father scored very high on promotes nonconformityand very low on authoritarianism. These parents wereantiauthoritarian and some were antiauthority, but theydid expect participation from the children in householdchores and good performance in school. Nonconformingparents were less passive and exerted firmer control thanpermissive parents.

Results. The results of this study are summarizedbelow by categories:

The ratings of boys and girls with permissive parentsdiffered somewhat. The 14 daughters of permissiveparents were not much affected by this style of parentalauthority, though they were somewhat suggestible andaimless. The boys, however, especially when comparedwith sons of authoritative parents, were not achievementoriented and tended to be hostile with peers, resistivewith adults, and rather aimless. They might be describedin the terms of Longstreth (1974) as having unsocializedpersonality problems.

The distinction between permissive and nonconform-ing parents, both relatively noncontrolling, was ofinterest particularly in the case of boys. Sons ofnonconforming parents were significantly more achieve-ment oriented and independent than were sons ofpermissive parents. By contrast, nonconforming parentsproduced daughters who were less achievement orientedand independent than daughters of permissive parents,although not to a significant degree.

When results for authoritarian parents were ex-amined, the 10 girls tended to be reasonably wellsocialized but also rather submissive, aimless, and notachievement oriented. The 16 sons of authoritarianparents were rather similar to those of permissiveparents. They were, however, somewhat more achieve-ment oriented and independent, while the 10 girls weresomewhat less so.

The children of authoritative parents were clearly

the most competent of those already discussed. For the12 boys, the profile is the exact opposite of thatassociated with permissive parents. These boys arefriendly, cooperative, tractable, and achievementoriented. They were not, however, as dominant andpurposive as we might wish. There is some indicationthat the extremely firm control of the authoritativeparents (even when compared with the authoritarianparents) impaired the development of independence inthese otherwise competent preschool boys. The sevengirls, by contrast, were markedly dominant, purposive,and achievement oriented. While friendly, they were notparticularly compliant. Authoritative parents producedthe daughters furthest from the feminine Sfereotype.

Discussion.2 The imposition of authority, evenagainst the child's will, appears beneficial to the childduring the first 6 years—referred to by Dubin and Dubin(1963) as the authority inception period. Indeed, powerserves to legitimize authority in the mind of the child, toassure the child that his parent has the power to protecthim and provide for his needs. The main way in whichparents exercise power in the early years is by manipu-lating the reinforcing and punishing stimuli that affectthe child. What makes a parent a successful reinforcingagent or an attractive model for a child to imitate is thepossession of effective power to give the child what isneeded. The parent has control over resources that thechild desires and is willing and able to provide the childwith these resources in such a manner and at such a timethat the child will be gratified and the whole familybenefited. Practically as well as morally, gratification ofthe child's needs within the realistic economy of thefamily is a precondition for the effective imposition ofparental authority. An exploited child cannot be con-trolled over a long period of time without rebellion. Theparent's ability to gratify the child and to withholdgratification, and to do so on bases that are internallyconsistent, legitimizes parental authority in the mind ofthe child.

The contrasting effects of authority viewed asjustified vs. authority viewed as illegitimate becomeparticularly apparent at adolescence. For example, Pikas(1961), in a survey of 656 Swedish adolescents, showedthat significant differences occurred in their acceptanceof parental authority, depending on the reason for thedirective. Authority that was based on rational concern

2 For a more complete discussion of the results, see Baum-rind (1971a).

26 SCHIZOPHRENIA BULLETIN

for the child's welfare was well accepted by the child,while authority based on the adult's desire to dominateor exploit the child was rejected. The former, whichPikas calls rational authority, is similar to what we havedesignated authoritative or firm control, and the latter,which he calls inhibiting authority, is similar to what weare calling restrictive-control. His results are supportedby Middleton and Putney (1963). These researchersfound that parental discipline seen by the child as eithervery restrictive or very permissive was associated withlack of closeness between parent and child and withrebellion against the parent's political views.

It is important to distinguish between the effects onthe child of authoritative vs. restrictive control. Restric-tive parents cover many areas of the child's life and needsystems with extensive proscriptions and prescriptions;they place arbitrary limits upon his autonomous strivingsto try out new skills and make decisions for himself. Ingeneral, as Becker (1964) indicates, restrictive disciplinedoes appear to lead to "fearful, dependent and submis-sive behaviors, a dulling of intellectual striving andinhibited hostility" (p. 197). If the child perceivesexpressions of parental authority as arbitrary and over-protective, that authority may not be accepted; be-havioral compliance may be accompanied by immaturityand rebellion. If the child sees parental control asjustified in terms of his own welfare, the result may beacceptance of that authority and independence in thechild.

The effects of permissive-Indulgent vs. harmoniousparental discipline should also be contrasted. Harmoni-ous parents, unlike permissive-indulgent parents, do notdispense unconditional love, although they do express ahigh level of unconditional commitment. This interestingpattern of child rearing was identified in study 2. Whilepattern membership was generally determined by mul-tiple criteria, the eight families placed in the harmoniouspattern had but one identifying characteristic incommon: The observer assigned to study each of thesefamilies would not rate the family on the construct, firmenforcement, in each case stating that any rating wouldbe misleading. The parent, while he or she almost neverexercised control, seemed to have control in the sensethat the child generally took pains to intuit what theparent wanted and to do it.

