Journal of Pediatric Psychology, Vol. 28, No. 7, 2003, pp. 505-516
© 2003 Society of Pediatric Psychology
The Role of Parents and Temperament on Children's Estimation of Physical Ability: Links to Unintentional Injury Prevention
University of Alabama at Birmingham
All correspondence should be addressed to David C. Schwebel, Department of Psychology, University of Alabama at Birmingham, 1300 University Blvd, CH 415, Birmingham, Alabama 35294. E-mail: schwebel{at}uab.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionThe Role of Parents...Objectives of the StudyMethodResultsDiscussionReferences |
|---|
Objective Unintentional injuries, the leading cause of pediatric mortality, are caused by a complex set of intrapersonal and environmental factors. The role of three critical variables— parental supervision, children's temperament, and estimation of children's physical abilities— was examined. Methods Sixty-four 6- and 8-year-old children completed a laboratory experiment with a parent. Both children and parents judged the child's ability to complete reaching, stepping, and crouching tasks. Parents also completed a parent-report measure of children's temperament. Results Both children and parents overestimated children's ability, although children did so more than parents. Parents of temperamentally impulsive and undercontrolled children judged that their children could complete tasks that were actually beyond the child's ability. Temperament also affected children's judgments while parents were known to be present or absent: Temperamentally impulsive and undercontrolled children were more accurate in their judgments when parents were standing next to them than when parents were hidden from view behind a one-way mirror. Conclusions The mechanism by which parental supervision might protect children from injury appears to be at least twofold: (a) Parents overestimate children's ability less frequently than children themselves, suggesting supervising parents could intervene to prevent children from attempting dangerous activities; and (b) children judge their physical abilities more cautiously when parents are present. Implications for temperament theory and for injury prevention are discussed.
Key words: injury; temperament; ability estimation; parenting; supervision; children.
|
Introduction |
|---|
|
TopAbstractIntroductionThe Role of Parents...Objectives of the StudyMethodResultsDiscussionReferences |
|---|
Unintentional injuries are the leading cause of death for children ages 1–18 (National Safety Council, 2001
). In 1998, 42% of the deaths among American children ages
5–14 were due to unintentional injuries; among this age group, the death
rate from unintentional injuries out-numbered the death rate from the next ten
causes of death combined (National Safety
Council, 2001). The cost of nonfatal injuries is also staggering.
Societal financial expenses are estimated at $174 billion per year, and the
impact on education is estimated at 14 million lost school days annually for
children ages 6–16 (National Safety
Council, 1991). Despite these statistics, scientific understanding
of the behavioral factors that place children at risk for injury, and
therefore the means to develop empirically based interventions, remains
incomplete. |
The Role of Parents in Children's Risk-Taking Behavior |
|---|
|
TopAbstractIntroductionThe Role of Parents...Objectives of the StudyMethodResultsDiscussionReferences |
|---|
Due to a range of developmental, temperamental, and cognitive factors, young children sometimes become dysregulated in their management of risk-taking behavior (Byrnes, 1998
) and therefore experience increased opportunity for
unintentional injury. Two risk factors proposed to contribute to dysregulated
risk-taking behavior are: (a) cognitive immaturity on tasks, such as
overestimation of physical ability, and (b) behavior patterns linked to
temperamental traits such as impulsivity and undercontrol
(Byrnes, 1998). Young children
do not have fully developed cognitive strategies and therefore frequently
overestimate their ability to successfully complete basic physical tasks
(Plumert, 1995). Among
6-year-old children, the tendency to overestimate physical ability correlates
to history of major injuries requiring professional medical attention
(Plumert, 1995) and to minor
injuries incurred on a daily basis
(Plumert & Schwebel,
1997). Similarly, children with impulsive and undercontrolled
temperamental tendencies take risks that might contribute to unintentional
injury (Schwebel & Plumert,
1999). Longitudinal research suggests that early temperamental
impulsivity and undercontrol is related to later injury risk
(Schwebel & Plumert,
1999).
Because adults have better-developed cognitive and impulse control skills,
parental supervision is recommended as an effective strategy to prevent
childhood injuries (e.g., Gärling
& Gärling, 1993;
Morrongiello, Midgett, & Shields,
2001; Peterson, Cook, Little,
& Schick, 1991; Peterson,
Ewigman, & Kivlahan, 1993;
Wills et al., 1997). Recent
research suggests that the quality and quantity of parental supervision is
related to reduced history of injury
(Morrongiello et al., 2001)
and that parents who intervene promptly with children who engage in dangerous
activities have children who take fewer risks and experience fewer injuries
(Morrongiello & Dawber,
2000).
One unresolved issue is how and why parental supervision might reduce
injury risk. Two possibilities emerge. First, parents might prevent children
from injury by physically or verbally intervening when children begin to act
in a dangerous way. If parents have cognitive skills and impulse control that
are more mature than children, then parents might recognize the potential
danger of a situation that children judge to be safe and could prevent
children from engaging in that dangerous activity. Alternatively, the mere
presence of a parent may cause children to better utilize self-correcting
strategies to make safe, nonrisky decisions. In other words, children may feel
pressure to behave cautiously in the presence of a parent, whereas they might
behave in a more risky manner when alone (see
Berndt, 1979;
Smith & Howes, 1994).
Another unresolved issue is whether the efficacy of parental supervision
differs across types of children. Researchers have long known that individual
differences affect injury rate (Manheimer
& Mellinger, 1967), but it is unclear whether the success of
parental supervision might vary for children of different ages, genders, or
temperamental traits. Previous work suggests that parents estimate older
children's pedestrian safety (Dunne, Asher,
& Rivara, 1992) and knowledge of safety rules
(Yarmey & Rosenstein,
1988) more accurately than they do those of children ages 5 and 6
and that mothers of daughters judge videotaped playground situations as more
risky than do mothers of sons
(Morrongiello & Dawber,
2000), but the actual effect of supervision on children of
different ages, genders, or temperaments is unknown.
