Perception of Father and Teacher by Pre-Adolescent Boys

This thesis examines the relationship between father perception by preadolescent boys and their per ception of anot her nonparental adult authority figure . Reflecting Piaget ' s theory that ch ildren apply perceived parental behavior patterns to other individuals (assimilation), it was hypothesized that boys ' perceptions of their fa t her would be similar to their perceptions of their teacher , a nonparental adult authorit y figure . A review of the li t erat ure established t he relevance of studying personality development through children's reports of perceptions . A review of these studies during the 1960 ' s then led to t he fo r mal statement of the hypothes is. The sub~ts were all the fi f t hand sixthgrade boys in a suburban parochial school . A modified version of the Children ' s Reports of Parent al Behavior Invent ory was admi n istered twice in a single session , one form for the father and one form for the teacher . A factor analysis of t he boys ' father and teacher perceptions was performed and compared to test t he similarity in perceptions. Within limits t he perceptions were similar. Two f act or dimensions were isolated and labeled Demanding and Accepting. They were extracted from intercorrelations for both teacher and father forms of the inventory. Further statisti~ analyses using subjects ' factor scores revealed certain real differences in boys' perceptions of the two adults . It is suggested that boys live in two dist inct contexts as their social horizon behins to emerge. Thev react in similar but not identical ways t o home and social situations. Research with more refined instruments may be necessary t o account for these. distinctions and how the child perceives them. The relevance of t he find i ngs to three theoretical issues is discussed : Erikson's stage of I ndustry v s . Inferiority; Piaget ' s theory of Assimilation; a nd Kagan's theoretical description of the acquisiti on of identification. Some limitations of the study are listed.