The atmosphere in these families was characterizedby harmony, equanimity, and rationality. While permis-sive parents avoided exercising control but were angryabout not having control, and authoritarian and authori-

tative parents exercised control willingly, harmoniousparents seemed neither to exercise control nor to avoidthe exercise of control. Instead they focused on achiev-ing a quality of harmony in the home and on developingprinciples for resolving differences and for balancedliving. These parents brought the child up to their levelin an interaction but did not reverse roles by actingchildishly, as did some permissive parents. Harmoniousparents lived parallel to the mainstream rather than inopposition to it.

The effects of harmonious child-rearing patterns onchildren appeared to be related to sex. The six daughtersof harmonious parents were extraordinarily competentand very similar in their scores on the child behaviormeasures. Their average Stanford-Binet \Q was 136 (thatof the entire sample was also high: 128). Compared togirls raised under the other authority patterns, these girlswere achievement oriented, friendly, -and independent.By contrast, the two boys whose parents were classifiedas harmonious, while cooperative, were notably submis-sive, aimless, not achievement oriented, and dependent.The harmonious pattern of child rearing seemed toproduce dysfunction in boys, if one can say much abouttwo cases, while the effect in girls was entirely positive(Baumrind 1971b).

Work in Progress with EarlyPrimary School Children

The children from study 2 averaged 8-3/4 yearswhen seen again. An additional 60 families are presentlybeing added to those in the core sample. The coresample comprises 89 nuclear families (38 girls and 51boys) in which the child resides with both his naturalparents and for whom data on both parents are availableat the two time periods (time 1 and time 2), plus anadditional 13 families disrupted by separation for whomdata on at least one parent are available at both timeperiods. From analyses done with the core sample ofnuclear families, there are some preliminary findings, notyet reported elsewhere, which will now be presented.

The data obtained during the child's 9th year, inaddition to interview and observational records, includemeasures of moral judgment, values, internal-externallocus of control, and social and sex role attitudes forboth the parents and the child, and, for the child only,measures of creativity and cognitive development. (Seesection entitled "Current Study1' in table 1 for asummary of these data sources.) Child ratings on the

ISSUE NO. 14, FALL 1975 27

Preschool Behavior Q-sort were based on 25 hours ofobservation, interview, and testing. Parent behaviorratings were based on 30 hours of similar scrutiny.

The findings to be presented are based upon ratingsonly and will consist of time 2 correlations betweenparent and child data, time 2 child effects associatedwith time 1 parent classification, and time 2 risk typecomparisons.

Time 2 Correlations between Parentand Child Data

Eight child behavior clusters emerged from a clusteranalysis of the child behavior ratings. These were similarfor boys and girls, except for cluster C. They defined athree dimensional space of social responsibility, socialindependence, and cognitive creativity. (See footnote 3of table 4 for descriptive cluster names.)

Table 4. Significant correlations among parent and child cluster scores at time 2(nuclear families only).1

Child clusters3

A

B

C

D

E

F

G

H

A

.30 mob

.46 fob4

.43 mog4

.35 mog

.35 mog

.33 mog

.39 fog

B

.35 mob

.31 mob

.43 fog

.28 mob

Parent clusters2

C D E F

.35 fog4

.32 fog

.31 mob

.27 mob

— .27 mob .37 mog

.37 mob.38 fob .29 fob

G

.38 fog

.32 mob

.52 fog

.29 fob

.40 fog

.28 fob

.39 mog

.39 fog

.32 mog

.35 fog

.38 fog

H

.29 mob4

.44 mob

.28 fob

I

.35 fog

.34 mog

.35 mog

.39 fog

'Correlations reported are significant at .05 or better. Study involves 37 girl and parent pairs and 51 boy and parent pairs.'Parent clusters: A = enforces/does not enforce directives; B = respects/does not respect child's reasoning ability (mother), per-

ceives own and child's individuality and identity clearly/unclearly (father); C = does not/does sex-stereotype (mother), encourages/discourages child's independence (father); D = possesses nontraditionaJ/traditionai values (mother), possesses nontraditional andnonsex-stereotypic/traditional and sex-stereotypic values (father); E = supports/rejects child; F = delegates/does not delegate house-hold responsibility to child; G = trains/does not train child cognitively; H = directs/does not direct child's regimen; I = possesses/lacksself-confidence as a parent.

'Child clusters: A = altruistic/egoistic; B = cooperative/obstructive; C = domineering/submissive (girls), dominant/submissive(boys); D = purposive/aimless; E = challenges self cognitively/avoids cognitive challenge; F = creative and differentiated/stereotypedand undifferentiated; G = socially independent/socially conforming; H = socially confident/withdrawn.

4 Fog = correlation between traits of fathers and of girls; mob = correlation between traits of mothers and of boys; fob = correla-tion between traits of fathers and of boys; mog = correlation between traits of mothers and of girls.

28 SCHIZOPHRENIA BULLETIN

From cluster analyses of the 84 parent behaviorscales rated for each parent, 9 clusters emerged. Theirnames suggest their meanings and, where they exist,mother/father differences. (See footnote 2 of table 4.)