A third unresolved issue is whether the proximity of a parent to the child
influences the effectiveness of supervision
(Wills et al., 1997). Little
empirical work has addressed this possibility directly, although logic
suggests supervision conducted from a close proximity might be more effective
than distant supervision.
|
Objectives of the Study |
|---|
|
TopAbstractIntroductionThe Role of Parents...Objectives of the StudyMethodResultsDiscussionReferences |
|---|
The present study was designed to examine the above issues. Four topics were of particular interest.
Children's Estimation of Physical Ability
Previous work suggests that children tend to overestimate their ability to
complete basic physical tasks (i.e., they judge that they can complete
physical tasks that are actually beyond their ability;
Plumert, 1995).
Underestimation of ability (i.e., judging that they cannot complete physical
tasks actually within their ability) is uncommon
(Plumert, 1995). In this
study, we expected to replicate these findings. We also expected that
overestimation would be moderated by age, gender, and temperament: Younger
children, boys, and temperamentally impulsive and undercontrolled children
would have a greater tendency to overestimate ability.
Parent's Estimation of Children's Physical Ability
Previous work has not examined parents' estimation of children's physical
abilities. In this study, we expected that parents would tend to overestimate
their children's ability to complete basic physical tasks, although they would
do so to a lesser degree than children. We also expected that parents would
underestimate their children's ability to some degree. The role of children's
age, gender, and temperament was of interest, but because little is known
about these processes, no a priori hypotheses were made concerning
the role of these individual differences.
The Role of Parental Presence in Children's Estimation of Physical
Ability
The presence of a parent was hypothesized to lead children to be more
cautious in estimating their own physical abilities—that is, we expected
that the presence of a parent would lessen the extent of overestimation of
physical ability. The role of children's age, gender, and temperament was also
of interest, but no a priori hypotheses were made concerning the role
of these individual differences.
The Role of Parental Proximity in Parents' Estimation of Children's
Physical Ability
If parents were standing immediately adjacent to their children, it was
hypothesized that their judgment of their children's physical ability would be
more accurate than if they were watching from several feet away. Thus,
increasing the proximity of the parent to the child was expected to decrease
both parental overestimation and underestimation of children's physical
abilities. The role of children's age, gender, and temperament was also of
interest, but again a priori hypotheses were not made.
|
Method |
|---|
|
TopAbstractIntroductionThe Role of Parents...Objectives of the StudyMethodResultsDiscussionReferences |
|---|
Participants
Sixty-four children (33 six-year-olds, 31 eight-year-olds; 34 boys, 30 girls) participated in a one-hour laboratory battery. Recruitment was targeted toward a wide range of ethnic and socioeconomic groups, so the sample was racially and socioeconomically diverse. The 6-year-old group had a mean age of 79.86 months (SD = 4.71), included 19 boys and 14 girls, and was 61% white, 30% African American, and 9% other ethnicities. The 8-year-old group had a mean age of 103.16 months (SD = 5.34), included 15 boys and 16 girls, and was 42% white, 48% African American, and 9% of other ethnicities (see Table I). Seven-year-olds were not included in the sample because (a) we sought to test a relatively broad range of development within a moderately sized sample but still maintain sufficient statistical power, and (b) previous research along a similar line of inquiry (Plumert, 1995
;
Plumert & Schwebel, 1997)
used 6- and 8-year-olds but omitted 7-year-olds. Parents provided written
informed consent for the study; children provided either verbal (age 6 and
under) or written (age 7 and over) assent. All research was approved by the
university institutional review board. Families received a small compensation
for participating in the study.
|
Procedure and Measures
Estimation of Ability Task. To assess estimation of
physical ability, the accuracy of both children and parents in estimating
children's ability to complete a series of physical tasks was computed. Tasks
were sometimes prepared within children's ability and sometimes beyond. In
total, children completed six blocks of four tasks. A figure depicting the
apparatus and a detailed explanation of the protocol is available elsewhere
(Plumert, 1995). Briefly, the
four tasks were the vertical reach, the horizontal reach, stepping, and
clearance tasks. The vertical reach task involved removal of a small toy from
a shelf while standing on tiptoes. The horizontal reach task involved reaching
from a squatting position to retrieve a small toy off a block without touching
hands or knees on the floor. The stepping task involved stepping from behind
one stick attached to the floor over a second, parallel stick. The clearance
task involved moving under a flexible bar attached to two posts without
knocking the bar down or putting hands or knees on the floor.
Tasks were set at one of four difficulty levels: (1) the well-within level, 13% below children's estimated maximum ability level; (2) the just-within level, at children's estimated maximum ability level; (3) the just-beyond level, 8% beyond children's estimated maximum ability level; and (4) the well-beyond level, 13% beyond children's estimated maximum ability level. Each task was individually scaled for each child using previously measured estimates of maximum ability (explained below) for reaching, stepping, and crouching.
Tasks that were well within and just within children's ability assessed
underestimation tendencies. In other words, if children or parents incorrectly
judged that a child could not complete a task that was actually within his or
her ability, they were underestimating the child's ability. Although not
directly relevant to risk for injury, frequent underestimation of ability
might have implications for children's social and motor development (see
Schwebel & Plumert, 1999).
Tasks that were just beyond and well beyond children's ability measured
overestimation tendencies. Inaccurate estimations that the child could
complete tasks that were actually beyond his or her ability indicated
overestimation of the child's ability. Overestimation of ability by children
was of particular interest since it may lead to unintentional injury
(Plumert, 1995;
Schwebel & Plumert,
1999).
Estimates of children's maximum ability levels were obtained before the test trials in a separate room. The actions used to derive estimates of the children's maximum levels of ability were similar but not identical to those performed in the tasks. For the vertical reach maximum ability level measurement, children stood on tiptoes and reached above their head. The height of their reach was marked at the second joint of the index finger. For the horizontal reach measurement, children squatted along a line and reached forward to touch the top of the experimenter's pencil. The experimenter moved the pencil increasingly further until children could not touch it, and the maximum reach was recorded. For the stepping measurement, children stepped forward beyond the experimenter's pencil, which was placed on the floor along a ruler; the pencil was placed at increasingly far distances until children could no longer complete it and the maximum step was recorded. For the clearance measurement, children squatted down and were asked to "get as small as possible, like a ball"; the point one inch above the arch of children's backs was recorded as the maximum ability level. After estimates were taken, one experimenter prepared the apparatus in the testing room while a second experimenter conducted brief measures with children unrelated to the present hypotheses.