for both teacher and father forms of the inventory. Furthe r statisti~ analyses using subjects ' factor scores revealed certain real differences in boys' percepti ons of the two adults . It is suggested that boys live in two dist inct contexts as their social horizon behins to emerge.    However , much of the ear ly research f ailed to account for the possible effects of father ing i n the child-rearing process. It is only since t he 1950' s that the literature ha s seriously begun to include more relevant evidence on f a ther-child interaction (Becker ,196 4 ). Compar ed with the effects of maternal child-rearing practices upon pers onality development in children, t her e has been a relative neglect of sim i lar resear ch i n r egard to fathers .
In his words, "• • • the child will tend to assimilate all other individuals into his (parental) schemes" ( Piaget , 1951, p. 262  Studies have found marked differences in father concept and style among various social and economic levels. In general, lower-c lass fathers seem to want more overt dom inance and authority in family matters than middleclass fathers. But, i n fact , when father inf luence at the two levels is studied , fathers from the former social class were found to have less actual influence and effect upon family members than fathers of middle-class families (Blood and Wolfe, 1960). Perhaps the greater extent of fathe r availabili ty or a more adequate self-concept are partly responsible for t he variability in father-child relationships between various socioeconomic l evels . In terms of support and control, Walters and St innett (1971) report that decade trends show middle-class parents to be more supportive and controlling of their children than lower-class parents, with their discipline based more on pursuasion and reason than on threats of physical punishment. Also , lower-class parents seem to give comparatively greater differential treatment of male and female children than do middle-class parents .
Biller suggests the influence of father occupation upon personality development. There is substantial evidence of a relationship (e . g ., Roe, 1957 ;Hurley and Hohn, 1971; Rose and Felton, 1971), but he thinks that the over-all quality of the father-son relationship is far mor e important than specific occupational i nfluence.
Cross-cultural research, according to Biller,6 s hows that children tend to be better adjusted in societies wher e the father is actively involved in the childrearing process. Romney (1965) found that cultures with low father availability require children to be more compliant tha n do cultures with high father availability.
Chi ldren in those societies tend to be more assertive and t o develop their individual potentialities more readily.
Constitutional f a ctors are apt to be the most ambiguous element of father-son interations to understand and the most difficult to control . Physical, emotional, and intellectual differences between parent and child often r equire a gr eat deal of sensitivity and acceptance in order for relationships to be successful . It is here that the direction of influence in a parent-ch i ld interaction i s often difficult to predict a ccura tely. Kysar (1968) lends s upport to Biller ' s suggestion that both mother and f a ther ca n be inf l uenced by constitut i onally predispose d differences in children. He reports some evidence tha t f a thers are less tolerant than mothers of intellectually handicapped chil dren. A father ' s values in regard to what is socially acceptable as masculine behavior has an in-7 fluenc e on his son. If the social milieu values physical agi lity in males , it is likely for a father to expect t he deve lopment of these abil i ties in a son , especially if t he father sees such qualities in himself . Lynn (1961 and 196 2 ) comments that the male sex role is largely learned from cultural e xpectat i ons; but we suggest further that masculine roles a re perceived by the boy as they are "filtered" through his i nterpretation of his father 's personality and expectations .
In what amounts to a fourth framework, Biller integrate s the socioeconomic, cultura l expectations, and constitutional considerations into a mult idimensional factor to acc ount more specifically for s ex-role development. In Biller concludes t hat an "int egrated descri ption" of paternal influences upon masculine sex-role development is fe asible. His formulation of masculine development focuses upon the constitutional and sociocultural f a c ti or discussed above, as well as a partial reliance upon social learnirig theory .
As described by Bandura and Walters (1963) , social l earning theory incorporates most of the currently disputed theories of identification and perception i nto the model i ng concept . Modeling , according to Bandura (1971) , more precisely denotes the behavioral phenomena involved in i mitation and identification. ·Referring to various authors (Mowrer, 1950 ;Parsons, 1955;Kohlberg , 1963) , he sees a lack of distinction betwe en t hese terms and an insufficient understanding of their cause-and-effect relationship in formulating behavior . Bandura regards the term imitation to imply too narrow a behavioral repertoire and the concept of identification is too diffuse to define in working terms. Instead , he proposes the modeling concept as a much broader explanation for "matching " types of be- As discussed in the preceding section, there are difference s i n child-rearing behaviors according to soci a l class. Kohn (1959) studied social class and parental authority, and found some fundamental differences between working-class a nd middle-class parents. His data showed that working-clas s parents were more concerned with controll ing observed public behavior and more likely to punish in terms of immediate consequences. Middle-clas s parents, however, ba se their punishment on their int er-11 pretation of the behavioral intent. The former is more c oncerne d with outward responsiblity and the latter with t he development of internalized standards of conduct.
More specifically, working-cla ss fathers respond in two ways : if the child's behavior does not compel specific attent ion, they are apt to ignore it, and, if it is suff iciently disruptive, they are more like ly to use physica l punishment than middle-cla ss fathers . Neither group r esorts to physical punishment as a first course of action. They set limits and then evaluate the situation in terms of the particular context.
Kohn describes three aspects of authority in the fami ly. First, there is the relative role of the mother and father in making family decisions . Then there is the relative role of the mother and father in setting limits on ch ildren' s freedom of movement or ·activity . And third, there is the frequency with which the mother or father resort to phys ical punishment to enforce obe dience .
Dubin a nd Dubin (1965) suggest that the child ' s perception of signi fican t others ' view of himself effects his self-imag e , which then influences his social behavior. Lane (1959) found that American boys tended not to express rebellious feelings in political matters because this would damage their ''buddy" relationship with their father. He thinks that this kind of :more of a brother than a father " r elationship is more significant in forming politica l belief s tha n the social indoctrination given through the regional s ocial envir onm ent. Apparently, American boys t end to feel s ome guilt if they sense that they might want to r ebel. Mor e recently, Block (197 2 )  c onnects , we s ugges t, the e a r l ier and current lite r a ture .