I will begin by summarizing the statistically signifi-cant correlations between parent and child cluster scoresobtained during the child's 9th year. (See table 4 for asummary of the main relationships between scores inparent clusters and child clusters at time 2. We have notyet analyzed the correlations across time.) Correlationssignificant at < .05 in table 4 are summarized here.

At time 1, parental warmth correlated with fewindices of child behavior. Warmth has proved morepredictive at time 2. For fathers of girls, warmth (E)predicts altruistic and cooperative behavior in the child(A, B), and for mothers of boys, warmth predictspurposive child behavior and creativity (D, F).

Parental practices that are stimulating and demand-ing (assignment of household chores, training child, andfirm enforcement) and parents' self-confidence are as-sociated, as they were at time 1, with assertivenessparticularly for girls. For girls, mother's delegation ofhousehold responsibilities (F) is associated with thechild's social independence (G). Cross-sex and within-sexcorrelations directed at training the child cognitively (G)are significant in many areas of child behavior. Findingsare particularly strong for fathers of girls. For fathers ofgirls, this cluster is significantly correlated with thecooperative, purposive, challenges self cognitively, andsocially confident cluster scores of girls and for bothmother and father with scores on girls' creativity andsocial independence. For both parents of boys, cognitivetraining is positively associated with purposiveness inboys and for fathers with boys' creativity. Firm discipline(A) continues to have clearly beneficial but somewhatdifferent effects at time 2. For girls, parental firmdiscipline (A) is associated with socially confidentbehavior and maternal discipline with both assertive (infact, domineering) behavior toward peers and creativityand socially independent behavior. Parental firm en-forcement is positively related to cooperation withadults in boys. Possesses self-confidence as a parent (I) isassociated for father/daughter pairs with purposive andsocially independent behavior and for mother/daughterpairs with creative and differentiated and socially inde-pendent behaviors.

Parents' willingness to offer justification for direc-tions and to reason with the child continues, as at time

1, to be associated with cognitively creative and sociallyindependent behavior for both sexes. Respect for child'sreasoning ability (B) for mother/son pairs is associatedwith purposive, creative, and socially confident behaviorin boys and for father/daughter pairs with sociallyindependent behavior in girls.

Socially confident behavior in boys was associatedfor both parents with nontraditional values (D), formothers with respects child's reasoning ability (B), andfor fathers with encourages child's independence (C)(but boys' social independence was negatively related tomother's avoidance of sex-stereotype (C)). Girls' socialconfidence was associated with parental firm discipline(A) and maternal cognitive training (G).

Time 2 Child Effects Associatedwith Time 1 Parent

Classification

Earlier in this paper, I summarized the child effectsassociated with the patterns of parental authority thatemerged when the children were of preschool age. Wehave just completed preliminary analyses of the effectson the child of those early patterns of socialization nowthat the child is 8-3/4 years old. Statements concerningstability of child behavior, stability of parent behavior,and stability of parent-child effects will be made.Probabilities associated with statements about high andlow cluster scores are from Mests comparing the groupunder discussion to all others of the same sex in thesample. (See table 5.)

Effects of Permissive Parenting

While at time 1 girls were not much affected bypermissive parenting, at time 2 the girls reared permis-sively tend to be passive and retiring. They are coopera-tive with adults, submissive, and somewhat sociallyconforming (probability level of .07) and withdrawn(.08). Sons of permissive parents, who at time 1were rather irresponsible, aimless, and lacking in achieve-ment orientation, are at time 2 lacking in socialresponsibility and independence. But because of thesmall number of subjects and high variance within thisgroup, the differences are not significant. On theaverage, parents classified as permissive at time 1 wouldstill be classified this way. These parents do not enforce

ISSUE NO. 14, FALL 1975 29

directives; parents of girls possess low self-confidenceand do not direct the child's regimen. Mothers of girlstend not to delegate household responsibilities (.07);mothers of boys tend not to engage in cognitive training(.09). Fathers of boys tend to lack self-confidence asparents (.09).

Effects of Nonconforming Parenting

While at time 1 daughters of nonconforming parentswere less achievement oriented and independent thandaughters of permissive parents, they are now more so.At time 2, daughters of nonconforming parents are morecooperative with adults and somewhat more altruistic(.06) than other girls. Sons of nonconforming parentsare more purposive, creative and differentiated, andsocially independent than other boys, as well as some-what more dominant (.10) and more willing to challengethemselves cognitively (.09). By time 2 their parents areat least as firm enforcers as other parents in the sample.Mothers continue to respect the child's reasoning abilityand to be supportive of their daughters. Fathers of girlsclearly perceive their own and their child's individualityand identity, and they tend to possess nontraditionaland nonsex-stereotypic values (.06). Parents of boystrain their child cognitively and fathers encourage thechild's independence. With opposite-sex children, non-conforming parents are self-confident. They are the mostself-confident parents among all the parent types,although this is not statistically significant for same-sexchildren. At time 2, these parents seem more warm,rational, and individually oriented than actually "non-conforming."