Once the apparatus was prepared, the experimenter led children into the testing room and gave instructions for the study. Parents learned directions simultaneously by watching the experimenter through a one-way mirror and listening on headphones attached to a microphone in the experimental room. Children were shown each of the four tasks and the experimenter explained how each was performed. Children were then presented with $12 of play money to use in the games. The experimenter explained that once children were positioned for the task, a timer would be set for 7 seconds. During those 7 seconds, children were instructed to think carefully about whether or not they could complete the task successfully. After the timer rang, the experimenter asked, "What do you think?" and children answered either affirmatively or negatively about their ability to complete the task. If children judged they could complete it, they were asked to attempt it. If they judged they could not complete it, the experimenter proceeded to the next task. Successful attempts were rewarded with an additional play dollar and unsuccessful attempts punished with the loss of a dollar. When children indicated that they could not complete the task, no money was exchanged.
During the 7-second latency, parents made their estimations about children's ability to complete the task. Parents had been previously instructed to circle their answers to two questions on a response sheet prepared for them: (a) Do you think the child will say he/she can complete the task? and (b) Do you think he/she can complete the task? Experimenters monitored parents to be sure they answered questions during the 7-second timer latency; on the rare occasion when parents responded late, those data were dropped from analyses.
Parents completed their ratings from one of two positions. For half of the six trial blocks, parents watched children from behind a one-way mirror, standing approximately 10 feet from the nearest task and 22 feet from the farthest. For the other half of the blocks, parents stood adjacent to children, within 3 feet of the task. Parental position was clumped so that the first three blocks were judged either from behind the window or standing near the child, and then the position was switched for the second three blocks. Initial parent location was randomly determined.
Each of the six trial blocks included each of the four tasks. Task order and task difficulty were randomly determined across trial blocks, with the stipulations that each child perform each task at each of six difficulty levels (i.e., once each at the well-within and well-beyond levels, and twice each at the just-within and just-beyond levels) and that each child complete each task once in each trial block. Between trial blocks, children worked with one experimenter on measures unrelated to the present hypotheses while a second experimenter adjusted the apparatus. Parents completed the temperament questionnaire described below ("Temperament") in a third room. This ensured that neither children nor parents would witness equipment adjustment and discover whether tasks might be easier or harder for the child. All activity in the testing room was videotaped through a one-way mirror to permit later coding.
After the six trial blocks were complete, children were permitted to "buy" prizes using the play money they had accumulated (in actuality, all children were given the same number of prizes). Children were also asked to perform the trials they had judged they were unable to complete successfully. This manipulation maintained the integrity of the game during the test trials but allowed the experimenter to determine whether tasks had been scaled correctly and to compare the accuracy of estimations with children's actual ability.
Incorrectly scaled tasks were removed from the data prior to analyses. A task was deemed correctly scaled if the child was able to perform a task that was at the wellwithin or just-within levels of difficulty or was unable to perform a task that was at the just-beyond or well-beyond levels of difficulty. The mean number of correctly scaled tasks for the well-within, just-within, just-beyond, and well-beyond levels of task difficulty was 3.86 (out of 4), 6.43 (out of 8), 6.25 (out of 8), and 3.52 (out of 4), respectively. After removal of incorrectly scaled tasks, accuracy proportions were calculated by dividing the number of correct estimations by the number of correctly scaled tasks at each level of difficulty.
Temperament. Parents completed an abbreviated version of
the Child Behavior Questionnaire
(Rothbart, Ahadi, & Hershey,
1994), a widely used measure of child temperament with good
psychometric properties (Rothbart et al.,
1994; Rothbart, Ahadi,
Hershey, & Fisher, 2001). Rather than asking parents to
complete the entire measure, four scales that assess temperamental traits
previously linked to unintentional injury
(Plumert & Schwebel, 1997;
Schwebel & Plumert, 1999)
were included: activity level, high-intensity pleasure, impulsivity, and
inhibitory control.
|
Results |
|---|
|
TopAbstractIntroductionThe Role of Parents...Objectives of the StudyMethodResultsDiscussionReferences |
|---|
Five issues were examined: (a) descriptive statistics and individual difference variables on children's temperament, (b) descriptive statistics and individual difference variables on children's estimation of ability, (c) descriptive statistics and individual difference variables on parent's estimation of children's ability, (d) whether presence of the parent affected children's estimation of their ability, and (e) whether standing near the child versus watching through the window affected parent's estimation of children's ability.
Children's Temperament
The four temperament scales had the following means, standard deviations,
and Cronbach alphas, respectively: activity level (4.84, 1.05, .88),
high-intensity pleasure (4.94, 1.02, .83), impulsivity (4.45, 0.85, .81), and
inhibitory control (4.89, 0.92, .84). Analyses of variance (ANOVAs) yielded no
significant age or gender differences on any of the scales. Because the four
scales were highly related to each other (with inhibitory control reversed,
Cronbach's alpha = .81) and because previous research suggests that all are
related to risk for unintentional injury
(Schwebel & Plumert,
1999), they were averaged into a single composite measure of
temperamental impulsivity/undercontrol for subsequent analyses. With
inhibitory control reversed, the average temperament score was 4.09
(SD = .77). An ANOVA yielded no age or gender effects.
Children's Estimation of Ability
For the well-within, just-within, just-beyond, and well-beyond levels,
respectively, the mean proportion of correct judgments by children was .97
(SD = .01), .95 (SD = .01), .31 (SD = .26), and .39
(SD = .29). Thus, children were quite accurate at judging tasks well
within and just within their ability, but children were much less accurate on
tasks beyond their ability.