Methodol.Qgy
A t raditional antag onism between ob j ective data and subje c tive experi ence has long existed in western thought. This dichotomous sepa ration in the meaning of knowledge has, in some measure, even affec ted research trends in c ontemporary socia l and behavioral sciences.
Note the distinctive approach between European subjectiv i t y and American experimental psychology. Dubin and Dubin (196 5) and Goldin (1969)  attitudes toward children was related to difficulties in father-child interaction or to marital conflict (Nichols , 1962). Yarrow and Campbell (1963) found that an adult and a child had quite different perceptions of the same stimuli.
A different approach g ives systematic attent ion to the child's actual experience of parental behav ior. This tec hnique yields more precise data on the dynamics of parent-child relationships a nd provides more i n sight into the dynamics of personality developmen t a nd t he socialization process (Dubin and Dubin, 1965). These studies consider the child ' s interpretation of a behav ior as the primary index of parental influence in paren t -child int eractions. Sieg elman (1966) found tha t pre-adolescent boys and gir ls who perceived their parents as punishing tended to be rated more introverted by their peers ; and children who were r ated as more extroverted by their peers , tended to perceive thei r par ents as more loving. In another study , the effect of parent perception appeared to relate to eventual occupational c hoice . Graduate students in psychol ogy who saw the dominant parent as acceptant were found likely to select person-oriented work in the field .
Students who saw their dominant parent as avoiding, tended t o pick less person-centered types of work (Medvene , 1969). Levenson (1973 ) found that perceived feeling s of parents as punishing or controlling is related to feelings of control from sources outside the person ad development   (Bronson et al. , 1959), measures of child adjustment (Berdie and Leyton, 1967), observers' reports of child behavior (Bronfenbrenner, 1961), school achievement (Morrow and Wilson, 1961), and differential criteria separating normal children from psychiatric patients (Wi lliams, 1958).
He thinks that many studies did not measure direct components of perceived parental behavior. Specific concepts describing the parent-child relationship were not prec isely identified and tested. Also, early studies did not sufficiently distinguis h paternal and maternal behavior. Finally, he cites the lack of differentiation between parental adjustment and marital adjustment from other parent-child interactions.
The development of high speed computers had made possible the use of a more refined statistical technique.
Factor analysis is a method of economy for determining the number and nature of underlying variables, or factors, from a larger number of correlations (Kerlinger , 1964, PP· 650-652) . Factors clarify the fundamental nature of varia bles and the relationship between them (Fruchter and Jennings, 1962). Zuckerman et al. (1958) and Schaefer (1961) used factor analys is to study parental attitudes in chi ld rearing. Milton (1968) used the technique to study actual child-rearing behavior . Siegelman (1965 and and Schaefer (1965b) have used factor analysis to study child- There is a discrepancy between Siegelman's demanding and Schaef er's third dimension, firm control vs. lax control.
However, dispite the minor differences in emphasis, Goldin (1969) concludes that both investigators tend to account for the same reported parental behaviors.
Some recent research has attempted to extend factor analytic methods in this area. Burger and Armentrout (1971) and Burger et al. (1973) report methods for estimating factor scores for reports of perceived parental behavior. Factor score refers to an invididual subject 's relative score contributing to each factor extracted from the intercorrelations. It is developed from the factor matrix and is calculated by an appropiate weighing of an individual ' s score on the original variables (Fruchter and Jennings , 1962). Assuming the stability of factor dimensions such as those isolated by the Children ' s Reports All the subjects were white and the sample was relat ively homogeneous in socio-economic status. Most of the fathers worked in skilled labor occupations ; a nd only five we re in occupations which require a college level educat ion . One s ubject was fatherless, but it was decided to include him , since he had been so for less than two years .
Including the subject, the mean number of children i n each family was 3.9 • None of the subjects was without sibling s. All fifty-one subjects were tested at the same time .
Part One (Fa ther Form) of the inventory was distributed by the wr i ter and two mal e assistants, one a graduate student and the other a n undergraduate . An initial introducti on i ndicated to the subjects tha t they were being a ske d t o help in a special project being conducted by graduate students a t t he University of Rhode Island . With the perm ission of the school officials, they were being a sked to t ell s ome r eal facts about t heir lives bot h at home and a t school.
I t was emphasized that the exercise wa s not a t est i n the usual sense. There were no "right" or "wrong " answers. No one would see their answers except those working on the project and the electronic computers that would help to figure out the results. Individual answer sheets were identified only by numbers.