Effects of Authoritarian Parenting

At time 1, daughters of authoritarian parents werereasonably well socialized, but rather submissive, aim-less, and nonachievement oriented. They are now undis-tinguished from the total sample except for beingactually higher in challenges self cognitively. Sons ofauthoritarian parents at time 1 were rather irresponsible,aimless, and nonachievement oriented. Now they aredistinguished only by being somewhat aimless (.07).Authoritarian parents of girls seem in many respects tobe opposite of nonconforming parents at time 2:Mothers of girls have little respect for the child'sreasoning ability, are nonsupportive, and do not train

their children cognitively; fathers of girls delegatehousehold responsibility to the child and tend to be highin enforcing directives (.09). Both parents of boys areundistinguished from the rest of the sample.

Effects of Authoritative Parenting

At time 1, the children of authoritative parents wereclearly the most competent. Daughters were markedlydominant, purposive, and achievement oriented. At time2, they are domineering, creative and differentiated, andsocially confident. At time 1 sons were friendly,cooperative, and achievement oriented, but not asdominant or purposive as we might wish. At time 2, theyare altruistic and cooperative. Authoritative mothers ofgirls at time 2 remain high on respects child's reasoningability, trains child cognitively, and possesses self-con-fidence as a parent. Fathers of girls also tend to trainchild cognitively (.08). Mothers of boys direct child'sregimen and are somewhat high on delegates householdresponsibility to child. Fathers of boys enforce directivesand perceive own and child's individuality and identityclearly.

Hypotheses Concerning the Family'sContributions to the Etiology of

Schizophrenic Reactions in Children

Most clinicians agree that schizophrenic reactions area form of disorganization of the personality, a failure toachieve or maintain ego integration. The critical featuresof schizophrenic reactions lie in aberrant symbolic pro-cesses through which the disturbed person alters internal-ized representations of reality in order to withdraw froma world grown untenable and from social interactionsthat precipitate insoluble conflicts for that person. Bymodifying perceptions of self and others, the distressedindividual defends a fortress in which some self-esteemand some consistency in cause-effect relations are pos-sible, even if the consensual logic of the culture must bedenied in order to do so.

Adolescence is a particularly stressful period becauseof added pressures toward the establishment of anidentity and of instrumental competencies. Symptomsof severe pathology that earlier were successfully sup-pressed or denied tend to emerge at this developmentalperiod. Schizophrenic thought processes, such as autistic

I ab

le b

. I i

me

I ch

ild

ett

ec

Tim

e 1

pare

ntal

auth

ority

typ

e A

ltru

isti

c/eg

oist

ic

Perm

issi

ve g

irls

(n =

10)

M2

48.7

SD

2 8.

1

Perm

issi

ve b

oys

(n =

7)

M

42.0

SD

11

.0

Non

conf

orm

ing

girl

s (/?

° 6

)

M

54.6

SD

6.

8

Non

conf

orm

ing

boy

s (n

= 4

)M

49

.0S

D

10.1

Aut

hori

tari

an g

irls

(n -

6)

M

47.1

SD

11

.9

Aut

hori

tari

an b

oys

(n =

12)

M

49.6

SD

10

.3

Aut

hori

tativ

e gi

rls (

n =

7)

M

50.1

SD

15

.8

its a

sso

ciat

ed

Coo

pera

tive/

obst

ruct

ive

55.4

3

7.1

45.7

10.7

55.0

3

4.3

49.8 1.9

47.3

12.7

48.7

11.6

47.8

10.9

I w

ith

tim

e 1

par

ent

cias

sii

Dom

inee

ring

(G

),do

min

ant

(B)/

subm

issi

ve

42.3

3

9.3

48.4

10.2

49.0 6.5

56.7 2.5

54.1

12.0

53.2

10.4

57.2

3

3.4

rica

tio

n.1

Tim

e 2

child

Q-s

ort c

lust

ers

Pur

posi

ve/

aim

less

45.6 9.9

47.5

13.6

51.9

11.4

63.0

3

2.8

51.7 9.6

47.1 8.6

53.3 8.7

Cha

lleng

es s

elf

cogn

itive

ly/a

void

sco

gniti

vech

alle

nge

46.6

12.1

46.3

10.8

54.1 9.2

58.1 9.4

56.4

3

6.2

48.7

10.6

50.2 8.5

Cre

ativ

e &

dif

fere

nti

ated

/st

ereo

type

d &

undi

ffer

entia

ted

46.0

11.2

44.0

12.5

49.3

13.0

59.3

3

7.2

54.5 9.9

48.6

10.5

55.0

3

5.1

Soc

ially

inde

pend

ent/

soci

ally

con

-fo

rmin

g

44.5 9.7

46.3 7.3

55.2 9.3

59.3

3

4.4

51.8

11.0

48.5 9.3

51.9 8.7

Soc

ially

conf

iden

t/w

ithdr

awn

44.2

10.5

46.0

10.5

49.3 8.4

53.7 6.1

51.1

4.8

52.6 8.7

57.9

3

9.5

LO O SCHIZOPHR z > c I— I— m —i

ISSUE NO. 14, FALL 1975 31

en inc>i din «-

co u>co r*.'in

co o>in

s *

q qco coin

<\i enin

r- egin tom

N en~ in in5 - ">n

1

o

<s8

in CNI

eri d

co CMen d

en d

8 2

cri d

q tod din t-

8 ?

en d

n

3

O) Q= 5 co<

q toin

co r-._d coin

CO CO

8?

