the effect of social influence on cognitive development and school performance

12
Scandinavian Journal of Psychology, 1989, 30, 52-63 The effect of social influence on cognitive development and school performance ELLEN HARTMAN”, TURID JOHANNE ERI’ and ANNE HELENE SKINSTAD’ 1Department of Psychology, University of Oslo, Oslo, Norway 2Vestlandets Nervesanatorium Solli, Bergen, Norway ’Department of Clinical Neuropsychology, University of Bergen, Bergen, Norway Hartmann, E., Eri, T. J. and Skinstad, A. H. (1989). The effect of social influence on cognitive development and school performance. Scandinavian/ournalof Psychology, 30.51- 63. Three months before school entrance a sample of 29 children and their mothers was tested for degree of deccntred child educability and degree of decentred maternal teaching. Mother and child were tested in two different situations, thus preventing interdependency between the measures of mother and child. Four months after school entrance, teacher judgements of school performance were obtained. A strong correspondence between degree of decentred child educability and degree of decentred maternal teaching was demonstrated. Degree of decentred maternal teaching and degree of decentred child educability were found to be good predictors of school performance, accounting for respectively 45 and 33% of the variance in school performance. In contrast a test of school readiness only accounted for 2% of the variance. A test of intelligence given after the teacher judgement accounted for 31% of the variance. The fact that the mother seems LO be a better predictor of her child’s school performance than the child himself, supports the assumption that parents, particularly mothers, are important mediators between the child and the outer world. Key words: Social influence, cognitive development, decentration, maternal teaching, child educability, school performance, mother-child interaction. E. Hanmann, Department of Psychology, University of Oslo, Box 1094, Blindern, N-0137 Oslo 3, Norway. Children differ widely in their intellectual and social achievements in school. Since the early 1960s a considerable amount of research has documented that environmental variables like social class and family background influence academic attainment in school and into adult- hood (Coleman, 1966;The Consortium for Longitudinal Studies, 1983;Cook-Gurnperz, 1973; Deutsch, 1973; Henderson, 1981; Hess, 1970; Laosa, 1981). Studies in Scandinavia have also demonstrated a significant relationship between social background and school performance (Hernes & Knutsen, 1971; Sandven, 1969; Svendsen, 1978). Different approaches have been used in order to identify the functional elements in these relationships. One of the main research traditions has been based on the view that, within certain biological limits, and interactive with genetic factors, human interaction and communi- cation influence cognitive development. Within this tradition a main assumption has been that intellectual development advances through dialogues between parent and child (Bee et al., 1969; Bernstein, 1971,1973,1975; Bernstein & Lundgren, 1983; Hartmann & Haavind, 1981; Hess & Shipman, 1965, 1967, 1968; Kamii & Radin, 1967; Laosa, 1981; Luria & Yudovich, 1959; Robinson & Arnold, 1977; Vygotsky, 1962; Wood & Middelton, 1975). Several studies have attempted to identify the characteristics of parent-child communication associated with variations in intellectual functioning and use this to explain the low school performance of many socioculturally disadvantaged children (Bereiter & Englemann, 1966; Deutsch, 1973; Gray et al., 1982; Hess & Shipman, 1965, 1967,1968; Kamii & Radin, 1967; Schweinhart & Weikart, 1983).

Upload: ellen-hartmann

Post on 30-Sep-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: The effect of social influence on cognitive development and school performance

Scandinavian Journal of Psychology, 1989, 30, 52-63

The effect of social influence on cognitive development and school performance

ELLEN HARTMAN”, TURID JOHANNE ERI’ and ANNE HELENE SKINSTAD’ 1Department of Psychology, University of Oslo, Oslo, Norway 2Vestlandets Nervesanatorium Solli, Bergen, Norway ’Department of Clinical Neuropsychology, University of Bergen, Bergen, Norway

Hartmann, E., Eri, T. J . and Skinstad, A. H. (1989). The effect of social influence on cognitive development and school performance. Scandinavian/ournalof Psychology, 30.51- 63.

Three months before school entrance a sample of 29 children and their mothers was tested for degree of deccntred child educability and degree of decentred maternal teaching. Mother and child were tested in two different situations, thus preventing interdependency between the measures of mother and child. Four months after school entrance, teacher judgements of school performance were obtained. A strong correspondence between degree of decentred child educability and degree of decentred maternal teaching was demonstrated. Degree of decentred maternal teaching and degree of decentred child educability were found to be good predictors of school performance, accounting for respectively 45 and 33% of the variance in school performance. In contrast a test of school readiness only accounted for 2% of the variance. A test of intelligence given after the teacher judgement accounted for 31% of the variance. The fact that the mother seems LO be a better predictor of her child’s school performance than the child himself, supports the assumption that parents, particularly mothers, are important mediators between the child and the outer world.

Key words: Social influence, cognitive development, decentration, maternal teaching, child educability, school performance, mother-child interaction.

E. Hanmann, Department of Psychology, University of Oslo, Box 1094, Blindern, N-0137 Oslo 3 , Norway.

Children differ widely in their intellectual and social achievements in school. Since the early 1960s a considerable amount of research has documented that environmental variables like social class and family background influence academic attainment in school and into adult- hood (Coleman, 1966; The Consortium for Longitudinal Studies, 1983; Cook-Gurnperz, 1973; Deutsch, 1973; Henderson, 1981; Hess, 1970; Laosa, 1981). Studies in Scandinavia have also demonstrated a significant relationship between social background and school performance (Hernes & Knutsen, 1971; Sandven, 1969; Svendsen, 1978).