To simplify analyses, scores on tasks within ability were aggregated to create a single measure of underestimation of ability, and scores on tasks beyond ability were aggregated to create a single measure of overestimation of ability. Correlations between the two within-ability scores and the two beyond-ability scores, respectively, were .47 and .62. Children were far more likely to judge ability accurately on tasks within their ability (M = .97, SD = .01) than tasks beyond their ability (M = .35, SD = .25), t(63) = 17.30, p < .01 (see Table I).
Age, Gender, and Temperament Effects. ANOVAs were conducted
to test for gender and age effects on the two estimation-of-ability scores.
Results suggest that girls tend to underestimate their ability (M =
.94, SD = .10) more than do boys (M = .98, SD =
.00), F(1, 63) = 4.26, p < .05; and 6-year-olds
(M = .29, SD = .24) tend to overestimate ability more than
8-year-olds (M = .42, SD = .24), F(1, 63) = 4.16,
p < .05. To test whether children's ability estimation was related
to their temperament scores, the aggregated temperament composite was
correlated to children's overestimation and underestimation scores with age
and gender controlled. Surprisingly, given previous findings to the contrary
(Plumert & Schwebel, 1997;
Schwebel & Plumert, 1999),
both correlations were non-ignificant and near zero.
Parent's Estimation of Children's Ability
For the well-within, just-within, just-beyond, and well-beyond levels,
respectively, parents' mean proportion of correct estimations was .85
(SD = .19), .73 (SD = .21), .61 (SD = .24), and .72
(SD = .26). Like the children, parents were quite accurate at judging
tasks well within, and to a lesser extent, just within, children's ability;
there were relatively few instances of underestimation of ability. On tasks
beyond children's ability, parents were more accurate than children, although
they still overestimated on nearly 40% of tasks just beyond their children's
ability.
As was done with the children's scores, scores on tasks within ability were aggregated to create a single measure of underestimation of ability, and scores on tasks beyond ability were aggregated to create a single measure of overestimation of ability. Correlations between the two within-ability scores and the two beyond-ability scores, respectively, were .32 and .23. Parents were more likely to judge accurately on tasks within children's ability (M = .79, SD = .16) than on tasks beyond their ability (M = .66, SD = .18), t(63) = 3.77, p < .01 (see Table I).
Age, Gender, and Temperament Effects. ANOVAs to test for gender and age effects on parents' underestimation and overestimation of children's ability yielded no significant results. To test whether parent's ability estimation was related to children's temperament scores, the aggregated temperament variable was correlated to parent's overestimation and underestimation scores, with age and gender controlled. The correlation with tasks within children's ability was nonsignificant, r(60) = –.05, but there was a significant negative correlation between parent's estimations on tasks beyond children's ability and children's temperament, r(60) = –.35, p < .01, suggesting parents of more controlled children judged children's ability more accurately than did parents of undercontrolled children. Put differently, parents of undercontrolled children expected that the children could complete tasks that they actually could not complete.
Comparison of Parent and Child Estimation of Ability. To compare parent's accuracy of estimating children's ability with the accuracy of the children themselves, a t test for tasks within children's ability yielded a significant difference, t(63) = 8.05, p < .01, with children (M = .97, SD = .01) judging more accurately than parents (M = .79, SD = .16). A similar test comparing estimation of tasks beyond children's ability was also significant, t(63) = –8.80, p < .01, although in this case parents (M = .66, SD = .18) judged more accurately than children (M = .35, SD = .25).
Children's Estimation of Ability as a Function of Parental
Presence
T tests were conducted to test for the effect of parental presence
on children's estimation of tasks within and beyond ability. Not surprisingly,
given the low rate of underestimation overall, tendency to underestimate
ability was similar across conditions of parental presence and absence,
t(62) = –.83, ns. On tasks beyond children's ability,
children whose parents were present in the room first tended to judge more
accurately (M = .42, SD = .27) than children whose parents
were absent first (M = .30, SD = .22), t(62) =
2.01, p < .05.
To further investigate the effect of condition and parental location, a repeated measures ANOVA was conducted with condition (parent present first vs. absent first) as a between-subjects factor, and estimation accuracy on tasks beyond ability during parent presence vs. absence as a within-subjects factor. It yielded a significant effect of condition, F(1, 62) = 5.93, p < .05, which was subsumed by a Condition x Estimation Accuracy interaction, F(1, 62) = 9.73, p < .01. As shown in Table II, children whose parents were present first judged 40% correctly but became more cautious afterward with parents absent (50% correct). Children who began the study with parents absent overestimated greatly (23% correct), but became more cautious with parents present (38% correct).
|
Age, Gender, and Temperament Effects. To test whether age or sex influenced the Condition x Estimation Accuracy interaction on tasks beyond ability, separate repeated measures ANOVAs were conducted with age or sex and condition as between-subjects factors and estimation accuracy on tasks beyond ability during parental presence vs. absence as a within-subjects factor. The ANOVA including age yielded the expected main effects of age, F(1, 60) = 4.68, p < .05, and condition, F (1, 60) = 6.58, p < .05, and interaction effect of condition by estimation accuracy on tasks beyond ability during parental presence versus absence, F (1, 60) = 9.71, p < .01, but did not indicate that children of different ages reacted differently to the presence of parents. The ANOVA including sex yielded the expected main effect of condition, F(1, 60) = 5.50, p < .05, and interaction effect of condition by estimation accuracy on tasks beyond ability during parental presence versus absence, F(1, 60) = 9.71, p < .01, but did not indicate that children of different sexes reacted differently to the presence of parents.