CO CM

8?

co

co incsi oiin

co o>

co encri cri

II

M

oSi

1 J_ «o _uo I1 »

O «

z Iu Si5- c .2?

o •>3-=

HI— Q ui i "1 u

logic, object inconstancy, or disassociation, interferenoticeably with school performance in high school andcannot be overlooked by adult authorities. The dis-tressed adolescent's social adaptation is drastically im-paired by excessive impulsivity, intolerance offrustration, and inability to role play effectively.

While genetic and biochemical causes for schizo-phrenic reactions are generally acknowledged, parentaleccentricities and abnormalities are also known toaccompany the constellation of symptoms just de-scribed, even when these symptoms do not assumepsychotic proportions. We have postulated that variousparental styles of child rearing that are neither abnormalnor eccentric may also predispose a child to suchsymptoms. As explained at the outset of thirarticle, ourdecision to study the antecedents of instrumentalcompetence was based, and continues to be, on theassumption that qualities basic to instrumental com-petence are essential to the health and successfulfunctioning of the individual. Conversely, a lack of thesequalities may result in poor adjustment and functioningand low self-esteem. We are further assuming that thevarious parental styles we have identified might bepredictive of a child's risk of the extreme maladjustmentand withdrawal of schizophrenia.

Time 2 Risk Type Comparisons

Using the parent PBR cluster scores discussed above,we correlated these scores with degrees of instrumentalcompetence in the children in our current sample; thatis, all subjects typed at time 1 for whom data areavailable at time 2 and whose original families are stillintact. The correlations were done separately formothers and fathers. Characteristics of types of childrenand parents and comparisons of these types are theresult of planned comparison by f-tests of the equalityof means of the two groups under consideration at anygiven time. As is shown in table 6, we arrived at threemain categories of risk for the children based oninstrumental competence: low risk; high risk, avoidant;and high risk, irresponsible.

Low Risk Group

For both boys and girls, an empirical type wasidentified, characterized by high scores on all child

32 SCHIZOPHRENIA BULLETIN

Table 6. Time 2 parent PBR cluster scores associated with child schizophrenicrisk types.

Mother PBR clusters

Risk typeRespacts Ponessss Delegates

Enforces child's nontradi- Supports householddirectivas reasoning s tioiul child responsibility

Trainschild

Directschild's

ability

B

stereotypevaluat

D

to child

F

cognitively regimen

solf-confidencsas a parent

Low risk girls

M 1

SD 1

Low risk boys(n -16)

MSD

High risk, avoidant girls

MSD

Paar independent

MSD

Peer conforming(n-5)

MSD

High risk, avoidant boys(n-6)

MSD

High risk, irresponsible girls

MSD

5732

8.1

50.19.0

45.13

8.3

62.23.9

3931-3

6.0

47 3.8.8

49.212.6

553J

9.4

62.88.8

45.03

10.6

50.19.1

40.83

10.7

46.48.7

45.515.9

49.810.2

48.39.7

53.012.0

62.12

10.4

45.77.6

49.27.1

48.813.1

49.511.1

50.310.6

52.810.1

56.99.9

49.59.9

49.17.6

50.06.7

51.312.5

52.67.2

46.911.6

47.99.6

46.114.1

49.811.2

44.76.9

543 '7.2

49.49.6

46.511.2

51.59.1

42.43

11.9

52.412.5

46.411.8

57.8'7.6

51.010.2

49.64.7

47.83.7

51.05.3

47.88.8

50.710.9

57.88.9

50.4J

7.2

53.19.2

5 6 31.6

50.312.1

43.26.9

50.113.1

67.8*7.4

47.410.9

4633

1 0 3

51.47.9

42.613

11.4

49.513.1

5137.6

High risk, irresponsible boys( /7-12)

M 46.3SD 6.4

52.07.7

54.07.2

51.86.5

49.963

49.411.2

61.18.8

42.53-3

11.247.410.0

' M = mean; SD = standard deviation. Across-sex means are standardized to SO and standard deviations to 10.3 Mean of type is significantly different from the mean of the rest of the total sample of the same sex at .05 or better.1 Mean of type is significantly different from the mean of low risk children of the same sex at .05 or better.Note: Characterizations of types of children and parents and comparisons of these types are the result of planned comparisons by

'-tests of the equality of means of the two groups under consideration at any given time. Statements are significant atp < .05 unlessotherwise noted.

ISSUE NO^14, FALL 1975 33

Paar indapandant

MSD

Paar conforming

in-8)MSD

43.67.1

47.66.1

49.712.1

53.15.3

51.26.7

55.37.4

47.63.7

53.96.8

46.673

51.56.4

50.311.0

49.012.1

49.58.2

51.911.5

39.6'15.0

43.99.6

47 312.8

47.69.4

Ail girlsin -38)

MSO

49.3n.i

49.611.7

50.311.4

50.79.5

49.610.7

48.99.5

52.09.8

51.413.0

60.893

All boys( / J - 5 1 )

