Child Social Competence – Attentional Control and Temperament

In the area of child development, cognitive and social processes are important topics of interest. There have been numerous research studies that have examined the relationships between the cognitive functioning of children as it relates to their social development and behavior. As the subject is rather complex, further investigation is needed to provide more in depth understanding of child developmental processes.

Temperament is an important variable in any discussion of child behavior. In his theory, Arnold Buss (1991) defines temperament as a subclass of personality traits that have a hereditary component, appear during the first year of life, and persist into later life. Emotionality, activity, sociability are the main components of his EAS theory, and show less variability than most assume. Buss contends that these traits are relatively stable, and are evident even before socialization typically occurs. Individual differences in temperament undoubtedly have bearing on a child’s behavior, social competence, and possibly on his or her cognitive abilities.

In a study that examined the relationships among emotionality, regulation, social skills and popularity in preschoolers, it was found that children who are less able to control their attention while engaged in social interaction have a harder time being socially successful (Eisenberg, Fabes, Bernzweig, Karbon, Poulin & Hanish, 1994). A number of interesting sex differences were also discovered. For both boys and girls, acting out and emotional intensity were negatively related to social skills. For boys, though, attentional control and constructive coping were positively related to both social skills and popularity, whereas for girls, avoidant coping was positively correlated with social skills. The effects of early socialization are mentioned as a contributing factor to the sex differences in the findings.

Posner and Rothbart (2000) explored the developmental mechanisms of self-regulation from a neurological standpoint. They determined that the two brain areas responsible for the regulation of attention and voluntary control are the prefrontal cortex and the anterior cingulate. A central concept to Posner & Rothbart’s work is plasticity, which is the neurological restructuring that takes place when new skills are learned. Within the first few years of life, neurological systems change and develop, allowing children increasing attentional control over their thoughts and actions. Attention and voluntary control are also essential to the regulation of emotion, temperament, conflict resolution, empathy and the development of the conscience. These findings are important because they allow for tracking of the development of the neurological systems of attention and personality.

Diamond (1988, 1990, 1991a), as cited in Gerstadt et al (1994), also believes the frontal cortex is necessary for inhibition, and for memory tasks as well. To test this idea, Gerstadt and colleagues (1994) gave a nonverbal Stroop-like task to children ages 4-7. The task required inhibition of natural associations and memory for the instructions that were given. Participants were asked to say “day” to a card depicting the moon, and “night” to a card with a picture of the sun. This task required a lot of voluntary control and attentional effort to remember the rules and respond correctly. As was expected, performance improved and response latency decreased with age (Gerstadt et al., 1994). In other words, children’s ability to control their attention showed development and maturity with age.

The purpose of the present study is to examine the relationships between cognitive, social and behavioural aspects of child development, including an investigation into the effects of temperament on those processes. Whereas Posner and Rothbart (2000) did not focus much on the social implications of attentional control, and Buss (1991) did not discuss the implications of temperamental differences on cognition, the present study attempts to explore all three variables and their relationships to one another. It was hypothesized that attentional control would be positively correlated with positive interactions, meaning that those who were best able to control their attention as measured by the Stroop task would be expected to show more instances of positive play behaviours. It was generally expected that the children who exhibit more inhibitory control would be the most socially successful. Similarly, those with low attentional abilities would be expected to have more negative social interactions or less interactions overall. Also, it was hypothesized that the more negative aspects of temperament, such as emotionality and shyness, will be negatively correlated with both the type of interaction (i.e. positive or negative) and the number of interactions overall.

Methods

Participants

One hundred and sixteen children in elementary schools in Canterbury, England and Dundee, Scotland served as participants. There were equal numbers of males and females, and participant age ranged from 4 years to 11 years (mean age= 8.20 years, SD= 1.67).

Our sample was chosen randomly after permission was given by the head teachers and classroom teachers at each school. Seven out of 29 classrooms did not participate in random sampling, due to specific requests from the teacher. Consent was obtained in the form of permission slips that were sent home with each child, and only those who returned the forms with a parent’s signatures were able to participate. Children with known developmental disorders, such as autism, Asperger’s syndrome, or ADHD, were excluded from the study.