Different approaches have been used in order to identify the functional elements in these relationships. One of the main research traditions has been based on the view that, within certain biological limits, and interactive with genetic factors, human interaction and communi- cation influence cognitive development. Within this tradition a main assumption has been that intellectual development advances through dialogues between parent and child (Bee et al., 1969; Bernstein, 1971,1973,1975; Bernstein & Lundgren, 1983; Hartmann & Haavind, 1981; Hess & Shipman, 1965, 1967, 1968; Kamii & Radin, 1967; Laosa, 1981; Luria & Yudovich, 1959; Robinson & Arnold, 1977; Vygotsky, 1962; Wood & Middelton, 1975). Several studies have attempted to identify the characteristics of parent-child communication associated with variations in intellectual functioning and use this to explain the low school performance of many socioculturally disadvantaged children (Bereiter & Englemann, 1966; Deutsch, 1973; Gray et al., 1982; Hess & Shipman, 1965, 1967,1968; Kamii & Radin, 1967; Schweinhart & Weikart, 1983).

Page 2: The effect of social influence on cognitive development and school performance

Scand J Psycho1 30 (1989) Social itifliience, cognitive development, school perfortnunce 53

These studies have generally been based on Bernstein’s ideas about social class differences in language and thought (Bernstein, 1971. 1973, 1975; Bernstein & Lundgren, 1983). Bern- stein’s main concern has been t o show how class relations generate distinctive forms of communication which transmit different underlying linguistic codes from one generation to the next. He distinguished two kinds of linguistic codes, “restricted” and ”elaborated”. It has been difficult to test Bernstein’s sociolinguistic theory of socialization. His theoretical for- mulations are very complicated and have been revised several times. Elaborated and restricted codes, furthermore, seem to be theoretical concepts that cannot be observed. His general assumptions of class-related patterns of language and modes of communication among middle-class and lower-class parents, and how these are mediated to their children can be empirically tested, however.

Empirical studies have generally demonstrated significant differences in speech and teach- ing strategies between middle-class and lower-class mothers. The language and the teaching strategies of middle-class mothers have, furthermore, much more in common with the teach- ing style of the school classroom than the language and the teaching strategies of mothers of low socioeconomic status or minority groups (Laosa, 1981). This discrepancy has been used to explain the inadequate school performance of lower-class and disadvantaged minority group children. However, the relationships between maternal teaching strategies and children’s cognitive performance are rather weak. Hartmann & Haavind (1981) among others discussed possible reasons for this, and concluded that the theory of social mediation had been tested with inadequate methods. For example, studies of cognitive socialization have demonstrated significant social class differences in maternal teaching strategies and language and tried to relate them to significant class differences in child intelligence. Social class, however, seems to be a too global category. The variation within the single social classes for both mothers and children may be great, even if the social class differences are not significant. It is necessary to separate the problem of how social mediation influences cognitive development from the problem of how differences in social class produce different forms of social mediation.

The use of social class or cultural group as variables has also been criticized by social scientists who view this research tradition as being biascd by a white American middle-class point of view, where the white middle-class mother is the standard reference (Baratz, 1969; Laosa, 1981). This may result in the prejudice that there is a kind of deficiency in the family system of all minority and lower-status families.

I n studies of the influence of maternal teaching behaviour, estimations of the child’s intellectual performance are often made from the same situation where the mother teaches the child a certain task. Hence. the measures of the mother’s and the child’s behaviour are interdependent. In other cases, IQ tests are used as independent measures, without any discussion of the relation between the processes studied by the IQ tests and those involved i n the maternal teaching tasks.

Taking this criticism into account, Hartmann & Haavind (1981) devised two methods, one for observing maternal teaching and one for observing child educability and related both to a common theory of cognitive development and social mediation. In accordance with the theory of Piaget (1950) and Piaget & Inhelder (1969), cognitive developmental change was seen as a gradually increasing degree of decentration. A cognitive structure is more decentred when it has changed in ways that make possible simultaneous consideration of relatively more aspects or views of an object or situation. Through the process of assimilation the individual incorpor- ates experiences into already existing structures. Through the process of accommodation, structures change according to the demands of reality. This makes possible assimilation of new impressions. The interactive process of assimilation and accommodation together form an equilibration of inner and outer change.

Page 3: The effect of social influence on cognitive development and school performance

54 E. Hartmann et al. Scand J Psycho1 30 (19h9)

Hunt (1961), Smedslund (1966) and others have pointed out how the environment cons- tantly produces imbalances to which the child tries to assimilate and accommodate. Hunt said that the environment evokes discrepancies between outer impressions and inner structuring, which the child tcnds to reduce. Smedslund asserted that communication conflicts may force the child to change his accustomed ways of thinking. If the discrepancies or conflicts between external impressions and inner structure are very great, the child will not be able to grasp them. No modification of structure will occur. The situation will either be irrelevant or perhaps anxiety provoking. If there is little or no discrepancy, neither will there be any change. The situation will be well known and perhaps boring. This is what Hunt (1971) called “the problem of the match”. Environmental influences on cognitive development can thus be only understood in relation to the cognitive capacities and learning styles which the child has already acquired. Only social interaction which is adapted to the child’s specific cognitive level of functioning and his need for understanding and meaning, can play a role in promoting the child’s cognitive development. It is such “matched” communication which leads to decentra- tion. However, parents vary in their capability to create such “matched” communication with their children, and their effect on the children’s intellectual development will differ accordingly.