MSD

50.59.3

50.38 3

49.89.1

49.410.5

50.39.6

50.810.5

48.510.1

48.093

49.410.2

Fathar PBR dustsn

Ride typeEnforces

directives

Perceives' _ P(Encouraojes

own & child's . M J 7 ^ nontrad. &child's

. nonsex-indepen- _̂ _ctance «™°*"»c

values

Individ. &ident.daarly

A B

Supportschild

Delegateshousehold Trains Directs

respon- child child'ssibility to cognitivsly regimen

child

F G H

self-confidenceas a parent

I

Low risk girls

MSD

54.08.4

55.58.7

47.112.5

50.711.5

54.49.4

52.39.1

58.78.0

55.510.3

56.46.1

Low risk boys(n-16)

MSD

52.39.4

51.98.8

50.21 0 3

47.712.4

50.99.9

47.310.2

50.69.5

50.18.7

48.513.3

High risk, avoktant girls'

MSD

Pear independentin-A)

MSD

Paar conformingin-S)

MSD

High risk, avoidant boysin - 6 )

MSD

46.61 2 3

S U9.1

42.91 3 3

45.18.7

4 9 310.6

59.21

6.4

41.83

5.1

48.712.9

43.08.0

44.06.4

42.39.8

5 2 310.0

60.28.2

47.07.2

51.111.2

56.63.0

48.713.8

43.63

9.4

46.314.3

41.55

3.5

ABJO'

6.8

49.74.0

46.71

8 3

5 1 39 3

61.1 1

7.2

44.6 ]

1 3

55.15A

57.075

53.62 3

50.99 3

47.55.6

52.34.1

51.011.0

48.76.0

50.33.2

45/48.7

34 SCHIZOPHRENIA BULLETIN

Table 6. Time 2 parent PBR cluster scores associated with child schizophrenicrisk types. (Continued)

Riifc type

High ri ik, irresponsible girlsin - 5 )

MSO

High rWc, irresponsible boysin - 1 2 )

MSO

Pear independentin-A)

MSO

Peer conformingin - 8 )

MSD

All girlsin - 3 8 )

MSO

Allboyt(/7-51)

MSO

Enforcesdirectives

A

49.75.1

47.66.9

50.57.6

46.06.5

49.211.3

50.69.1

Perceive!own & child's

individ. &ktant.dearly

B

45.018.5

4BJ52>9.4

3BS1'IS

50.67.6

50.01 1 5

50.09.0

Encourageschild's

indepen-dence

C

44.913.6

54.66.6

53/46.1

66.37.2

47.210.9

62.18.9

Ftther PBR dusters

Possessesnontrad. &

nontex-stereotypic

values

D

49.9123

5 0 3S3

46.05.0

53.23.6

5 0 310.7

49.39.6

Supportschild

E

50.78.6

47.67.1

4 U4.1

50.76.1

61.010.6

49.29.7

Dal agateshousehold

respon-sibility to

child

F

49.43.7

51.710.5

53.013.5

51.19.6

50.610.2

50.510.6

Trains

childcognitlvely

G

53.211.6

48.410.0

48 JO10.4

48.71 0 5

52.310.6

48.39.3

Directs

child'sregimen

H

4 5 313.1

46.411.2

45.414.5

47.010 3.

61.010.8

49.29.5

Possessesserf-

confidenceas a parent

1

51.65 3

45.79.0

40.211.3

48/46 3

53.37.1

47.611.3

behavior clusters that signified that these children-designated low risk—were altruistic to peers, sociallyindependent and confident, purposive, creative anddifferentiated, and willing to challenge themselves cogni-tively. In addition, boys were cooperative with adultsand dominant, and girls were somewhat domineering(.10).

The child-rearing practices associated with these highlevels of instrumental competence differed for boys andgirls in that parents' socialization practices were mostlynormative for the sample of low risk boys while for girlsthey were distinctive. The exceptions for low risk boys

were that their mothers appeared somewhat directive(H—.08)-by virtue of the lack of directiveness in themothers of high risk boys—and that the fathers weresomewhat more perceptive of their individuality andidentity (B—.12). Parents of low risk girls, however, wereclearly firm guides to their daughters. They gave theirdaughters cognitive training (G) and directed the child'sregimen (H—.05 mother, .13 father). Their mothersdelegated responsibility to them in the household (F)and enforced directives (A), and their fathers tended todo both, but not significantly. Fathers perceived theirown and their daughters' individuality and identity

ISSUE NO. 14, FALL 1975 35

clearly (B). Their mothers, apparently aware of thesuccess of their child-rearing efforts, possessed self-confidence as parents (I). They respected their daugh-ters' reasoning ability (B) as well. Parents were assupportive (E) and conforming (D) as the average parent.Why are there different patterns of child rearing associ-ated with instrumental competence (low risk behavior)for boys and for girls in this analysis? A plausiblehypothesis is that peer group pressure and societalexpectations will produce instrumentally competentbehavior in boys without special parental action sincesuch behavior is regarded as "normal" for latency-ageboys, while high instrumental competence (indepen-dence and assertiveness) is not "normal" for girls.Parents of girls must press demands for logical thinkingand assumption of responsibility since peers and teachersfail to make these demands on girls as they routinely doon boys.

High Risk, A voidant Group

We designated as high risk, avoidant children thosewho were withdrawn. The girls in this type were alsostereotyped and undifferentiated compared to the restof the girls in the sample. Two high risk, avoidant typesemerged for girls, differing from each other in that apeer-independent group was socially independent com-pared to the rest of the sample while the other groupwas peer conforming in that they scored low on thesocial independence cluster. The peer-independent groupwas extremely stereotyped and undifferentiated, aimless,and even more withdrawn (.06) than the peer-conform-ing girls. These girls seemed to be independent of theirpeer group more from lack of contact than from apositive sense of individuality and independence.