Measures

Temperament. Classroom teachers completed an EAS questionnaire for each child. The 20-item questionnaire, rated on a 5-point Likert scale, is designed to assess each child on four aspects of temperament. Each item was rated as to how well it described the child, with 0 being “not at all” descriptive and 4 describing the child “very well”. There were five items for each of the subscales of emotionality (e.g. “Child often fusses and cries”), activity (e.g. “When child moves about, he/she usually moves slowly”), sociability (e.g. “Child finds people more stimulating than anything else”), and shyness (e.g. “Child takes a long time to warm up to strangers”). Frequency counts were calculated for each subscale.

Cognitive functioning. There were two Stroop tasks used during this experiment, one a verbal and one a non verbal task. The non-verbal task was used to control for the younger children who may have been unable to read. Their performance on the verbal task would be much better, since there would be no natural tendency to read the word as opposed to naming the ink color, which was the task at hand. The Stroop task requires both inhibitory control of attention and memory for task instructions.

Observational play. Observers recorded both the type of interaction (i.e. positive or negative) and whether the child was playing alone, with one other, or with a group. Positive interactions included behaviors such as smiling, laughing, holding hands, sharing, and taking turns. Negative interactions included behaviors such as crying, teasing, physical fights, rejection and isolation. Situations that were neutral were left up to the judgment of the observer. Experimenters made note and counted each time the child’s play situation changed. For example, if a child was sitting alone crying, and was then approached by another child who shared his or her toys, the interaction was noted as changing from “alone/negative” to “one other/positive”.

Materials

For the non-verbal Stroop task, two sets of cards were used, one each for the experimental and control conditions; each set was made up of 20 cards. Half of the cards for the experimental condition had a white background with a picture of a yellow sun in the center, and the other half had a black background with a picture of a white crescent moon and stars. For the control condition, half of the cards had a red and blue checkerboard pattern, and the other half had a blue background with two red squiggly lines in the center. For the verbal Stroop task, one card was used for each of the experimental and control conditions. The control card was comprised of two columns of nouns (e.g. “shoe”, “dog”, etc.), each word printed in red, blue, black, green, or yellow ink. There were a total of 20 words on each of the cards. The experimental card was comprised of color words (e.g. “black”, “yellow”, etc.), each of which was printed in a color of ink that did not match the word itself. For example, the word “black” was printed in red ink, “green” in blue ink, and so on. A stopwatch was used to record time.

Procedure

The study took place over a period of two weeks, during which the experimenters sat in the classrooms to observe. Teachers were given an EAS questionnaire for each child as soon as participants were chosen, to allow enough time for their completion. Observations were done on three separate days, for 10-minute periods each, during either free play in the classroom, outdoors on the playground, or at lunch time. Experimenters were careful in following the children as they moved about while remaining distant enough so as not to influence the child’s behavior by making them aware of the fact that they were being watched.

Each child was brought to a quiet area to be given the Stroop task, to minimize distraction. A demographic information sheet was filled out with the child before testing began. The demographic information that was collected included information about age, gender, siblings, pets, hobbies, and favorite subjects in school.

For the non-verbal Stroop test, there was a brief training period before the experimental trials began. Participants were instructed to say “day” to the black card with the moon and stars in the experimental condition (squiggle card in the control condition), and to say “night” to the sun card (checkerboard pattern in the control condition). After instructions were given, the experimenter then did a practice trial with four of the cards to ensure that the directions were fully understood. If the child made any errors during the practice trial, both the training and practice trials were repeated until the child reached 100% accuracy. After the practice trial, the four cards used were set aside, and the experimental trial was carried out with the remaining 16 cards. Participants were shown the cards one at a time, in random order, as the experimenter recorded their responses.

For the verbal task, participants were asked to name the color of the ink in which each word was printed, as the experimenter went down each column, pointing to the words one at a time. Tests conditions were alternated so that half the subjects were first given the control and then experimental condition and half were given the experimental and then control condition, on both the verbal and nonverbal tasks. Time recording began once participants gave their first response, and ended as soon as they gave their last. On average, the procedure took approximately fifteen minutes to complete.