To test whether temperamentally impulsive and undercontrolled children might be more influenced by the presence of a parent than other children, a median split created groups of high and low temperamental control and a repeated measures ANOVA was conducted with condition and temperament as between-subjects factors and estimation accuracy on tasks beyond ability as a within-subjects factor. As previously found, there was a significant main effect of condition and a significant interaction effect of condition by estimation accuracy on tasks beyond ability with parents present vs. absent. Both findings were subsumed, however, by a Temperament x Condition x Estimation Accuracy interaction, F(1, 60) = 4.05, p < .05. As shown in Figure 1, temperamentally controlled children judged 49% of tasks beyond their ability correctly with parents present first and 47% with parents absent. Corresponding numbers for temperamentally controlled children with parents absent first were 36% and 25%. The two-way Condition x Overestimation of Ability interaction within the temperamentally controlled group was not significant in follow-up tests.
|
Temperamentally undercontrolled children judged 33% correctly with parents
present first and then improved to 53% after their parents were absent. Those
with parents absent first judged only 19% correctly initially with parents
absent but improved to 41% correct with parents present in the second half of
the study. The two-way interaction was highly significant, F(1, 29) =
12.54, p < .01. Thus, among the temperamentally controlled
children, the presence of a parent had little effect. Among temperamentally
undercontrolled children, however, the presence of parents had a great effect,
including a lasting effect in the group whose parents were initially present
and then departed. Learning Effect. One limitation of the
above analyses is that they do not account for learning that might occur: That
is, children might learn from the early tasks and therefore perform better on
later tasks regardless of whether parents were present or not
(Plumert, 1995). To test for
this possibility, a repeated measures analysis of covariance (ANCOVA) was
conducted with estimation accuracy on tasks beyond ability during the first
half of the tasks versus the second half of the tasks as a within-subjects
effect, and condition (parents present first vs. absent first) covaried. The
main effect of estimation accuracy was not significant, indicating that
children did not judge the first and second half of the tasks differently
after co-varying for the effects of condition. There was a main effect of
condition, F(1, 55) = 5.89, p < .05, with children whose
parents were in the room first judging more accurately across the study than
children whose parents were absent first.
In summary, it appears that the presence of a parent reduced children's overestimation of ability. The effect of parental presence persisted over the course of the study for those children whose parents departed after half of the study was complete. This pattern was particularly true for children who were temperamentally impulsive and undercontrolled. It did not seem to be affected by children's age or gender, and it occurred beyond the effect of children learning to complete tasks more accurately over the course of the experiment.
Parent's Estimation of Children's Ability as a Function of Proximity
to Child
Next, we considered whether parent's estimation of children's ability
differed depending on the location of parents. A t test comparing
parent's estimation of tasks within children's ability yielded a significant
difference: Parents were more accurate at judging tasks within children's
ability when they were standing directly next to children (M = .82,
SD = .18) than while watching children from several feet away through
a one-way mirror (M = .74, SD = .25), t(63) = 2.47,
p < .05. The corresponding t test comparing parent's
estimations of tasks beyond children's ability was not significant,
t(63) = .22.
To further investigate the finding on parents' estimation of tasks within children's ability, a repeated measures ANOVA with condition (parent present first vs. absent first) as a between-subjects factor and estimation accuracy of tasks within children's ability while near children (present) versus behind the window (absent) as a within-subjects factor was conducted. It yielded the expected effect of estimation accuracy, but the interaction effect was not significant. Age, Gender, and Temperament Effects. A repeated measures ANOVA was conducted with age and condition as between-subjects factors and parents' estimation accuracy of tasks within children's ability while near children versus behind the window as a within-subjects factor. It yielded the expected main effect of location of the parent, F(1, 60) = 6.05, p < .05, as well as an interaction effect of condition by age, F(1, 60) = 4.93, p < .05. Further investigation of the interaction effect suggested that parents of 6-year-olds who were in the room first estimated 81% of tasks within their children's ability accurately while parents of 6-year-olds who were behind the window first estimated 76% of tasks within their children's ability accurately. Parents of 8-year-olds showed the opposite pattern: Those in the room first estimated 72% of tasks within children's ability correctly while those behind the window first estimated 85% of the tasks correctly.
A similar repeated measures ANOVA was conducted with sex and condition as between-subjects factors and estimation accuracy of tasks within children's ability while parents were near children versus behind the window as a within-subjects factor. It yielded just one significant effect, a Sex x Condition x Location of the Parent interaction, F(1, 60) = 4.73, p < .05. Parents of boys correctly judged 87% of the trials when they were in the room first, but only 70% of the trials when they were behind the window. When they were behind the window first, parents of boys judged correctly on 82% of the tasks while in the room and 80% of the tasks while behind the window. A follow-up repeated measures ANOVA with the sample of boys, using condition as a between-subjects effect and parent location as a within-subjects effect, yielded the expected main effect of location, F(1, 32) = 4.39, p < .05, but the interaction was not significant.
Parents of girls showed a different pattern. For parents of girls who were in the room first, they correctly judged 74% of tasks correctly in the room and 74% of tasks when behind the window. Parents of girls who were behind the window first judged 72% of tasks correctly behind the window and 86% of tasks correctly while in the room. A followup repeated measures ANOVA with the sample of girls, using condition as a between-subjects effect and parent location as a within-subjects effect, yielded no significant results. Thus, although many of the trends did not reach traditional significance levels, parents of boys underestimated the most when behind the window, particularly after being in the room first. Parents of girls underestimated more frequently than parents of boys, except when they were in the room after underestimating from behind the window.
A final repeated measures ANOVA was conducted with condition and temperament as between-subjects factors and estimation accuracy of tasks within children's ability as a within-subjects factor. It yielded a significant main effect of estimation accuracy, but this effect was subsumed by a significant Temperament x Estimation Accuracy interaction, F(1, 60) = 9.41, p < .01. As shown in Table III, parents judged temperamentally controlled children with approximately the same accuracy whether they were in the room (M = .78, SD = .19) or behind the window (M = .80, SD = .22). For children rated as temperamentally undercontrolled, however, parents were far more accurate at judging ability while standing near children (M = .86, SD = .17) than while watching from behind the window (M = .69, SD = .27).
|
Learning Effect. To ensure that learning did not affect parent's ratings over the course of the study, a repeated measures ANCOVA with estimation accuracy of tasks within children's ability on the first half of the tasks versus the second half of the tasks as a within-subjects effect and condition (parents present first vs. absent first) covaried yielded a main effect of estimation that was subsumed by an Estimation Accuracy x Condition interaction, F(1, 62) = 6.06, p < .01. Parents who were in the room first averaged a mean of .81 (SD = .21) tasks judged correctly during the first half of the study, while parents behind the window first averaged a mean of .76 (SD = .26) tasks judged correctly during the first half. During the second half of the study, the proportion of tasks judged correctly were reversed: Parents who were in the room first judged a mean of .72 (SD = .25) tasks correctly while behind the window during the second half; parents who were behind the window first judged a mean of .84 (SD = .15) correctly while in the room for the second half. This pattern is not surprising: Stated simply, parents judged tasks within children's ability more accurately while in the room, and very little learning occurred over the course of the study.