There is some evidence that part of the difficultieswith peer interaction of peer-independent girls may havebeen due to unsureness about their sex roles, which theymight have tried to compensate for by conforming to afeminine stereotype. They were more compliant withadult authority than any other risk type (mean = 55)although the differences were not significant. Theirmothers were extreme in their rejection of sex stereo-typing for either themselves or their children (C), anemphasis that has consistently been shown to backfire inour studies. Their fathers were highly controlling of theirdaughters' regimens (H) but below average on delegatinghousehold responsibility; this appeared to reflect a lack

of confidence in their daughters' competence despite thehigh value they placed on perceiving their own and theirdaughters' individuality and identity clearly (B). Theywere also highly supportive (E). Both parents trainedtheir daughters cognitively less than parents of low riskgirls. These girls may not have been provided with strongrole models and seemed overprotected.

The peer-conforming girls avoided cognitive chal-lenges, in addition to being socially withdrawn. Theirparents seemed unable to function affirmatively asparents. They did not delegate household responsibility(F—.09 mother, .05 father). Mothers were lax inenforcing directives (A), tended not to respect theirdaughters' reasoning ability (B-.11), and had lessself-confidence as parents (I) than parents of low riskgirls. Their fathers perceived their own and the child'sindividuality and identity unclearly (B)' and did notdirect the child's regimen (H). They did not do as muchcognitive training (G) as fathers of low risk girls. Itwould appear that the parents of "peer conforming-avoidant" girls provided their daughters with littlemotivation to develop competence.

One high risk, avoidant type emerged for boys. Inaddition to being withdrawn, they were submissive,aimless, stereotyped and undifferentiated, and sociallydependent, avoided cognitive challenge, and tended tobe egoistic (.06) in comparison to the rest of the sample.They were obstructive in comparison to the low riskboys. Surprisingly, their parents showed few significantchild-rearing differences from the rest of the sample as awhole, although the fathers supported the child (E)more than those of other boys and, compared to fathersof low risk boys, were less firm enforcers (A—.12), whilethe mothers directed the boys' regimens somewhat lessthan mothers of low risk boys (H-.06).

High Risk, Irresponsible Group

We designated as high risk, irresponsible thosechildren who were egoistic in their lack of altruisticconcern for peers. Girls were socially conforming andtended to be somewhat creative and differentiated (.10).They were aimless and withdrawn in relation to low riskgirls. The parents of high risk, irresponsible girls did notdiffer significantly from the sample as a whole.

Two high risk, irresponsible types emerged for boys.Like the girls "avoidant" subtypes, one group was peerindependent and the other peer conforming. The peer-

36 SCHIZOPHRENIA BULLETIN

independent group scored quite high on challenges selfcognitively, while the peer-conforming group scored low.The peer-conforming boys were also extremely obstruc-tive toward adults and socially confident. (The peer-independent boys tended in that direction, but not sig-nificantly.)

Peer-independent boys were more dominant than thesample as a whole, but not significantly more so thanlow risk boys. They were also creative and differenti-ated. Their mothers were less firm enforcers (A) thanmothers of other boys. Their fathers were extremelyunclear about perceiving their own and the child'sindividuality and identity (B) and strongly rejected thechild (E). Father/son conflict was apparent By rejectinghis son emotionally and, perhaps competitively, thefather denied the boy's actual abilities and individuality.The son was egoistic and dominant, generalizing from hiscontentious relationship with the father and assertinghimself with his peer group since he could not do so athome.

Mothers of peer-conforming boys did not sex-stereotype (C) and tended to possess nontraditionalvalues (D—.06). Their fathers possessed nontraditionaland nonsex-stereotypic values (D) and were somewhatless firm enforcers (A—.07) than fathers of low riskboys.

Summary

A general summary of our findings is as follows:Parental behaviors that put consistent pressure onchildren to test their limits, cognitively and socially,while supporting their efforts with praise and encourage-ment, produce high levels of instrumental competence.Parents who behave in this way encourage their childrento probe reality and to meet life head on, thusfacilitating their development of instrumental com-petencies. Parents who are consistent and who success-fully enforce their directives provide the young childwith the necessary structure to facilitate effectivefunctioning and clear thinking. By keeping the channelsof communication open and by using reason when thechild balks, the parent alerts the child to the possibilityof being an effective agent of change. These are theprocesses by which shared foci of attention are estab-lished and maintained among family members in authori-tative families; they may well explain the generally

positive effects in our sample of the high parentalcontrol of authoritative parents.

More specifically I would like to offer the followinghypotheses relevant to the family etiology of schizo-phrenia:

• Extremely competent children, whom we havepostulated as being at low risk for schizophrenia, arenever products of homes in which both parents demon-strate extreme noninvolvement in caretaking (i.e., wherethey neglect both their control and support functions).However, many children from such neglectful homes donot show psychotic, delinquent, or extremely neuroticbehavior despite their adverse family conditions.