Results

An alpha value of .05 was chosen and used for all analyses unless otherwise specified. A series of one-sample t-tests revealed that the Stroop test was effective. Verbal error rate t(115)= 5.44, p<.001, verbal reaction time t(115)= 6.20, p<.001, and nonverbal reaction time t(115)= 2.84, p<.01 were all found to be significant. Nonverbal error rate, however, was not significant, t(115)= 1.23, p>.05. The effectiveness of the Stroop test was further confirmed by our finding that the verbal error rate was significantly higher than the nonverbal error rate (t (115) = 4.24, p<.001).

It was hypothesized that attentional control, operationally defined by low error rates, would be positively correlated with positive behavioral interactions. Analyses revealed that a higher verbal error rate was positively correlated with instances of positive alone play, r= .19, p<.05, and positive play with one other, r = .18, p <.05. Children who made more errors, indicative of less attentional control, had more instances of positive play interactions. Verbal error rates also correlated positively with negative group play, r= .20, p<.05, and total play with one other, r= .21, p<.05.

It was also predicted that the more negative aspects of temperament, such as emotionality and shyness, would be negatively correlated with both the type of interaction (i.e. positive or negative) and the number of interactions overall. The EAS subscale of emotionality was negatively correlated with positive group play, r= -.21, p< .05, and total group play, r= -.18, p< .05. In other words, more emotionally reactive children tended to have less positive group play and less group play overall. The activity subscale was positively correlated with positive play with one other, r= .20, p< .05, positive group play, r= .21, p< .05, and total play with one other, r= .20, p< .05. Active children tended to engage in more instances of positive play with one other and in groups, and more instances of total play with one other overall.

The association between the Stroop measure and the EAS subscales were also examined. Verbal error rate was positively correlated with scores on the emotionality subscale, r= .20, p< .05, so that children with greater negative affect tended to make more mistakes. Reaction time on the verbal Stroop task, however, was negatively associated with emotionality, r= -.18, p< .05. Children with less negative affect took longer to complete the verbal task. Verbal reaction time was also found to correlate positively with both positive play alone, r= .23, p< .05, and total alone play, r= .22, p< .05. Those who took longer to complete the task tended to have more instances of positive alone play and alone play overall.

Age was found to be related to the Stroop and for the behavioral observational measure. Verbal error rate was found to correlate negatively with age, r= -.19, p< .05, so that older children tended to make fewer errors on the verbal Stroop task. Also, age was found to correlate negatively with a number of play situations, including positive play with one other, r= -.29, p< .01, total positive play, r= -.19, p< .05, negative alone play, r= -.22, p< .05, negative play with one other, r= -.31, p< .01, negative group play, r=-.29, p< .01, total negative play, r= -.37, p< .001, total alone play, r= -.22, p< .05, total play with one other, r= -.34, p< .001, and total play overall, r= -.24, p< .01. As age increased, the types of play interactions changed. No significant gender differences were found on any of the measures.

Discussion

Interpretation of Findings

This study was done in order to find relationships among cognitive, social and behavioural aspects of child development. Temperament was also explored as a variable having influence on of those processes. The primary hypothesis, that high attentional control would be positively correlated with positive behavioral interactions, was disconfirmed by data analysis. Children who made more errors on the verbal Stroop task tended to engage in more instances of positive play alone or with one other, more instances of overall play with one other and more instances of negative group play. These results are not what was expected, and seem to reflect rather complex socialization processes. A child who is less able to control his or her attention may be more likely to fail at group interaction, making him or her more likely to play either alone or on a one-on-one basis. That child would probably be more happy in those play situations then, as he or she would avoid further negativity and rejection from a larger group. Previous literature has suggested that children with less attentional control would be less socially successful (Eisenberg et al., 1993), which conflicts with the findings of this study.