In their study, Hartmann & Haavind (1981) found that the main variation in maternal teaching strategy reflected the mothers’ degree of decentred teaching, and the main variation in child educability reflected the children’s degree of decentred educability. The correlation between maternal teaching and child educability was 0.55 @<0.01) in a primary and 0.51 (p<O.Ol ) in a retest study, indicating a strong correspondence in degree of decentred maternal teaching and decentred child educability.

Mothers with a high degree of decentred teaching had an informing and rational approach based on consideration of common rules. Special care was taken to explain possibilities for making choices and plans. However, they also left much initiative to the child, thus both encouraging activity and keeping control of the child’s learning and understanding and eventual need for more instruction. Restraining behaviour was almost lacking. On the whole these mothers invited the child to explore, debate, take the role of another and see his own behaviour from different perspectives. Hartmann & Haavind called this teaching strategy for “informing”. Mothers with low degree of decentred teaching, on the other hand, had a rather imperative approach with a minimum of information and justification with references to reciprocal roles. Opportunities to explain how to choose and make plans were ignored. On the whole they showed little sensitivity to the child’s process of learning and understanding. The child was prevented from reflecting, considering and searching for rational principles. He was trained to be passive and obey authority. This kind of teaching strategy was called “imperative”.

Children with a high degree of decentred educability were active and verbal children with an assertive and concentrated approach to learning and an adequate responsiveness to instruc- tions. They were orientated towards logical principles as a guide for behaviour, and were able to make plans and decisions based on fairly complex and elaborated judgements. When they did not understand or were unsure, it did not result in passivity, but in active claiming for more information. Hartmann & Haavind called this kind of educability “active mastering”. On the other hand, children of a low degree of decentred educability had a fundamental passive approach to learning and instruction. Solutions to problems were reached by superficial imitation or irnpuisive guessing rather than by reflection. Generalization to rational principles was lacking, and decisions and plans were vague and determined by fortuitous factors. This kind of educability was named “passive”.

The maternal strategies described above are in many ways similar to the descriptions given

Page 4: The effect of social influence on cognitive development and school performance

Scand J Psycho1 30 (19x9)

by other researchers (Bee el a l . , 1969: Bernstein, 1975; Brophy. 1970; Cook-Gumperz, 1973; Hess & Shipman, 1965, 1967, 1968; Kamii & Radin, 1967; Laosa, 1981). The strategies are, however, more closely related to observed maternal behaviour and more directly based on relations between different variables. Child educability has not been correspondingly elabor- ated by other authors. The strength of the relations between maternal teaching and child educability exceeds what is ordinarily found, and focuses on variables that are operative as mediating processes between mother and child.

The active mastering child with high degree of decentred educability is supposed to have acquired cognitive structures and learning styles that will match well with the teaching strategies and the demands of the school. In accordance with Hunt’s (1971) concept “the problem of the match”, the school is supposed to have a stimulating and motivating effect on the active mastering child. His adaptation to the educational system will probably be smooth and successful. For the passive child, however, the gap between acquired capacities and the demands of the school is supposed to be very great. This child will probably have great difficulties in grasping and managing the education of the school. Further, the teaching strategies of the informing mother are supposed to resemble the teaching strategies of the school, whereas the strategies of the imperative mother are not. The informing mother probably thus has better prepared her child for the school than the imperative mother. In accordance with Laosa (1981) it is predicted that the children of informing mothers will adapt with greater ease to the demands of the school than the children of imperative mothers. The aim of the present study thus is to analyse the concepts of “decentred child educability“ and ”decentred maternal teaching” as predictors of children’s adjustment and attainment during the first year of elementary school. As the relationship between maternal teaching and child educability is very strong, the predictions from the educability variables and the teaching variables will often coincide. It will, however, be interesting to see which of the two will be the strongest predictor of school adjustment. Lastly. the choice of maternal behaviour does not reflect a belief that this relationship is the only influential one for cognitive development. But in our culture the mother is mostly the main caretaker in the family, and her daily interactions with the child consititute a predominant aspect of his childhood environment (Haavind, 1987).

Generally school adjustment has been predicted either by tests of intelligence or by specially devised tests of school readiness. In Norway the most common method for evaluating school readiness has been Sandven’s test of school readiness (Sandven. 1971). Sandven’s test is supposed to be a good predictor for school adjustment. Further school adjustment during the first year of elementary school in Norway is often tested by Kuhlmann-Anderson’s ability test (Svendsen. 1978). Kuhlmann-Anderson’s ability test is supposed to give a valid measure of school adjustment. Thus positive relations are expected between Sandven’s and Kuhlmann- Anderson’s tests as well as between these tests and the children’s school adjustment and the measures of decentred child educability and decentred maternal teaching.

