6
SOCIAL LEARNING THEORY OF THE FAMILY: AN ANALYSIS KEVIN MARJORIBANKS The University of Adelaide, Auatralia Recent revearch which has investigated relationships between family environ- ments and the cognitive performance of children has lacked a sound theoretical framework and has trivlalized the nature of the relations between structural social forces and social- sychological environment variables. In the present study, a social learning tieory of the family which links cognitive performance with distal and proximal environment variables is tested. Factor scaling is used to construct new environment scales to measure the proximd learning environment of the family. Canonical analysis is used to examine the theory. The findings suggest that our understanding of the differences in the cognitive performance of children will be enriched by starting from a theoretical per- spective which analyzes relationships between structural social forces and the family learning environment defined in terms of parents’ expectations and the cognitive stimuli and reinforcement provided in families. At present our understanding of the mechanisms by which family environments are related to children’s cognitive performance is extremely limited (e. g., Bron- fenbrenner, 1974; Jencks, 1972; Jensen, 1973; Moos & Insel, 1974; Schulman, 1970; Walberg, 1971). A review of environmental research (e. g., Bloom, 1964; Dave, 1963; Marjoribanks, 1972a; Mosychuk, 1969; Plowden, 1967; Vernon, 1969; Williams, 1973, 1974) suggests that the investigations have not been theoretically grounded and have failed to identify the important dimensions of family learning environments. Williams proposes that future investigations of relations between family environments and the cognitive performance of children should be guided by a social learning theory of family environments (also, see Bijou, 1971). The theory proposes that differences in the cognitive performance of children are related to between-family variations in the expectations that parents have for childrens’ behaviors associated with cognitive performance, and in the kinds of cognitive stimuli and reinforcement provided within the family. Although the theory attempts to isolate significant environment dimensions, it examines only proximal social-psychological variables and ignores distal social measures such as social status characteristics and family structure indicators. Bernstein and Davies (1969) and Halsey (1975) suggest that a theory of learning environments is inadequate if it fails to include those structural forces that make up social class. Halsey proposes that . . . . .it is essential to insist that the effect of social class on educational ex- perience is not to be thought of as one factor from which parental attitudes and motivations to succeed are independent. A theory which explains ed- ucational achievement as the outcome of a set of individual attributes has lost the meaning of those structural forces which we know as class. (p. 17) Therefore, it would seem appropriate to include both structural social forces and social-psychological environment variables in a social learning theory of the family. The present study examines relationships between the cognitive performance Requests for reprints should be sent to Kevin Marjoribanks, Dept. of Education, University of Adelaide, Adelaide, South Australia 5001. 457

Social learning theory of the family: An analysis

Embed Size (px)

Citation preview

Page 1: Social learning theory of the family: An analysis

SOCIAL LEARNING THEORY OF T H E FAMILY: AN ANALYSIS KEVIN MARJORIBANKS

The University of Adelaide, Auatralia Recent revearch which has investigated relationships between family environ- ments and the cognitive performance of children has lacked a sound theoretical framework and has trivlalized the nature of the relations between structural social forces and social- sychological environment variables. In the present study, a social learning tieory of the family which links cognitive performance with distal and proximal environment variables is tested. Factor scaling is used to construct new environment scales to measure the proximd learning environment of the family. Canonical analysis is used to examine the theory. The findings suggest that our understanding of the differences in the cognitive performance of children will be enriched by starting from a theoretical per- spective which analyzes relationships between structural social forces and the family learning environment defined in terms of parents’ expectations and the cognitive stimuli and reinforcement provided in families.

At present our understanding of the mechanisms by which family environments are related to children’s cognitive performance is extremely limited (e. g., Bron- fenbrenner, 1974; Jencks, 1972; Jensen, 1973; Moos & Insel, 1974; Schulman, 1970; Walberg, 1971). A review of environmental research (e. g., Bloom, 1964; Dave, 1963; Marjoribanks, 1972a; Mosychuk, 1969; Plowden, 1967; Vernon, 1969; Williams, 1973, 1974) suggests that the investigations have not been theoretically grounded and have failed to identify the important dimensions of family learning environments. Williams proposes that future investigations of relations between family environments and the cognitive performance of children should be guided by a social learning theory of family environments (also, see Bijou, 1971). The theory proposes that differences in the cognitive performance of children are related to between-family variations in the expectations that parents have for childrens’ behaviors associated with cognitive performance, and in the kinds of cognitive stimuli and reinforcement provided within the family.