In summary, results from the analyses of parent estimation suggest that parents were more accurate on tasks within children's ability while standing next to the children compared with standing behind the one-way mirror. This was particularly true for parents of children who were temperamentally impulsive and undercontrolled and for parents of girls who had previously underestimated their girls' ability from behind the window.
|
Discussion |
|---|
|
TopAbstractIntroductionThe Role of Parents...Objectives of the StudyMethodResultsDiscussionReferences |
|---|
Results suggest that both children and parents overestimate children's physical abilities, although children do so to a greater extent than parents. Children, and in particular temperamentally undercontrolled children, overestimate to a lesser degree while parents are in the room. Parents, and in particular parents of temperamentally undercontrolled children, underestimate children's ability more frequently while standing several feet away.
Parental Supervision, Temperament, and Children's Ability
Overestimation
Injury scholars generally concur that parental supervision is among the
best means to prevent childhood injuries
(Gärling & Gärling,
1993; Morrongiello et al.,
2001; Peterson et al.,
1991,
1993;
Wills et al., 1997), but the
present results are among the first to empirically address how and why
parental supervision might increase children's safety. Results suggest the
mechanism is at least twofold: a) Parents more accurately identify children's
physical limitations, and so presumably intervene to prevent dangerous
activities; and b) children engage in environments with greater caution when
parents are present, and this effect persists even after parents leave the
environment. The contribution of each factor depends on temperament. The
first, that parents can intervene to prevent injury, seems somewhat more
important for temperamentally controlled children, while the second, that
children behave differently in the presence of a parent, seems somewhat more
important for temperamentally undercontrolled children.
Along with poor parental supervision, temperament is among the most
implicated of risk factors for children's unintentional injuries
(Manheimer & Mellinger,
1967; Matheny,
1986; Schwebel & Plumert,
1999). The pathway by which an impulsive, undercontrolled
temperament might lead to injury risk is proposed to include multiple direct
and mediated pathways (Schwebel &
Plumert, 1999). One mediated pathway that has been proposed is
that overestimation of physical ability might mediate the relation between
temperament and injury: Children who are temperamentally undercontrolled might
overestimate their physical ability, which in turn leads to injury
(Schwebel & Plumert,
1999). The present results suggest parental supervision might
influence that mediated pathway, since impulsive and undercontrolled children
tend to overestimate ability less frequently in the presence of a parent. If
so, there are important implications to injury prevention: Parental
supervision might prevent temperamentally at-risk children from overestimating
their physical ability, which in turn might prevent those children from
injuring themselves.
Another way to conceptualize the role of temperamental differences in
injury risk is to consider Temperament x Environment interactions.
Extending the early "goodness of fit" model advocated by Thomas
and Chess (1977), temperament
theorists (Wachs & Kohnstamm,
2001) and empiricists (Bates,
Pettit, Dodge, & Ridge, 1998;
Schwebel, Brezausek, Ramey, & Ramey,
in press) have recently emphasized the importance of considering
Temperament x Environment interactions as predictors of children's
behavior. Parents generally recognize that the intensity of parental
supervision should vary based on the dangers in the child's immediate
environment. Thus, for example, young children are thought to need careful
supervision in the yard because it poses a high degree of risk for injury
(Peterson et al., 1993). The
present results suggest that perhaps supervision should vary not just in
response to the setting the child is engaging in but also in response to the
individual differences of children. Some children—temperamentally
controlled 6-year-olds or most 13-year-olds, for instance—do not need
careful supervision in playrooms that parents have prepared for safe play. On
the contrary, other children—temperamentally undercontrolled 6-year-olds
or most 3-year-olds, for instance—need relatively careful supervision in
all areas of the home, including the playroom. Thus, the interaction between
the setting the child engages in and any individual difference(s) within the
child might both contribute to the level of supervision the parents should
provide under different circumstances.
Implications for Injury Prevention
Injury scholars have long debated the relative value of environmentally
oriented injury prevention strategies versus person-oriented injury prevention
strategies (e.g., Christophersen,
1989). Environmentally oriented strategies, which generally take
the form of an alteration in the environment (e.g., placing mulch under
playground equipment), are geared to prevent injury in all individuals,
whether they are at high risk of injury or not. Person-oriented strategies
generally take the form of altering an individual's behavior through
education, behavior modification, or other means (e.g., teaching children not
to walk in front of another child riding on the swingset). Robertson
(1994,
1997) has argued that
person-oriented interventions will not be as effective as environmentally
oriented ones because most injuries are caused by the environment (e.g., the
safety equipment and design of a motor vehicle) rather than by individual
behavior. Present results affirm the possibility that individual behavior by
both the child and the parent is an important aspect of risk for injury and
suggest that both person- and environmentally oriented prevention strategies
should continue to be explored. As an example, interventions geared toward
parents of children with extreme levels of temperamental impulsivity and
undercontrol might be valuable. We have recently proposed that mental health
professionals might consider incorporating injury prevention techniques,
including education on the importance of parental supervision, into counseling
sessions with families of children with oppositional defiant disorder and
attention deficit /hyperactivity disorder
(Schwebel, Speltz, Jones, & Bardina,
2002).