Social influence, cognitive deveioprnent. sciioai perforrnunce 55

METHOD

Subjects The sample consisted of29children. aged six; six to seven: four, and their mothers, and constituted 65% of the families invited to participate in the study; 20% of the children had attended full-time kindergarten. Most of the children had attended some kind of short-time pre-school lessons the last year before school entrance. The children were enrolled in the same school in Bergen.

Classification of social class was based on income, standard of housing and education of both parents. The families belonged to three different social classes: 24% upper middle class, 34% lower middle class and 42% working class. The percentage of working class people is unusually high, perhaps because many

Page 5: The effect of social influence on cognitive development and school performance

Scand J Psycho1 30 (1Y8Y) 56 E. Hartmann et al.

upper middle class families refused to participate in the study. The mothers were the main caretakers of the children in all the families; 11% of the mothers were working full time, 80% had part-time jobs.

Procedure and instruments Three monrhs before school artendunce the children and their mothers were tested with the methods for studying child educability and maternal teaching (Running Horses) developed by Hartmann & Haavind (1981). The children were also tested with Sandven’s test of school readiness (Sandven, 1971), a common Norwegian test for determining school-readiness.

Four months afrer school arrendance the main teacher of the child evaluated the child along seven variables of intellectual and social school performance.

Afterwards the children were tested with Kuhlmann-Anderson’s ability test (Svendsen, 1978)-an intelligence test often used in Norway for children in the first year of elementary school.

The results from the tests three months before and four months after school attendance were then analysed.

Running Horses In order to evoke social and cognitive processes theoretically related to the present theory of cognitive development and social mediation, Haavind & Hartmann (1977) developed a special game called Running Horses. It is a board game similar to Ludo. There are two players. Each one has two horses. Chance-throws of four ch ipde te rmines how far one can go in a move. Different rules decide where one is allowed to move. Two horses give the players constant possibilities of deciding which horse is most profitable to move in order to gain advantages and avoid obstacles. Winning thus depends both on luck and on skill.

The structure of the game and the social matrix created by playing it is well suited for generalization to social conditions of daily life and for eliciting cognitive processes and learning and teaching strategies relevant for acquisition of concrete operations.

To prevent interdependency between the measure of mother and child, the child‘s behaviour was analysed when learning the game from the investigator, and the mother’s when teaching the game to another child of the same age. In a control study it was demonstrated that the mother‘s way of teaching does not differ when teaching her own child or another child of the same age (Haavind & Hartmann,

The research was carried out in the home of the family where mother and child were supposed to feel safer and act more naturally. The conversation was recorded and relevant non-verbal behaviour was noted.

The instruction of the child was based on what Hunt (1971) defined as “intrinsic motivation”. The quantity of new information was always balanced against the child’s receptivity. The rules were presented during the game but ahead of the child’s need for them. Opportunities to plan and choose were regularly explained though adjusted to the child’s understanding. To obtain smooth communication and coopera- tion with the child, the training was not standardized in the strict sense of the word but continuously adapted to the child’s level of motivation and need for feedback and new information. The essential element was to have the child understand that he should try his best, play as cleverly as possible, win the race, answer questions, listen to explanations and be task-orientated. Children find this situation unthreatening and rather amusing.

The purpose of the analysis of child educability behaviour was both to diagnose cognitive processes and to record the child‘s learning style. Behaviour indicating the emergence of concrete operations was the main basis for diagnosing the child’s cognition. The degree of decentration was most clearly indicated by the child‘s mastery of the rules and by his ability to take several aspects into consideration when making plans and decisions. His learning style was traced from his reaction to cognitive demands and according to how appropriate or inappropriate his behaviour was for the learning process.

The following nine variables for describing child educability were used:

1. Rule masrering is coded every time the child is spontaneously capable of using a rule in a correct as well as efficient way.

2. Decision-making is coded every time the child spontaneously manages to set up a choice where two or more alternatives are discriminated.

3. Planning is coded every time the child spontaneously puts forward adequate plans. 4. Tusk-irrelevant discerning is coded every time the child makes or sticks to plans and decisions that are

1977).

clearly not task-orientated.

Page 6: The effect of social influence on cognitive development and school performance

Scand J Psvchol 3I) (l9SY) Social in fliience, cognitive developrnent, sclzool performance 57

5 . Adequate interest is coded every time the child expresses behaviour which indicates interest or alert

6. Inarfeqrtate reaction is coded when the child expresses negative, irrelevant or over-excited reactions. 7 . Passivity is coded when the child is passive or unresponsive to questions or in other ways expresses

8. Competing is coded every time the child spontaneously focuses on the competitive aspects of the game. 9. Adequate emotion is coded when the child expresses emotional involvement that is relevant.

The mother was taught the game twice, first some days ahead of the observation of the mother, and then just before she was going t o teach the game. Care was taken to explain the purpose of thc research. Training continued until the mother mastered the game. Opportunities to plan and to choose were constantly demonstrated during the training. Questions about the rules were always answered. Questions about how to behave, however, were met by suggestions to act in the way she found most natural. Three mothers refused to teach the game. The dropouts were the most insecure of the mothers. Their children, participating in the study, all mastered the game in a very inferior way.

