Mean (SD)
Range
Number of items
Reliability α
f
Significance of age as covariate
Gender main effects
Mean by gender
Female
Male
Community problem behavior
2.12 (1.02)
1–7
6
0.82
4.64* (548)
ns
8.52*
2.00 (0.06)
2.26 (0.06)
Alcohol use
2.49 (1.65)
1–7
3
0.81
4.73** (548)
5.86*
3.84t
2.61 (0.94)
2.34 (0.10)
School problem behavior
1.35 (0.68)
1–7
3
0.74
10.65** (548)
ns
21.28**
1.29 (0.04)
1.49 (0.04)
High-risk problem behavior
2.51 (2.56)
0–12
12
0.75
6.62** (548)
ns
12.62**
2.16 (0.15)
2.93 (0.16)
Sad depression
2.46 (1.20)
1–7
6
0.87
14.39** (547)
ns
28.59**
2.71 (0.07)
2.17 (0.07)
Angry depression
2.68 (1.40)
1–7
3
0.81
0.24NS (548)
ns
ns
2.65 (0.08)
2.73 (0.09)
Anxiety
3.01 (1.09)
1–7
5
0.68
2.44NS (545)
ns
4.54*
3.10 (0.06)
2.90 (0.07)
Self-esteem problems
3.12 (0.51)
1–7
7
0.81
3.67* (543)
ns
6.15*
3.17 (0.03)
3.06 (0.03)
To create the externalizing behavior scales (community problem behavior, school problem behavior, and substance use), a factor analysis was conducted on the 12 items that were created to assess the level that adolescents in the study engaged in problem behaviors (Note: Factor analyses will not be reported here to save space; these analyses can be obtained from the authors). The items asked adolescents to indicate how often in the past 6 months they had engaged in a variety of illegal and risky behaviors. Responses were given on an 8-point Likert-type scale (1 = never, 2 = once, 3 = 2 to 3 times, 4 = 4 to 6 times, 5 = 7 to 10 times, 6 = 11 to 20 times, 7 = 21 to 30 times, and 8 = 31 or more times). Community problem behavior consisted of the items related to engaging in risky behavior, doing something dangerous for the thrill of it, disobeying parents, damaging property, lying to parents, and having contact with the police. These items involve problem behaviors that adolescents engage in at home and in the community. The alcohol use scale consisted of the items related to getting drunk, drinking alcohol, and skipping school, all items involving alcohol use and behaviors associated with the use of substances (i.e., skipping school in order to use substances). School problem behavior consisted of items related to being suspended from school, fist fighting, and being sent to the principal’s office. As can be seen on Table 1, these behaviors were relatively infrequent for this, and there are some gender and age differences in frequency of behavior. Because of the low frequency of these behaviors across the sample, we created a high-risk problem behavior variable that consisted of the number of times an adolescent gave a response on a problem behavior item that fell in the upper quartile for the sample. This item yielded responses from 0 to 12 and further descriptive characteristics for this variable can be found in Table 1.
We also ran a factor analysis on the 9-item depression scale on which adolescents provided responses on a 7-point Likert scale with options ranging from 1 = almost never to 7 = almost always (derived from the Symptoms Checklist-90; SCL-90-R; Derogatis, Rickesl, & Rock, 1976) and created two subscales: angry depression and sad depression. Angry depression included the items related to feeling so angry that you wanted to smash or break something, feeling like you couldn’t control your temper, and feeling as if you wanted to hit or hurt someone, while sad depression included items related to feeling hopeless, not caring anymore, feeling very sad, having thoughts of ending your life, and feeling unhappy. Again, as can be seen on Table 1, there were gender differences in sad, but not angry, depression, and no age differences.
The anxiety scale was comprised of six items that assessed social (e.g., How much do you worry that boys/girls don’t like you?) and academic (e.g., How much do you worry about doing badly on tests? How fast does your heart beat when you take a test?) anxiety with responses given on the same 7-point Likert scale used for the depression scale. There were no age but significant gender differences in anxiety symptoms across the sample, with girls reporting higher levels of anxiety than did boys.
Self-esteem was measured using seven items from the Harter self-worth scale (Harter, 1982). This scale consists of seven forced-choice items that range from 1 (very much not like me) to 4 (very much like me). Items assess children’s general self-evaluations of the way they do things, feel about themselves, act, and whether they feel like a good person. Higher scores reflect more self-esteem problems. Analyses revealed gender but not age differences on this measure as reported on Table 1.