Although the theory attempts to isolate significant environment dimensions, it examines only proximal social-psychological variables and ignores distal social measures such as social status characteristics and family structure indicators. Bernstein and Davies (1969) and Halsey (1975) suggest that a theory of learning environments is inadequate if it fails to include those structural forces that make up social class. Halsey proposes that

. . . . .it is essential to insist that the effect of social class on educational ex- perience is not to be thought of as one factor from which parental attitudes and motivations to succeed are independent. A theory which explains ed- ucational achievement as the outcome of a set of individual attributes has lost the meaning of those structural forces which we know as class. (p. 17)

Therefore, it would seem appropriate to include both structural social forces and social-psychological environment variables in a social learning theory of the family. The present study examines relationships between the cognitive performance

Requests for reprints should be sent to Kevin Marjoribanks, Dept. of Education, University of Adelaide, Adelaide, South Australia 5001.

457

Page 2: Social learning theory of the family: An analysis

458 Psychology in the Schools, October, 1976, Vol. 13, No. 4.

of children, structural characteristics of family environments, and the social psy- chological environment of the family defined in terms of parents’ expectations and the kinds of cognitive stimuli and reinforcement provided within the family.

METHOD In order to test the social learning theory, new scales were developed to mea-

sure parents’ expectations and the stimulus-reinforcement variable. From research (e. g., Bernstein, 1961; Bing, 1963; Bloom, 1964; Coleman, 1966; Plowden, 1967; Strodtbeck, 1958; Weiss, 1969; Wiseman, 1967) which had investigated relations between family environments and the cognitive performance of children, Marjori- banks (1972,b) collected a pool of 200 environment items. Six-point rating scales were devised for each item. An interview schedule, including the 200 items, was constructed and administered to the parents of 185 families in Southern Ontario, Canada. The families were assigned to two categories, one classified as middle class and the other as lower class. An equally weighted combination of the oc- cupation of the head of the household and a rating of his (or her) education was used to assess the social class classification.

In the final sample there were 90 families classified as middle class and 95 classified as lower class. Each family had an ll-year-old boy attending school, and most of the items in the interview schedule assessed relationships between the parents and the boy. Also, each boy in the sample completed the SRA Primary Mental Abilities Test, which provides verbal, number, spatial, and reasoning ability scores.

In the present analysis the family interview data and the mental ability scores were used to test the social learning theory of the family. From the responses to the 200 environment items it was decided to construct new environment scales to assess parents’ expectations and the stimulus-reinforcement variable. The structural social forces used to define the family environment included father occupation, education of mother and father, family size, and crowding ratio of the home.

Learning Environment of the Family Scales Two sets of items were selected from the pool of 200 items. One set consisted

of 30 items which in terms of face validity were considered to assess parents’ ex- pectations, while 80 items were chosen as potential measures of cognitive stimulus and reinforcement in the family. Factor scaling (Armor, 1975) was used to construct the new environment scale8 and also to estimate the theta reliabilities of the scales. Within the two categories of items the responses were factor-analyzed using princi- pal-component analysis with varimax rotations. In the parents’ expectations set of items, two factors were identified. After eliminating those items with low factor loadings (< .40) and removing items which loaded strongly on both factors, 13 items remained. Principal-component analysis was used to refactor the 13 items, and two factor scales were isolated which were labeled parents’ expectations for the child and parents’ expectations for themselves.

When principal-component analysis was used to examine the stimulus-re- inforcement items, four factors were identified after rotation. Again, items with small loadings (< .40) and items which appeared on more than one factor were

Page 3: Social learning theory of the family: An analysis

Social Learning Theory of the Family 459

eliminated. The remaining items on each factor were refactored, which produced four clearly defined factors, and they were labeled parents’ concern for the use of English, parents’ knowledge of child’s educational progress, family involvement in educational activities, and parents’ reinforcement of educational expectations.

Because factor scaling guarantees relatively equal factor loadings on scales, the score for each environment scale was obtained by summing the item scores on each scale. The new environment schedule consists of 50 items collapsed into six scales. The environment scales, their theta (0) reliabilities, the number of items in each scale, and the means and standard deviations of the scores on each scale are shown in Table 1.

TABLE 1. ENVIRONMENT SCALES, RELIABILITIES. MEANS, AND STANDARD DEVIATIONS

Environment Scale Theta (e) Number reliability of items

Parents’ expectations for the child .94 10 Parents’ expectations for themselves .74 3 Parents’ concern for the use of English .93 12

Family involvement in educational activities .a1 10 Parents’ reinforcement of educational expectations .90 9

Parents’ knowledge of educational progress .91 6

Mean

32.9 8 . 9

29.3 17.7 20.8 29.6

Standard deviation

11.6 2 . 6

13.0 6 . 6 7 . 9 8 . 1

Note: 8 refers to the reliability of the refactored scales.