The purpose of the analysis of maternal teaching was to classify those aspects of her behaviour that from a theoretical point of view was related to how cognition is stimulated or inhibited, The general idea was that pointing to relevant aspects and taking them into account simultaneously as well as stimulating the child to act independently could promote decentration. On the other hand cognitive development might be delayed when the mother restricts the child's possibilities to reflect or act on account of his own understanding.

1. Informing is coded for statements where the mother gives information about the rules of the game. Informing is coded once for each item of information. A certain criterion of efficiency has to be passed in order to score for this category.

2 . Anficipution is coded every time the mother talks about what may or might happcn in the game according to the possibilities the rules crexc: .

3 Demonstration of alternatives is coded every time the mother talks about the playcr's possibilitics of choosing which horse to move. Explicit focusing of at least two alternatives is required.

4. Seeking feedback is coded every time the mother overtly tries to check how the child undcrstands the game.

5 . Imperurive feedback is coded every time the child expresses uncertainty or acts against the rules. and thc mother responds to this with an order without any explanation. The scores. however. are calculated as a percentage of all situations requiring feedback.

6. Ordering is coded every time the mother spontaneously directs the child or acts on behalf of the child without explaining why.

7 . Restriction is coded for different types of maternal behaviour that are supposed to confuse the child: very vague and unintelligible information. instructions discordant to previous given information. ignoring situations where the child obviously is in need of feedback. etc.

understanding.

uncertainty by just doing nothing.

The following nine variables for describing maternal teaching were used:

8. Competition is coded whenever the mother speaks about competitive elements of the game. 9. Emofionalsupport is coded every time the mother gives overt support and encouragement to the child,

Judgement of school performance The purpose of the judgement of school performance was to evaluate the child against a set of variables that included cognitive, emotional and social processes which are important in relation to the demands of elementary school. Further, the variables had to include psychological processes which the teacher may be able to form an opinion about after four months of schooling. The programme for the Nonvegian elementary school and information from experienced teachers were the main basis for composing the following seven variables for judgement of school performance:

1. lnifiotive was defined a s unsolicited activity which is appropriate in a training situation. Initiative expresses itself through curiosity and an active attitude towards learning, alert interest and motivation towards the teacher's instructions, etc.

The teacher judged initiative on the basis of the above given definition and according to whether they experienced the child's participation in the classroom as joyful or reluctant.

2 . Concentration was defined as the ability over time to keep one's attention on certain tasks. Per- severance and attention are important aspects of good concentration.

The teacher judged the child's concentration on the basis of his perseverance, attention and ability to

Page 7: The effect of social influence on cognitive development and school performance

58 E. Hartmann et al. S c a d J Psycho1 30 (1989)

3.

4.

5 .

6 .

7.

A

work thoroughly in the class. Inferior concentration expresses itself when the child is acting out, daydreaming or in other ways is absent-minded. Apprehension and utilization oftraining was defined ils the ability to understand information and later use it in an adequate way. Memory is an important element of this variable.

The teacher judged the children on this variable according to how they experienced the child‘s ability to apprehend information and instructions and his ability to work independently. Generalization of learned material was defined as the ability to remember essential elements of a problem and transfer them to the solution of other problems. Synthetical ability and incipient ability to abstract cognition are important aspects of this variable.

The teacher judged the children on this variable according to the above given definition. The instruction to the teacher also included examples indicating superior and inferior ability to make gcnemlizations. School achievement was defined as the results the child had accomplished in verbal and written school subjects.

The teacher judged the children on this variable on their general impression of the child’s perfor- mances. The teacher was further asked to compare each child’s skill according to what one would ordinarily expect that chiidren manage after three to four months of school attendance. Social contact was defined as contact between teacher and pupil that might promote learning.

The teacher judged the children on this variable on the basis of his impression of the child’s contact with him. The instruction to the teacher also included examples indicating good and inferior teacher- pupil contact. Atfainnlent in school was defined as the child’s general ability to adapt to the school system.

The teacher judged the child’s attainment according to how confident he seemed to be at school, whether he did his homework and how he generally related to the routines and rules of the school.

five-point scale was used for each variable instead of a seven-point scale, because the teachers were inexperienced judges (Anastasi, 1969). The teachers were informed about the study some weeks before the evaluation took place. Definitions of the variables and the scales were given both written and verbally during a structured interview where the teachers’ judgements were noted. In order to reduce the possibility that the teacher would judge each child equally on all variables, every child was judged on each variable separately. The teachers were invited to use all the points of the scales. The scores on the seven variables of school performance were summed up in a sum score which givesa general estimation of school performance.

Sandven’s test of school readiness and Kuhlmann-Anderson’s ability test Sandven’s test of school readiness is a group test, given to children about half a year before school entrance. It is composed in order to measure intellectual abilities on the following functions: memory, verbal apprehension, apprehension of proportion and form, reasoning and apprehension of numbers and quantity. The test is standardized for Norwegian conditions and has demonstrated high reliability and ialidity.. The correlation with other tests of intelligence is high (Gahnstrom & Skerfing, 1973; Sandven, 1971).

The school selected for this research had a standard offer to all pre-school children of being tested with Sandven’s test of school readiness half a year before entering school. The families in this research were required to say yes to this offer. Everybody said they would do so, but only 70% of the children in the study were tested. The others did not show up on the day of testing and thus fell out on this test.