Continuity of Behaviors from Childhood to Adolescence
Our first question was whether or not there is continuity of behavior patterns from childhood to adolescence. Is that shy child more likely than the more outgoing child to become anxious or depressed in adolescence? Does that difficult, impulsive, aggressive kid end up delinquent or abusing substances in adolescence?
To begin to look at this question, we examined the relation between parents’ ratings of their child’s personality and temperament when they were in elementary school and the adolescent outcomes. The parents of the children in our sample rated their children on 44 personality/temperament items when the sample was in 3rd, 4th, and 6th grades. We combined mothers’ and fathers’ ratings when available, or used either mother or father ratings if the other were not available. A factor analysis of the 44 items was conducted that yielded seven factors, accounting for about 60 % of the variance: difficult personality (noisy, impulsive, and physically aggressive), prosocial (leader, assertive, outgoing, athletic, competitive), sensitive (shy, nervous), intellectual (persistent, intelligent), artistic (imaginative, creative, curious), caring (affectionate, caring), and relaxed (flexible, laid back). We then constrained the model to three factors: externalizing (difficult), internalizing (sensitive, artistic, caring, and relaxed), and prosocial (prosocial and intellectual). The following table contains the correlations of these three child personality/temperament constructs as reported by parents with the adolescent outcomes (Table 2).
Table 2
Correlations of parents’ ratings of childhood behavior with adolescent outcomes
Externalizing behavior problems in elementary school | Internalizing behavior problems in elementary school | Prosocial behaviors in elementary school | |
|---|---|---|---|
Community problem behavior | ns ns | ns ns | ns ns |
Alcohol use | ns 0.23* | ns 0.23* | ns 0.20* |
School problem behavior | ns ns | ns ns | ns ns |
High-risk problem behavior | ns ns | ns −0.17* | ns ns |
Sad depression | ns ns | ns ns | ns ns |
Angry depression | ns 0.38*** | ns ns | ns ns |
Anxiety | ns 0.32** | ns ns | ns ns |
Self-esteem problems | ns | ns | ns |
0.38*** | 0.20* | −0.33*** |
As can be seen, none of the behaviors rated by parents of girls were related to later adolescent behaviors. In contrast, parents who rated their sons as more externalizing (i.e., being more difficult, aggressive, noisy) during childhood as compared to sons who were rated lower on these items had adolescents who indicated they used alcohol more, had more angry depression, had higher levels of anxiety, and had more self-esteem problems. Boys rated as having more internalizing problems (e.g., sensitive, artistic, shy) by their parents were more likely to have fewer high-risk problem behaviors but more self-esteem problems as compared to boys rated with fewer internalizing problems. Finally, boys who were identified as presenting several prosocial behaviors, including being a leader, athletic, and outgoing, were more likely to use alcohol, and have fewer self-esteem problems in adolescence than were boys with fewer prosocial behaviors.
So, when pediatricians experience children in their waiting rooms who are difficult or fearful or outgoing, how these behaviors manifest themselves in adolescence depends on the gender of the child and the kind of problem they are exhibiting. Boys who exhibit more externalizing behaviors may be most vulnerable to problems in adolescence, especially mood problems and alcohol use. Our findings are consistent with others who have found that difficult child temperament measured even as early as six months of age predicted later externalizing behavior problems at age 17 (Olson, Bates, Sandy, & Lanthier, 2000). Although no one has looked at gender differences in continuity of behavior over time, Schonberg and Shaw (2007) looked at the relationship between early temperament (measured at age 2) and later conduct problems (measured at ages 5 and 12) in a group of low-income boys. The researchers found a significant relationship between the measures at age 5, but not age 12. It is interesting to note that boys rated high on externalizing behaviors in childhood reported more internalizing behaviors (depression, anxiety, self-esteem problems) in adolescence. It may be that the externalizing behaviors were a reflection of the boys’ negative emotions during childhood, or that the boys developed these internalizing problems in adolescence in response to parental interventions throughout their childhood. This is an area that needs further exploration.