• Instrumental competence in children, particularlyin girls, is facilitated by firm parental control. (In noneof the high risk groups did either parent enforcedirectives to an extent above the mean; in the low riskgroup of girls, however, mothers were firm enforcers andboth parents were directive and delegated householdresponsibilities.)

• Marked nontraditionality or actively nonconform-ing attitudes in parents seem to be associated withcompliant and withdrawn rather than independent be-havior, particularly in girls. (Note time 1 effects on girlsof the nonconforming pattern of parental authority andthe nontraditionality of parents of peer-independentgirls in the high risk, avoidant group and peer-conforming boys in the high risk, irresponsible group.)

• Parental warmth is not linearly related to instru-mentally competent (low risk) behavior, particularly in.girls. (Note that contrary to generally held views,paternal punitiveness was positively associated withindependence in girls in the preschool studies and thatfathers of peer independent girls in the high risk,avoidant group were highly supportive.)

• Willingness to train the child cognitively is veryhighly associated with outstanding competence in thechild. (Note the positive effects of rational methods ofdiscipline in the preschool studies and the consistentlypositive effects of training the child cognitively at time2.)

References

Baldwin, A. L ; Kalhorn, J.; and Breese, F. H. Theappraisal of parent behavior. Psychological Monographs,63(4, whole no. 299), 1949.

ISSUE NO. 14, FALL 1975 37

Baumrind, D. Child care practices anteceding threepatterns of preschool behavior. Genetic PsychologyMonographs, 75:43-88, 1967.

Baumrind, D. Current patterns of parental authority.Developmental Psychology Monographs, 4(1, Part 2),1971a.

Baumrind, D. Harmonious parents and their pre-school children. Developmental Psychology, 4:99-102,1971b.

Baumrind, D. An exploratory study of socializationeffects on black children: Some black-white compari-sons. Child Development, 43:261-267, 1972a.

Baumrind, D. From each according to her ability.School Review, 80:161-197, 1972b.

Baumrind, D., and Black, A. E. Socialization prac-tices associated with dimensions of competence inpreschool boys and girls. Child Development, 38:291 -327,1967.

Becker, W. C. Consequences of different kinds ofparental discipline. In: Hoffman, M. L , and Hoffman, L.W., eds. Review of Child Development Research. Vol. 1.New York: Russell Sage Foundation, 1964. pp. 169-208.

Becker, W. C, and Krug, R. S. A circumplex modelfor social behavior in children. Child Development,35:371-396,1964.

Becker, W. C ; Peterson, D. C ; Luria, Z.; Shoemaker,D. J.; and Hellmer, L. A. Relations of factors derivedfrom parent-interview ratings to behavior problems offive-year-olds. Child Development, 33:509-535, 1962.

Block, J. The Q-sort Method in Personality Assess-ment and Psychiatric Research. Springfield, III.: CharlesC Thomas, Publisher, 1961.

Coriat, I. H. The psycho-analytic approach to educa-tion. Progressive Education, 3:19-25, 1926.

Dubin, E. R., and Dubin, R. The authority inceptionperiod in socialization. Child Development, 34:885-898,1963.

Emmerich, W. Continuity and stability in early socialdevelopment Child Development, 35:311-332,1964.

Goodman, P. Compulsory Mis-education. New York:Horizon Press, 1964.

Kagan, J., and Moss, H. A. Birth to Maturity: AStudy in Psychological Development. New York: JohnWiley & Sons, Inc., 1962.

Longstreth, L. E. Psychological Development of theChild. 2d ed. New York: The Ronald Press Company,1974.

Maslow, A. H. Motivation and Personality. NewYork: Harper & Row, Publishers, 1954.

Middleton, R., and Putney, S. Political expression ofadolescent rebellion. American Journal of Sociology,68:527-535,1963.

Naumburg, M. The Child and the World. New York:Harcourt, Brace, 1928.

Neill, A. S. Summerhill. New York: Hart PublishingCo., Inc., 1964.

Parsons, T. The Social System. New York: The FreePress, 1951.

Pikas, A. Children's attitudes toward ratjpnal versusinhibiting parental authority. Journal of Abnormal andSocial Psychology, 62:315-321, 1961.

Rogers, C. R. A Therapist's View of Personal Goals.Pendle Hill Pamphlet 108. Wallingford, Pa.: Pendle Hill,1960.

Schaefer, E. S. Converging conceptual models formaternal behavior and for child behavior. In: Glidewell,J. C, ed. Parental Attitudes and Child Behavior. Spring-field, III.: Charles C Thomas, Publisher, 1961.

Spock, B. M. 777e Common Sense Book of Baby andChild Care. New York: Duell, Sloan, and Pearce, 1946.

Spock, B. M. Baby and Child Care. 2d ed. NewYork: Pocket Books, 1957.

Spock, B. M. Raising Children in a Difficult Time.New York: W. W. Norton & Company, Inc., 1974.

Acknowledgment

The program of research discussed in this paper issupported by the Grant Foundation, Inc., and theNational Institute of Child Health and Developmentunder Research Grant HD-02228.

The Author

Diana Baumrind, Ph.D., is a Research Psycholo-gist, Institute of Human Development, and Princi-pal Investigator, Family Socialization and Develop-mental Competence Project, University ofCalifornia at Berkeley.