In addition 93% of the children were tested with Kuhlmann-Anderson’s ability test. It is a group test that measures intellectual abilities mainly on the following functions: memory, reasoning, apprehension of concept, apprehension of similarities and dissimilarities. Strong correlations have been demonstrated between Sandven’s test and Kuhlmann-Anderson’s ability test (Sandven, 1971). The test was presented by the teacher a short time after the teacher had made his judgements of the children’s school performance.

RESULTS

Maternal teaching and child educability Principal factor analysis without rotation of the child variables and of the maternal variables gave the same results as found by Hartmann & Haavind (1981). In both samples most of the variables are related to the first factor which accounts for 60% of the variance among the

Page 8: The effect of social influence on cognitive development and school performance

Scand J Psycho1 30 (1989)

children and 61% among the mothers. In both samples the first factor seems to be a bipolar dimension reflecting respectively the children's degree of decentred educability and the mothers' degree of decentred teaching. The first principal factor score without rotation of each child and each mother is thus used as an individual measure of degree of decentred child educability and decentred maternal teaching in the further analysis. High positive scores for the children reflect a decentred, active mastering style of educability and high negative scores a centred, passive style. High positive scores for the mothers reflect a decentred informing teaching strategy and high negative scores a centred, imperative strategy. The correlation between the mothers' and the children's first principal factor scores demonstrated the same close relationship ( r=0 .52 , pCO.01) between degree of decentred maternal teaching and decentred child educability as found by Hartmann & Haavind (1981).

Sociril influence, cognitive tleveloptnent, scliool performance 5 Y

School performunce

The mutual correlations between the seven variables of the teacher's judgement of school performance are summarized in Table 1.

All the correlations, except for some involving Social contact, are significant on the 0.01 level. Social contact shows no significant relation to Initiative and Concentration. but the other correlations are significant on the 0.05 level.

Predictors of school performance

Multiple regression analyses were used in order to discriminate which of the independent variables were the best predictor of the teacher's judgement of the children's school perfor- mance. The following predictors were used: the child's measure of decentred educability, the mother's measure of decentred teaching, the child's score on Sandven's test of school readi- ness and the child's score on Kuhlmann-Anderson's ability test. The teacher's sum score of school performance was used as a dependent variable.

Table 1. Spearman rank correlation between the seven variables of teacher jitdgemenr of school perfor- mance (N=?Y)

Apprehension Generali- and utili- zation of zation of learned School- Social Attninmant Sum

Variable Initiative Concentration training material achievement contact to school score

Initiative 1.000

Concentration 0.7547** l.OW

Apprehension and utilization of training 0.8391 * * 0.7-85* * 1 .ooo Generalization of learned material 0.8785'* 0.7485** 0.8983" 1.000

School achievement 0.8634** 0.8096** 0.9391'* 0.9368'* 1.000

Social contact 0.2947 0.2689 0.3539' 0.4?06* 0.3506' 1.ooO

Attainment in school 0.6215" 0.6305*' 0.6728'' 0.7504" 0.7030** 0.5514** 1.000

Sum score 0.8966** 0.859" 0.94 15' * 0.9570* * 0.9396' * 0.5056'* 0.8054*' 1.000

'p<O.05; * *p<o.o1.

Page 9: The effect of social influence on cognitive development and school performance

60 E. Hartriinnn et al. Scand J Psycho1 30 (1989)

Table 2 . hfirtripfe regression anulysis of school performance

Teacher judgement of school performance

Predictors r ~

'7r of variance

Degree of decentred child educability

Degree of decentred maternal teaching (N=29) 0.57" * 33* *

(N-36) 0.67'' 45*= Sandven's test of school readiness (,V=20) 0.048 .. 1 Kuhimnnn- Anderson's ability test (N=27) 0.56' a 31"

**p<0.01.

The results from the multiple regression analysis are presented in Table 2 . The analysis shows that the best single predictor of school performance is degree of

decentred maternal teaching, accounting for 45% of the total variance in the teacher's evaluation of school performance. Degree of decentred child educability is also a good predictor of school performance, accounting for 33% of the variance. Kuhlmann-Anderson's ability test, presented some weeks after the teacher's evaluation took place, accounts for 31 % of the variance. In contrast Sandven's test of school readiness only accounts for 2 7 ~ of the teacher's judgement.

There is also no relation between Sandven's test and the measure of decentred child educability (r=0.092). The correlation between the test of Sandven and Kuhlmann-Ander- son's ability test is only 0.24 and not significant.

As there were 30% dropouts on the Sandven test, the tested sample and the dropouts were compared on the following measures: mean score of decentred child educability, mean score on Kuhlmann-Anderson's ability test and mean sum score of the teacher's judgement of school performance. The comparison is presented in Table 3, and shows that the two groups are highly similar on every measure.

DISCUSSION

The strong positive relationships between maternal teaching as well as child educability on Running Horses, and teacher judgement of school performance are congruent with the hypotheses.