27
A personal data sheet was completed giving such informa ti on as age, grade, number of siblings, and father's occupation. Then specific directions for Part One were given along with three practice items.
Each item was read aloud by either the writer or one of the assistants alternating every third page. The subjects were timen time to circle the appropiate response alternative for each item .
After Part One of the inventory was completed the subjects were given a twenty minute rest period. Then Part Two (Teacher Form) was distributed and appropiate direc tions given . It was explained that this form was similar to the first one, but that the items referred to their feelings about school, and specifically teachers.
While doing each item , they were asked to think of their Gym teacher. Items were written with reference to him rather than the father as in Part One.
It is the situation of the male gym teacher which made this sample population an ideal one for the present study. The subjects have contact with him at least one period a week; and, since he also teaches science to higher grades , there is some interaction with him at other times during the school week. He is the only male teacher in the scho ol a nd, except for occasional visits by the parish priests , the subj ects do not have regular contact with other male authority figure s . The wr iter has observed the interaction between the teacher and the sub jects on several occasions both during gym period and other school situations .
Upon completion of Part Two, the inventory was collected and the s ubjects returned to their classr ooms .
Total time for the entire testing procedure was approximately two a nd one-half hours .

ANALYSES
A scoring procedure similar to earlier s t udies using the CRPBI was used . Nominal values of three, two, and one were assigned to responses of "Like , Somewhat Like, and Not Like my Father/Teacher," respectively.
For each sub ject there was a father and teacher score sheet matching the two forms of the inventory. Since  (Harman, 1960 , p. 132 ). Factor ma trices were obtained and t hese were rotated orthogonally by the varimax method. Factor scores were also obtained using the BMD08M Factor Analysis Program .
The same analyses were performed on the 18 scales from the teac her inventory . The r esults of the factor analyses were then compared to test the hypothesis.

RESULTS
The separate intercorrelations of the 18 scales for the father and teacher inventories yielded two similar and corres ponding factor matrices which were then rotate d or- Kendall rank correlation coefficients (tau) were obt ained between corresponding matrices for the first and second father-teacher factor s , r espe ctiv ely . Similar correlations were calculated between these results and the f i ndings of an earlier s t udy (Burger and Armentrout, 1971). The previous study had extr acted a third factor, but only the f i rst t wo were used for comparison with the two factors is olated in the present study . (Table II).
The ( Burger and Armentrout and Arsenault (Father) .81 Burger and Arm.entrout and Arsenault (Teacher) • 7 2 Arsenault (Father) and Arsenault (Teacher) . 83 \..,.) \..,.) each father factor, the subjects were rank ordered by factor scores a nd divided into three equal groups of 35 17 subjects each. Kendall rank order correlations were calculated between these groups and three groups of each subjec t's corresponding teacher factor score.
The ( Rejection ; Psychological Autonomy vs . Psychological Control, a nd Firm Control vs . Lax Control (1965b ) . Since t hen , s tudies using a revised version of the CRPBI have yielded the same general factor dimensi ons .
As previously discussed, Armentrout (1970 ) questions the advisability of using the three dimension factor structur e with subjects below the seventh grade. His findings with boys and girls f rom a middle-class subur ban community tend to show that t he ex tent of perceive d parental control was associated with both the degree of perceived parental rejection and perceived firmness of parental discipline . He suggests that young children ' s views of their partnts' behavior are not sufficiently di fferenti ate d to s upport a more complex three-dimensiona l factor structure , which might be more appropiate for the percept i ons of older children and adults.  (Schaefer , 1965b).
Apparently elements from the second and third factors of previous studies were coupled into the first factor of the present study. The perceptions are stronger and more unpleasant than the "psychological " character of a similar factor in other studies. The label for thi s factor will have to reflec t the subjects' inability to perceive distinctions between overt limit setting behaviors and more subtle covert behaviors .
The first factor is called Demanding. Roe and Siegelman (1963 ) describe demanding parents, and we suggest other a uthority figures too, as requiring high standards in act ivities , imposing strict regulations, demanding unquestioning obedience, hugh punitiveness, limiting friend-ships, and making little effort to discover a child's feelings about things.