The secondary hypothesis, that temperamental traits such as emotionality and shyness would be negatively associated with positive interactions and total play interactions overall was only partially confirmed by data analysis. Children high on the trait of emotionality were found to be less likely to engage in positive group play and total group play overall. However, there were no significant relationships found between shyness and behavioural interaction. One would expect the shy child to be less socially successful as well. Activity was positively related to more instances of positive play with one other, positive group play and total play with one other. It is logical finding that active children tend to engage in more instances of play, especially group play. Similarly, one would expect that the highly social child would be very socially successful and engage in many instances of positive play. It was interesting that no other significant relationships were between the other EAS subscales (i.e. sociability and shyness) and the behavioral measure. These findings agree somewhat with previous literature that has found emotional intensity to be negatively correlated with social skills (Eisenberg et al., 1993), although no significant sex differences were discovered.

The finding that verbal error rate was negatively correlated with age is also in line with what one would expect. Older children make fewer mistakes than do younger children on the verbal Stroop task because as age increases, performance on the task improves as inhibitory mechanisms mature. Gerstadt et al. (1994) found that performance on the Stroop task improved and that response latency decreased with age. There was no found association between any of the EAS subscales and age, however, which attests to the theory of temperament proposed by Buss (1991) and the idea that temperamental traits are stable. Both of these findings are in line with the previous literature.

Another interesting finding is that verbal error rate was positive correlated with emotionality scores, such that those children who made more errors tended to have higher emotionality ratings. These children, given their temperamental disposition to be more negative, could have been distracted by their mistakes which caused them to err in subsequent trials. They may have also been more fearful or apprehensive about making mistakes, or have had an expectancy that they would not do well. Contrastingly, a higher verbal reaction time was associated with less emotionality. Children who had less negative affect took longer to complete the verbal Stroop task. A possible explanation is that those children who took longer are more invested in self-regulation and self-monitoring, and were thus more careful to not make mistakes. The finding that age was related to lower error rates further attests to the idea that self-regulatory mechanisms and attentional control mature with as one gets older.

There were a number of negative correlations found between age and play interactions, that again provide grounds for exploration into childhood social behaviors. Play and the dynamics of social interaction undoubtedly change with age. Older children were found to have less negative play interactions overall, which is understandable as their reasoning and conflict resolution skills improve over the years. It is also important to keep in mind when interpreting these results that data analyses were based on frequency counts. So whereas young children, having shorter attention spans, changed activities and playmates many times within one observation period, older children tended to remain in one play situation longer. Older children also tend to be more involved in organized group games (e.g. football, rounders), which makes them less likely to play alone or with just one other child. These and similar situations caused older children to have lower frequency counts overall. The results seem to be very much in line with childhood socialization processes.

Limitations

The primary limitation to this study is the large number of experimenters (n=29) that were involved in data collection. Inter-rater reliability was unable to be established because of the individual assignments at each school (i.e. no two observers were ever in the same classroom).This may have been most problematic concerning the behavioural observation measure, since it left judgments about a child’s interactions solely up to the individual experimenter. Also, there was no coding system in place for specific types of behaviours that defined either positive or negative play.

Another point of interest concerns the verbal Stroop task, which lacked the training component of the nonverbal task. The instructions that were given only asked the child to name the color of ink that the words were written in, and then the experimental trials began. Experimenters were unable to warn the children against reading the word, as that may have affected the results. There was therefore no true way of knowing whether or not the child completely understood the instructions. This was especially true for those participants who were given the experimental condition first. Those who were given the control condition first were at least able to have somewhat of a practice run, during which they had no need to inhibit their natural tendency to read, and could name the ink colors easily for the non-color words. For the others, though, there was no way to determine if they were making errors due to lack of inhibition or a lack of understanding of the task.

Response time was not recorded for each individual trial as was the case in previous studies (Gerstadt et al., 1994), which produced much more variability in reaction times. Random error, such as a participant stopping to cough or distractedly looking around before responding, could have affected results and prevented significant relationships from being revealed. More careful time recording procedures, in which an average response time is used rather than an overall response time, should be considered in future research.

Future Directions

Despite some limitations of research design, this study has provided an interesting look at the relationships between cognition, socialization and behavior processes in childhood. A number of insightful relationships have been revealed and this study provides a good basis for future research in the area of child development, specifically research into the relationships between attention and social success.