RESULTS AND DISCUSSION The results in Table 2 show the zero-order correlations between the environ-

ment variables and the mental abilities. For verbal and number abilities, the six environment scales and the social status and family structure variables have moderate to high concurrent validities, while for reasoning and spatial abilities, the environment variables have low to moderate concurrent validities. When the environment variables are combined into a set of predictors, they account for a large percentage of the variance in verbal and number scores and a moderate percentage of the variance in spatial and reasoning scores. Instead of examining the zero-order correlations in greater detail, the covariate structure of the relations between the environment and mental abilities was analyzed using canonical analysis. (Bock & Haggard, 1968; Darlington, Weinberg, & Walberg, 1973 ; Hotelling, 1935).

In the canonical analysis, the six environment scales, social status character- istics, and family structure indicators formed a set of predictor variables and the four mental abilities formed a set of criterion variables. The analysis revealed that the first two canonical correlations, .788 and .437, were significant (probabili- ties less than .001).

Relations between the environment variables, the mental abilities, and the two canonical variates are shown in Figure 1. The values used in the figure, canoni- cal loadings, are the correlations between the original measures and canonical variates of the same battery: predictor measures with predictor canonical variates and criterion measures with criterion canonical variates.

For a solution which is easier to interpret, the first variate has been rotated orthogonally through number ability. The canonical loadings show that with

Page 4: Social learning theory of the family: An analysis

460 Psychology in the Schools, October, 1976, Vol. 13, No. 4.

TABLE 2. CORRELATIONS BETWEEN ENVIRONMENTAL VARIABLES AND MENTAL ABILITIES - Mental Abilities

Environment Variables ~~

Verbal Number Spatial Reasoning

Parents’ expectations for child .64** .67** .21** .35** Parents’ expectations for themselves .34** .29** .23** .23** Parents’ concern for English .48** ,25** .19** .27** Parents’ knowledge of educational

progress .56** .47** .06 .19** Family involvement in activities .54** .44** .26** .30** Parents’ reinforcement of expectations .58** .63** .31** .36** Father occupation .43** .30** .31** .29** Father education .29** .27** .26** .22** Mother education .39** .33** .21** .16* Fi+mily size - .32** - .33** - .04 - .03 Crowding ratio - .34** - .34** - .07 - .09

100Rg 64.76 48.16 15.36 20.80

* D < .05 **p < .01 100R* indicates the total amount of variance in the mental abilities associated with the environ-

ment variables.

respect to the rotated first canonical variate, high scores on number a i d verbal abilities are associated with high scores on parents’ expectations for the child, parents’ reinforcement of educational expectations, parents’ knowledge of child’s progress, and related to low scores on family size and crowding ratio of the home. Loadings on the first variate also show that verbal and number ability scores are related more highly to the environment variables than are reasoning and spatial abilities.

After removing the variance of the first canonical variate from predictors and criteria, the loadings on the second rotated variate reveal that the social status characteristics and environment scales are related to differentially developed abilities. High scores on parents’ concern for English, father occupation, and family involvement in educational activities are associated with high scores on verbal, spatial, and reasoning abilities, but associated with lower number ability scores. That is, the second canonical variate suggests that the development of discrepant verbal, spatial, and reasoning abilities in relation to number ability might be facilitated in families characterized by high parent-son interaction. The inter- actions take place in activities such as reading, conversations at mealtimes and in the evenings, and outings with the family, and they involve a purposeful teaching of vocabulary and the correction of syntactical errors in language use.

The canonical analysis suggests that global environment variables, such as social status characteristics and family structure indicators, and social-psychological environment variables operate together to develop certain potential cognitive abilities and leave others relatively underdeveloped (Walberg & Marjoribanks, 1973). Therefore, the canonical analysis provides initial support for the social learning theory of the family by showing that cognitive performance is related to:

Page 5: Social learning theory of the family: An analysis

Social Learning Theory of the Family 46 1

I

Number

I .C

- 5 X prents? e:

for child .8

upare]

-.4

-.5

-.6

mental ab i l i ty

X enviroment scale

-.7

A soc ia l structural variaBle

ctat ions *verhal

' reinforoment

x family lnvolvaent

xconcem fo r English Reasoning.

;ion. A father occupation Xparenta, expectations

for themselves

fa& Spatial I' education

/ .1 .2 -3 .4 .5 .6 .7 I1

fa& Spatial education

.1 .2 .3 .4 .5 .6 .7 I1 '

FIGURE 1. Canonical Loadings of Environment Variables and Mental Abilities.

(a) the expectations that parents have for intellectually relevant behaviors for themselves and for their children, the cognitive models that parents provide as one means by which appropriate behaviors can be learned, the opportunities that are provided for the learning and practice of the behaviors, and the reinforcement that the family provides for the behaviors, and (b) the interrelationships between the global and social psychological environment variables.