38
The second factor loads heavily on more nurturant and positive perceptions. It is very . similar to the first factor generally found in studies using the CRPBI.
This factor is called Accepting.
Despite the perceived father-teacher similarity in the fac tors, a close examination of the factor loadings sugge s ts some subtle differences. Neither factor in the teac her matrices have heavy loadings with negative sign values . The uni-polar nature of these loadings is evident when compared with the negative loadings in the two father matrices. In terms of perceiving a full range of behavi ors on a pos itive-to-negative continuum, the teacher is seen with a sense of attenuation by the subjects as compared to their perceptions of the father . The teacher is not seen as accepting or demanding as strongly. The perceptions of the teacher appear more functionally oriented, while the percept ions of the father are more expressive in nature .
This attenuation in the perception of the teacher may reflect the greater emotional intensity of a child's oneto-one relati onship with the parent. At home the child is one of a few at most, if t here are siblings. At school he is one of many and the teacher-child interactions may be "diluted" in intensity.

39
The data i n Table IV s hows the effect of "diluted " percepti ons by the su'Qjects within the family. Boys with the greater number of siblings saw their father more strongly on ,the Demanding or functional factor, as a teacher might be seen in school. In comparison, boys with only one s ibling perceived their father as more Accepting or expressive, . denoting a more one-to-one kind of relationship. These findings suggest that not only pers onality a nd behavior of the pa rent is salient to parental percepti on by children, but also the number of. siblings with which the parent must interact .in the fami ly. Further resea rch with a more refined instrument may account f or these variables.

VI CONCLUSIONS
The findings of this s t udy a re relevant to some important theoretical issues in child development. A child' s perceptions of hi s parents are salient to personality development at every age. But the pre-adolescent years may be more crucial t han is commonly thought because this period is a transitional one from early childhood to adolescence. Erikson (1963, pp. 258-261) describes the school age child as beginning to be a "worker ;" as he realizes that if he is to survive he must be able to exist and progress outside the family. He learns to win recognition by his ambition and industry in producing things. A tendency for ambivalence toward home life may begin at this time. The child knows he is accepted a t home but resents parental control of his energies. It is this aspect of his relati onship with his parents which, in his estimation, dominate s. Sexual latency is a time of '~isch ievous " achievement prepa ring the child to handle the identity problems to be encountered during adolescence. The child must learn to balance a need for acceptance at home with the demands of living in a larger society.
This sense of tension between the home and the emerg-ing social horizon is suggested by the factor score data.
Boys live in two distinct contexts and they react in similar but not completely identical ways to both situations.
By late pre-adolescence the child should have established mental structures which are sufficiently stable for dealing with the environment . The practice of "assimilation" begun during infan ce should be well perfected (Ginsburg and Opper, 1963. p . 18). The way a child sees a parent should be relatively stable, in order that this

4.3
It is suggested that the findings might be relevant to the acquisition and maintenance of identification as described by Kagan (1958). Note the prevalence of parent perception type notions. The first criterion is Results may also be considered as tentative due to the possibility of subject fatigue over the testing period, length of the instrument , and failure to counterbalance the administration of the Father-Teacher forms of the inventory .