Another group of children who might be specifically targeted for
intervention are those who have a history of frequent previous injury. In the
present study, neither children nor parents seemed to learn from their
estimation experiences; after controlling for condition, estimation accuracy
on the first three tasks was similar to accuracy on the later tasks. This
jibes with the finding that a previous history of injury is among the best
predictors of future injury (Bijur,
Golding, & Haslum, 1988;
Jaquess & Finney, 1994) and
suggests that the "repeat accident victim" should be targeted for
intervention efforts.
Unexpected Results, Limitations, and Future Directions
Although most of the results from this study replicate and support previous
work, one result was surprising and fails to replicate previous work:
Children's temperament was unrelated to a tendency toward overestimation.
Previous work found both concurrent
(Plumert & Schwebel, 1997;
Schwebel & Plumert, 1999)
and longitudinal (Schwebel & Plumert,
1999) links between temperamental impulsivity/undercontrol and
overestimation of ability. It is difficult to speculate what might have led to
the present findings. They may be spurious, due to peculiarities in this
sample. They may also represent the fact that children's overestimation of
ability is likely driven by a wide range of intrapersonal, interpersonal, and
environmental factors—with temperament playing only a modest role.
This study has limitations. Most prominently, it was a laboratory-based
experimental study. Although the experimental design provides the opportunity
to examine behavior in a controlled manner, it also presents drawbacks. The
consequences for erroneous judgments of physical ability in our laboratory
were minimal for children (loss of a play dollar) and essentially nil for
parents (emotional reactions in what most parents engaged in as a mildly
competitive activity). In real-world environments, where consequences for
mistakes are greater, children and parents might be more cautious and
therefore err more frequently on the side of ability underestimation
(Rosenstock, 1974). Further,
the experimental design led to the opportunity to conduct a large number of
analyses and therefore elevated the experiment-wide risk of Type I error.
Future directions warrant discussion, in closing. This study contributes to an important trend in injury research: How do multiple risk factors, influencing children through developmental stages, contribute to risk for unintentional injury? Longitudinal studies are sorely needed. They might follow individual differences (e.g., temperament, gender), social influences (e.g., parental supervision, peer relations), cognitive influences (e.g., children's estimation of physical ability), and community influences (e.g., socioeconomic status, race/ethnicity) with the objective of identifying how these factors interact to place particular children at increased or decreased risk for injury. Beyond further examination of etiological factors, empirically based research on the role of supervision in injury prevention would be valuable. How can parents be trained to supervise their temperamentally impulsive and undercontrolled children more closely? What type of supervision is sufficient to keep children safe? Finally, future research should continue to balance controlled laboratory-based studies such as the present one with studies having greater ecological validity.
|
Acknowledgments |
|---|
Thanks to Oni Hasan-McDade, Natasha Holloway, Sheryl Roque, Julie Sherrod, and Adam Summerlin for assistance with collecting and coding the data and to Michael Windle for comments on an earlier draft of the manuscript. Thanks to the administrators, parents, and children from Fair-field, Homewood, and Tarrant Public Schools for their assistance and participation. This research was supported by funds from the Department of Psychology and School of Social and Behavioral Sciences, University of Alabama at Birmingham.
Received September 24, 2002; revision received January 14, 2003; revision received March 3, 2003; accepted March 5, 2003
|
References |
|---|
|
TopAbstractIntroductionThe Role of Parents...Objectives of the StudyMethodResultsDiscussionReferences |
|---|
Bates, J. E., Pettit, G. S., Dodge, K. A., & Ridge, B. (1998). Interaction of temperamental resistance to control and restrictive parenting in the development of externalizing behavior. Developmental Psychology, 34, 982 –995.[CrossRef][Web of Science][Medline]
Berndt, T. J. (1979). Developmental changes in conformity to peers and parents. Developmental Psychology, 15 , 608 –616.[CrossRef][Web of Science]
Bijur, P. E., Golding, J., & Haslum, M. (1988).
Persistence of occurrence of injury: Can injuries of pre-school children
predict injuries of school-aged children? Pediatrics, 82
, 707
–712.
Byrnes, J. P. (1998). The nature and development of decision-making: A self-regulation model. Mahwah, NJ: Erlbaum.
Christophersen, E. R. (1989). Injury control. American Psychologist, 44, 237 –241.[CrossRef][Medline]
Dunne, R. G., Asher, K. N., & Rivara, F. P.
(1992). Behavior and parental expectations of child pedestrians.
Pediatrics, 89, 486
–490.
Gärling, A., & Gärling, T. (1993).
Mothers' supervision and perception of young children's risk of unintentional
injury in the home. Journal of Pediatric Psychology, 18
, 105
–114.
Jaquess, D. L., & Finney, J. W. (1994). Previous
injuries and behavior problems predict children's injuries. Journal
of Pediatric Psychology, 19, 79
–89.
Manheimer, D. I., & Mellinger, G. D. (1967). Personality characteristics of the child accident repeater. Child Development, 38, 491 –513.[CrossRef][Web of Science][Medline]
Matheny, A. P. (1986). Injuries among toddlers:
Contributions from child, mother, and family. Journal of Pediatric
Psychology, 11, 163
–176.
Morrongiello, B. A., & Dawber, T. (2000). Mothers' responses to sons and daughters engaging in injury-risk behaviors on a playground: Implications for sex differences in injury rates. Journal of Experimental Child Psychology, 76 , 89 –103.[CrossRef][Web of Science][Medline]
Morrongiello, B. A., Midgett, C., & Shields, R.
(2001). Don't run with scissors: Young children's knowledge of
home safety rules. Journal of Pediatric Psychology, 26
, 105
–115.
National Safety Council. (1991). Accident facts: 1991 edition. Chicago: Author.
National Safety Council. (2001). Injury facts: 2001 edition. Chicago: Author.
Peterson, L., Cook, S. C., Little, T., & Schick, B. (1991). "Mom lets me go there": The role of environment and supervision in children's minor injuries. Children's Environments Quarterly, 8 (3/4), 15–23.