The results suggest that degree of decentred maternal teaching is a good predictor of the

Table 3. Comparison bemeen the tested sample and the dropouts on Sandven's test of school readiness

Tested sample Dropouts (N=20) (N=9) Mean scores Mean scores

Degree of decentred child educability 0.04 0.01

Teacher's judgement of school performance 21 -_ Kuhlmann-Anderson's ability test 58 56

17

Page 10: The effect of social influence on cognitive development and school performance

Scand J Psycho1 30 ( I W Y ) Social in fhence, cognitive developnient, school performance 61

child’s adaption to the education system of the school, better than any of the child variables. Combination of maternal teaching and child educability only slightly raises the predictability. The fact that the mother seems to be a better predictor of her child’s school performance than the child himself gives further support to the assumption that parents, particularly the main caretaker, which in this study is the mother, are important mediators between the child and the outer world.

The positive significant relationship between Kuhlmann-Anderson’s test and the teacher judgement of school performance is congruent with the hypothesis. I t is, however, important to notice that this test of cognitive ability, administrated a few weeks after the teacher judgement had taken place, shows lower predictability than the variables of maternal teaching and child educability calculated seven months earlier.

It is amazing and incongruent with the hypothesis, that Sandven’s test of school readiness shows no relationship with the teacher judgement and very low insignificant relation with Kuhlmann-Anderson’s test. Sandven (1971) found significant relations between the test of school readiness presented half a year before school attendance and teacher judgement of school performance (r=0.58) and Kuhlmann-Anderson’s ability test (r=0.78) done half a year after school attendance. The lack of relation between Sandven’s test and the other measures in this study is difficult to understand, and the reasons are probably manifold. The sample is small, and as many as 30% of the children were not tested with Sandven’s test. However, the tested sample and the dropouts d o not differ according to mean scores on decentred child educability, Kuhlmann-Anderson or the teacher’s judgement of school performance. Thus, there is little reason to think that the dropouts have changed the representativeness of the sample in ways that can explain the lack of relation between Sandven’s test and the other measures. As the correlation with Kuhlmann-Anderson’s test also is very low, one reason for the insignificant results for Sandven’s test may be that there were some unknown irregularity in the school’s administration of the test. The results, however, also give support to the view that this test have lost some of its predictive value as the elementary school has changed.

It is difficult t o judge whether the strong relation between maternal teaching and child educability, and between maternal teaching and the school performance of the child, shows that maternal communication influences the child’s cognitive development, or “only” his competence in adaptation to elementary school. It seems reasonable, however, to argue that decentred. informing teaching creates social interaction that is adapted to the child’s level of understanding and in addition stimulates him to be active and use a variety of cognitive processes. Frequent encounter with such influence is likely both to have promotive effect on the child’s cognitive functioning and to make him better prepared for a smooth and successful transition to the school.

On the other hand centred, imperative teaching will very often not match the child’s level of functioning. The child is left to passively follow instructions, often without understanding why. Exposure to such adult-child interaction has probably both a hampering effect on cognitive development and gives little preparation for taking advantage of the teaching of the school.

Hunt (1979) has pointed out that the environment that surrounds the child early in life is likely, without major change of milieu, to be the environment that in many ways will continue to surround him in the future. This is what he has called the “continuing contex of develop- ment”. Thus the aspects of maternal teaching that exercise an influence on the child’s educability at ages six to seven will probably continue to operate in years to come.

Further longitudinal studies are needed for illumination of the possible long-term effects of maternal teaching and child educability for children’s cognitive development and competence in school and in society.

Page 11: The effect of social influence on cognitive development and school performance

62 E. Hartmann et al. Scand J Psycho1 30 (1989)

This study was partly financed by the Norwegian Research Counsil for Science and the Humanities.

cand . psychol. degree.

educability.

The present article is based on data from the thesis of T. J . Eri & A. H. Skinstad (1977) for their

H. Haavind and E . Hartmann developed the methods for testing maternal teaching and child

REFERENCES

Anastasi, A. (1969). Psychological testing, 3rd edn. London: Macmillan. Baratz, J. C. (1969). A bi-dialected task for determining language proficiency in economically disadvan-

taged Negro children. Child Development, 40, 889-901. Bee, H. L., van Egeren, L. F., Streissguth, A. P., Nyman, B . A. & Leckie, M. S . (1969). Social class

differences in maternal teaching strategies and speech patterns. Developmenial Psychology, 1 , 727- 734.

Bereiter, C. & Englemann, S. (1966). Teaching dbadvanraged children in the preschool. Englewood ClifIs, NJ: Prentice Hall.

Bernstein. B. (1971). Class, codes and control. Vol. 1. Theorerical siudies towards a sociology of language,. London: Routledge & Kegan Paul.

Bernstein, B. (1973). Class, codes and control, Vol. 2 . Applied siudies towards a sociology of language. London: Routledge & Kegan Paul.

Bernstcin, B. (1975). Class, codes and conirol, Vol. 3. Towards a theory of edi~caiional rransmission. London: Routledge & Kegan Paul.

Bernstein, B . & Lundgren, U. P. (1983). Maki, koniroll och pedagogik. Lund: Liber Forlag. Brophy, J . E. (1970). Mothers as teachers of their own pre-school children. Child Developtnent, 41 ,

Coleman, J . S. (1966). Equality of educaiional opporfunify. Washington, DC: US Office of Education. The Consortium for Longitudinal Studies (1983). As fheiwig isbent. tnstingeffecisofpreschoolprogra~ns.