The present study goes beyond much of previous research which has investi- gated the environmental correlates of cognitive performance. New environmental scales have been developed from a theoretical framework which suggests that family environments can be investigated from a social learning theory perspective with dimensions involving expectations, cognitive stimuli and reinforcements as the broad categories of behaviors that vary between families. Canonical analysis shows that the structural social environment variables should not be trivialized, as has been the case in much previous research, to the point where differences in parental attitudes and behaviors are conceived of as separate factors, rather than

Page 6: Social learning theory of the family: An analysis

462 Psychology in the Schools, October, 197’6, Vol. 13, No. 4.

as an integral part of the social situation of children. If the social principles that are related to the shaping of attitudes and behaviors are underestimated, then it is likely that attempts to influence parental attitudes and behaviors and eventually the cognitive performance of children may be made much more difficult.

REFERENCES ARMOR, 1). J. Theta reliability and factor scaling. In H. L. Costner (Ed.) Sociological methodology

1973-1974. San Francisco: Jossey-Bass, 1974, 17-50. BERNSTEIN, B. Social structure, language and learning. Educalional Research, 1961, 3, 163-176. BERNSTEIN, B., & DAVIES, B. Some sociological comments on Plowden. In R. Peters (Ed.) Per-

spedives on Plowden. London: Routledge & Kegan Paul, 1969, 55-83. BIJOU, S. W. Environment and intelligence: A behavioural analysis. In R. Cancro (Ed.) Intelli-

g e m : Genetic and environmental injhences. New York: Grune & Stratton, 1971, 221-239. BING, E. Effect of child-rearing practices on the development of differential cognitive abilities.

Child Development, 1963, 34, 631-648. BLOOM, B. S. Stability and change in human characteristics. New York: Wiley, 1964. BOCK, R. D., & HAGGARD, E. A. The use of multivariate analysis of variance in behavioural re-

search. In D. K. Whitla (Ed.) Handbook of measurement and assessment in lehuvioural sciences. Reading, Mass. : Addison-Wesley, 1968, 100-142.

BRONFENBRENNER, U. Developmental research, public policy, and the ecology of childhood. Child Development, 1974, 46, 1-5.

COLEMAN, J. S., et al. Equality of edwutionat opportunity. Washington, I ) . C.: U.S. Government Printing Office, 1966.

DARL!NQTON, R., WEINBERG, S., & WALBERO, H. J. Canonical variate analysis and related tech- niques. Review of Educational Research, 1973, 43, 433-454.

D.~vE, R. H. The identification and measurement of environmental process variahles that are related to educational achievement. Unpublished doctoral dissertation, University of Chicago, 1963.

HALSEY, A. H. Sociology and the equality debate. Oxford Review o/ Education, 1975, 1, 9-23. HOTELLING, H. The most predictable criterion. Journal of Educalional Psychology, 1935,26, 139-142. JENCKS, C. Inequality: A reassessmeni of the effect of family and schooling in America. New York:

Basic Books, 1972. JENSEN, A. R. Educability and group differences. London: Methuen, 1973. MARJOIUBANKS, K. Ethnic and environmental influences on mental abilities. American Journal

of Sociology, 1972(a), 78, 323-337. MARJORIBANKS, K. Environment, social class and mental abilities. Journal of Educational Psychology,

1972(b), 63, 103-109. Moos, R. H., & INSEL, P. M. Issues in social ecology. Palo Alto, Cal: National Press Books, 1974. MOSYCHUK, H. Unpublished doctoral

NUNNALLY, J. C. Psychometric theory. New York: McGraw-Hill, 1967. PLOWDEN, B., et al. Children and their primary schools. London: H.M. Stationery Office, 1967. SCHULMAN, L. S. Reconstruction of educational research. Review of Educational Research, 1970,

40, 317-396. STRODTBECK, F. L. Family interaction, values and achievement. In D. C. McClelland (Ed.) Talent

and Society. Princeton: Van Nostrand, 1958, 135-194. VERNON, P. E. Intelligence and cultural environment. London: Methuen, 1969. WALBERG, H. J. Models for optimizing and individualizing school learning. Interchange, 1971, 2,

15-27. WALBERG, H. J., & MARJORIBANKS, K. Differential mental abilities and home environment: A

canonical analysis. Developmental Psychology, 1973, 9, 363-368. WEISS, J. The identification and measurement of home environment factors related to achievement moliva-

tion and self esteem. Unpublished doctoral dissertation, University of Chicago, 1969. WILLIAMS, T. Abilities, environment and attainments. Paper presented a t the New York meetings

of the American Sociological Association, 1973. WILLIAMS, T. Competence dimensions of family environments. Paper presented at the Chicago meetings

of the American Education Research Association, 1974. WISEYAN, S. The Manchester survey. In B. Plowden, Children and their primary schools. Vol. 2.

London: H.M. Stationery Office, 1967, 347-400. WOLF, R. M. The identification and measurement of environmental process variables related to intel-

ligence.

Differential home environment and mental ability patterns. dissertation, University of Alberta, 1969.

Unpublished doctoral dissertation, University of Chicago, 1964.