Peterson, L., Ewigman, B., & Kivlahan, C. (1993). Judgments regarding appropriate child supervision to prevent injury: The role of environmental risk and child age. Child Development, 64 , 934 –950.[CrossRef][Web of Science][Medline]
Plumert, J. M. (1995). Relations between children's overestimation of their physical abilities and accident proneness. Developmental Psychology, 31, 866 –876.[CrossRef][Web of Science]
Plumert, J. M., & Schwebel, D. C. (1997). Social and temperamental influences on children's overestimation of their physical abilities: Links to accidental injuries. Journal of Experimental Child Psychology, 67, 317 –337.[CrossRef][Web of Science][Medline]
Robertson, L. S. (1994). Child injury control: Surveillance and research questions. American Journal of the Medical Sciences, 308, 88 –91.[Web of Science][Medline]
Robertson, L. S. (1997). Mistaken assertions on
reducing motor vehicle injury. American Journal of Public
Health, 87, 295
–296.
Rosenstock, I. M. (1974). The health belief model and preventive health behavior. Health Education Monographs, 2 , 354 –386.[Web of Science]
Rothbart, M. K., Ahadi, S., & Hershey, K. L. (1994). Temperament and social behavior in children. Merrill-Palmer Quarterly, 40, 21 –39.
Rothbart, M. K., Ahadi, S. A., Hershey, K. L., & Fisher, P. (2001). Investigations of temperament at three to seven years: The Children's Behavior Questionnaire. Child Development, 72 , 1394 –1408.[CrossRef][Web of Science][Medline]
Schwebel, D. C., Brezausek, C. M., Ramey, S. L., & Ramey, C. (in press). Interactions between child behavior patterns and parenting: Implications for children's unintentional injury risk. Journal of Pediatric Psychology.
Schwebel, D. C., & Plumert, J. M. (1999). Longitudinal and concurrent relations among temperament, ability estimation, and injury proneness. Child Development, 70 , 700 –712.[CrossRef][Web of Science][Medline]
Schwebel, D. C., Speltz, M. L., Jones, K., & Bardina, P.
(2002). Unintentional injury in preschool boys with and without
early onset of disruptive behavior. Journal of Pediatric
Psychology, 27, 727
–737.
Smith, E. W., & Howes, C. (1994). The effect of parents' presence on children's social interactions in pre-school. Early Childhood Research Quarterly, 9 , 45 –59.
Thomas, A., & Chess, S. (1977). Temperament and development. New York: Brunner.
Wachs, T. D., & Kohnstamm, G. A. (Eds.). (2001). Temperament in context. Mahwah, NJ: Erlbaum.
Wills, K. E., Christoffel, K. K., Lavigne, J. V., Tanz, R. R.,
Schofer, J. L., Donovan, M., et al. (1997). Patterns and
correlates of supervision in child pedestrian injury. Journal of
Pediatric Psychology, 22, 89
–104.
Yarmey, A. D., & Rosenstein, S. R. (1988). Parental predictions of their children's knowledge about dangerous situations. Child Abuse and Neglect, 12, 355 –361.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  B A Morrongiello, M Corbett, and R J Brison Identifying predictors of medically-attended injuries to young children: do child or parent behavioural attributes matter? Inj. Prev., August 1, 2009; 15(4): 220 - 225. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Damashek, N. A. Williams, K. Sher, and L. Peterson Relation of Caregiver Alcohol Use to Unintentional Childhood Injury J. Pediatr. Psychol., May 1, 2009; 34(4): 344 - 353. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. A. Sturm, R. M. Mays, B. Juliar, R. M. Bair, S. Perkins, and G. D. Zimet `What, Me Worry?': Parents' Reasons for Worry and Nonworry about Adolescent Risk for Sexually Transmitted Infections (STI) J Health Psychol, November 1, 2008; 13(8): 1060 - 1071. [Abstract] [PDF]
|
|
|
|
|
|
B. A. Morrongiello and S. Matheis Understanding Children's Injury-risk Behaviors: The Independent Contributions of Cognitions and Emotions J. Pediatr. Psychol., September 1, 2007; 32(8): 926 - 937. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. R. Porter, L. A. Crane, L. M. Dickinson, J. Gannon, J. Drisko, and C. DiGuiseppi Parent Opinions About the Appropriate Ages at Which Adult Supervision Is Unnecessary for Bathing, Street Crossing, and Bicycling Arch Pediatr Adolesc Med, July 1, 2007; 161(7): 656 - 662. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. K. Barton and D. C. Schwebel The Roles of Age, Gender, Inhibitory Control, and Parental Supervision in Children's Pedestrian Safety J. Pediatr. Psychol., June 1, 2007; 32(5): 517 - 526. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. K. Barton and D. C. Schwebel The Influences of Demographics and Individual Differences on Children's Selection of Risky Pedestrian Routes J. Pediatr. Psychol., April 1, 2007; 32(3): 343 - 353. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. A. Morrongiello, M. Corbett, M. McCourt, and N. Johnston Understanding Unintentional Injury Risk in Young Children II. The Contribution of Caregiver Supervision, Child Attributes, and Parent Attributes J. Pediatr. Psychol., July 1, 2006; 31(6): 540 - 551. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. C. Schwebel, A. L. Summerlin, M. L. Bounds, and B. A. Morrongiello The Stamp-in-Safety Program: A Behavioral Intervention to Reduce Behaviors that Can Lead to Unintentional Playground Injury in a Preschool Setting J. Pediatr. Psychol., March 1, 2006; 31(2): 152 - 162. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. A. Morrongiello Caregiver Supervision and Child-Injury Risk: I. Issues in Defining and Measuring Supervision; II. Findings and Directions for Future Research J. Pediatr. Psychol., October 1, 2005; 30(7): 536 - 552. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. C. Schwebel and B. K. Barton Contributions of Multiple Risk Factors to Child Injury J. Pediatr. Psychol., October 1, 2005; 30(7): 553 - 561. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. C. Schwebel and C. M. Brezausek The Role of Fathers in Toddlers' Unintentional Injury Risk J. Pediatr. Psychol., January 1, 2004; 29(1): 19 - 29. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||