Cook-Gumperz, J . (1973). Social control and sociafizalion. London: Routledge & Kegan Paul. Dcutsch, C. P. (1973). Social class and child development. In B. M. Caldwell & H. N. Ricciuti (Eds),

Review of child development research, Vol. 3 . Chicago: University of Chicago Press. Eri. T. & Skinstad, A. H. (1977). Barns fdelsesmessige og infellektuelle tilpasning iil skolen self i lys av

barns innlreringsstil og mpdres pedagogiskestil. Hovedoppgave, Institute of Psychology, University of Bergen.

Gahnstrom, E. & Skerfing, M. (1973). N y Crrkolingsmodell for nybcjrjare. Stockholm: Pedagogisk sentrum.

Gray, S. W . , Ramsey, N. K. & Klaus, R. A. (1982). From3 fo20: Theearly trainingproject. Baltimore: University Park Press.

Haavind, H. & Hartmann, E. (1977). Mothers as teachers and their children as learners. Reports from the Instituie of Psychology, University of Bergen, No. 2.

Haavind, H. (1987). Liten og stor. Oslo: Universitetsforlaget. Hartmann, E. & Haavind, H. (1981). Mothers as teachers and their children as learners. In W. P.

Robinson (Ed.), Communication in development. European Monographs in Social Psychology. 21. London: Academic Press.

Henderson, R. W. (1981). Home environment and intellectual performance. In R. W. Henderson (Ed.), Parent-chifd interaction: Theory, research and prospecfs. New York: Academic Press.

Hernes, G. & Knutsen, K. (1971). Levekdrsunders0kelsen. Utdanning og ulikhet. NOU 45. Oslo: Universi tetsforlaget .

Hess, R. D. (1970). Social class and ethnic influences on socialization. In P. H. Mussen (Ed.), Car- michael's manual of child psychology. New York: Wiley.

Hess, R. D. & Shipman, V. C. (1965). Early experience and the socialization of cognitive modes in children. Child Development, 36,869-886.

Hess, R. D . & Shipman, V. C. (1967). Cognitive elements in maternal behavior. In J. P. Hill (Ed.), Minnesota symposium on child development 1 . Minnesota: University of Minnesota Press.

Hess, R. D. & Shipman, V. C. (1968). Maternal influence upon early learning. In R. D. Hess & R. M. Bear (Eds), Early education. Chicago: Aldine Press.

Hunt, J. McV. (1961). Intelligence and experience, New York: Ronald Press.

79-94,

New Jersey: Lawrence Erlbaum Associates.

Page 12: The effect of social influence on cognitive development and school performance

Scand J Psycho1 30 (19x9)

Hunt. J. McV. (1971). Using intrinsic motivation to teach young children. Educarionul Technology, 2, 78-80.

Hunt. J . McV. (1976). Environmental programming to foster competence and prevent mental retardation in infancy. In R. N. Walsh & W. T . Greenough (Eds), Environinents us fherapyfor brain dysfunctions. London: Plenum Press.

Social influence, cognitive development, rchool performance 63

Hunt, J MrV. (1979). Psychological development: Early experience. Annual Review of Psychology. Kamii, C. K. & Radin, N. L. (1967). Class differences in socialization practices of Negro mothers. Journal

of Marriage and rhe Family, 29, 302-310. Laosa, L. M. (1981). blaternal behavior: Sociocultural diversity in modes of family interaction. In R. W .

Henderson (Ed.), Parenr-child inreracrion: Theory, research, and prospect. New York: Academic Press.

Luria, A. R. & Yudovich, F. (1959). Speech and the developmenr of menralprocesses in the child. London: Staples Press.

Piaget, J . (1950). Thepsychology of intelligence. London: Routledge & Kegan Paul (originally published. 1947).

Piaget, J . & Inhelder, B. (1969). The psychology of rhe child. London: Routledge B Kegan Paul (originally published, 1966).

Robinson, W. P. & Arnold, J . (1977). The question-answer exchange between mothers and their chil- dren. European Journal of Social Psychology, 7, 151-163.

Sandven, J . (1969). Recruitment to the gymnasium. Scundinavian Journal of Educurional Rereurch. 13, 29-68.

Sandven, J . (1971). Der reoreriske og merodiskt grurinlug for modefiherspr@ving. Oslo: Universitets- forlaget.

Schweinhart. L. J . & Weikart. D. D. (1983). The effect of the Perry preschool program on youth through age IS . A summary. In the Consortium of Longitudinal Studies. As rhe fwig is bcnr. Lusting effects of prrschool prograrm. New Jersey: Lawrence Erlbaum Associates.

Srncdslund, J . (1966). Les origins socialea de l a dtkcntration. In F. Brcsson & G. dc h.lontmollin (Eds). Pxychologie el epislenrologie generiqite. Paris: Dunod.

Svendscn. D. (1978). Testrcsultatcr og sosial bahgrunn. Unpubliahcd manuxript. Institute of Psychol- ogy, University of Bergen.

Vygotsky, L. (1962). Though1 and language. Ncw York: Wilry. Wood. D. & Middelton. W. (1975). A study of assisted prohlcm-solving. Britisfz Journal of Psychology,

66. 181-191.

Received 3 Octobcr 1988