Parenting Practices and Beliefs in Childhood
Does parenting matter? In her second edition 2009 book, The Nurture Assumption, Judith Rich Harris challenges the assumption that parenting matters in how a child turns out. Her view is that the countless studies that have demonstrated how various parenting practices and behaviors are correlated with a variety of child outcomes fail to control for genetic influences. Many healthcare providers familiar with Harris’ work may counsel parents to not worry so much about what they do as parents and instead watch the child’s peer group, which she claims is the most important shaper of the child’s behavior. Nevertheless, many researchers have questioned Harris’ analyses and find that parent behaviors and practices have a profound impact on children’s academic outcomes (Bronstein et al., 1996; Steinberg, Lamborn, Dornbusch, & Darling, 1992), their mental health (Aseltine, Gore, & Colten, 1994; Lewinsohn, Gotlib, & Seeley, 1997; Steinberg, Lamborn, Darling, Mounts, Dornbusch, 1994), and their involvement with delinquency and alcohol use (Barnes, Farrell, & Banerjee, 1994; Baumrind, 1991). In addition, during regular pediatric checkups, parents often report to healthcare providers that they and their child have a lot of conflict and that they are worried about their child’s behavior. How realistic are these conflicts and fears? Do they foreshadow some problematic behavior in the future, or should healthcare providers not be concerned?
In our work, we examined parenting strategies from early childhood when the children were in kindergarten, first, and third grades and two groups of strategies emerged: harsh and gentle discipline. Harsh discipline included items that assessed parents’ use of threats, criticism/anger, showing disappointment, withdrawal of affection, guilt, and physical punishment to control child behavior. The second factor, labeled gentle discipline, consisted of the items related to the use of explanation/reasoning, praise/affection/kindness, and tangible rewards to control child behavior. Because this construct did not map onto any of the outcomes of interest, we only looked at the harsh parenting strategies. We looked at the correlation of these practices for mothers and fathers with both behaviors reported 3 years later (prosocial, externalizing, internalizing) and the adolescent outcomes for boys and girls separately. As can be seen on Table 3, not surprisingly the harsh parenting practices were more strongly related to childhood temperament than to later adolescent behaviors. Mothers’ harsh parenting was related to both boys’ and girls’ prosocial (negatively related), externalizing, and internalizing problems. Fathers’ harsh parenting was related to girls’ prosocial and internalizing behaviors and boys’ externalizing behaviors. The stronger relationships between mothers’ parenting practices and the behaviors may reflect the fact that the mothers in this sample reported being the primary caretakers and disciplinarians of their children. Interestingly, although the harsh discipline strategies were related to childhood behaviors, they were related to only a few adolescent problem behaviors. Most notably, fathers’ harsh discipline strategies were related to boys’ community and school problems in adolescence.
Table 3
Correlations of parental practices and beliefs in childhood and adolescent outcomes
Harsh discipline moms (Wave 1–2) | Harsh discipline dads (Wave 1–2) | Maternal conflicted relationship with child (Wave 4) | Paternal conflicted relationship with child (Wave 4) | Maternal worry about child (Wave 4) | Paternal worry about child (Wave 4) | |
|---|---|---|---|---|---|---|
(N girls ≈ 192) | (N girls ≈ 145) | (N girls ≈ 155) | (N girls ≈ 98) | (N girls ≈ 155) | (N girls ≈ 98) | |
(N boys ≈ 157) | (N boys ≈ 127) | (N boys ≈ 144) | (N boys ≈ 102) | (N boys ≈ 144) | (N boys ≈ 102) | |
Childhood personality/temperament (Waves 1 and 2) | ||||||
Prosocial | −0.19** | −0.14* ns | −0.27*** ns | ns −0.22* | −0.30*** −0.21** | −0.23* −0.28** |
−0.16** | ||||||
Externalizing | 0.27*** 0.24*** | ns 0.18* | 0.38*** 0.38*** | ns 0.31** | 0.48*** 0.39*** | 0.34** 0.44*** |
Internalizing | 0.21*** 0.19*** | 0.16* ns | 0.23** ns | ns 0.22* | 0.21** ns | ns ns |
Adolescent behaviors | ||||||
Community | ns | ns | ns | ns | 0.23** | ns |
Problem behavior | ns | 0.16t | ns | ns | ns | ns |
Alcohol use | ns ns | ns ns | 0.20* ns | ns ns | 0.24** ns | ns ns |
School problem
Related posts: Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
Get Clinical Tree app for offline access
| ||||||
