Few broad observational measures of preschool-age children's temperament and behavior currently exist. Studied the Child Temperament and Behavior Q-Set (CTBQ-Set) as a naturalistic observation measure to tap the major domains of temperament and behavior in preschoolers. Pairs of observers rated the behavior of a community sample of preschoolers during 2 independent home visits using q-sort methodology. The CTBQ-Set scales displayed good levels of internal consistency and convergent and discriminant content saturation. The scales displayed good interrater reliability at each observation and moderate test-retest stability between observations. The construct validity of the CTBQ-Set scales was supported by correlations with mothers' ratings of their children's behavior using the Child Behavior Checklist for 2- and 3-year-olds (CBCL/2-3) and the Children's Behavior Questionnaire. The CTBQ-Set shows promise as an observer-rated measure of preschoolers' behavior and temperament in their natural home environment.

Examined proximal and contextual factors most strongly related to externalizing behavior among young children growing up in low-income, mother-headed families. Participants were 50 low-income single mothers and their preschool-age children who were visited twice in the home setting. Measures of proximal (low levels of supportive parenting, high levels of punitive disciplinary practices, low levels of maternal emotional well-being) and contextual (low maternal support, high levels of family stress) risk were assessed in relation to maternal reports of child externalizing behavior and an index of negative child behavior during a clean-up task. Child defiance during the clean-up task was highly associated with punitive maternal control in the same situation but had no other direct correlates. However, multiple risk factors representing both proximal and contextual variables were associated with variations in children's behavior problem scores. Mothers of children with high behavior problem scores reported lower feelings of self-efficacy in handling child care and emotional stressors, more frequent use of punitive child disciplinary practices, and lower feelings of satisfaction with the quality of their supportive resources than others. Maternal self-evaluations of coping efficacy mediated the relation between perceived support and child behavior problems, suggesting that constructs of personal control are important to represent in future studies of highly stressed parents.

Examined reliability and validity of the parent version of the Child Symptom Inventory (CSI-4) in 247 boys between 6.0 and 10 years 11 months old referred for evaluation of behavioral and emotional problems. The CSI-4 is a behavior rating scale whose items correspond to the symptoms of disorders defined by the Diagnostic and Statistical Manual of Mental Disorders (4th ed. [DSM-IV]; American Psychiatric Association, 1994). Results indicated satisfactory internal consistency reliability, test-retest reliability, and temporal stability over a 4-year period for most symptom categories. CSI-4 ratings converged and diverged in a theoretically consistent pattern with respective scales of the Child Behavior Checklist (CBCL; Achenbach, 1991a) and the Diagnostic Interview for Children and Adolescents-Revised-Parent Version (DICA-P; Reich, Shayka, & Taibleson, 1991). Discriminant validity was established in that boys with specific DICA-P diagnoses received significantly higher corresponding CSI-4 parent symptom ratings than boys not so diagnosed. Clinical utility (sensitivity, specificity, positive predictive power, negative predictive power) was evaluated for screening cutoffs based on categorical (DSM-IV) and dimensional (normative distribution of Symptom Severity scores) scoring methods.

Many multivariate statistical techniques that previously appeared prohibitive to the applied researcher are now becoming increasingly accessible given the advances and user-friendly nature of the graphic interfaces associated with many software applications. Structural equation modeling (SEM) is one such technique that has seen an exponential growth in its development and usage. However, up to less than a decade ago, most of the texts devoted to SEM were relatively complex, with a sound foundation in matrix algebra necessitated. However, there has been a spate of introductory texts that have inundated the marketplace recently, each with their own emphasis (e.g., type of software) and philosophy. The purpose of this article is to provide a comprehensive review of the various introductory texts, some of which are software specific and some that are more general. Also, for those who have gained more than a cursory familiarity with SEM, a brief review of more advanced texts is provided.

Studied 4th- through 8th-grade students in a rural school district (N = 382, 201 girls; M age = 12.48) who, at the end of the academic year, completed self-report measures assessing functional impairment, self-concept, reactance, and autonomy experiences in relation to parents. Confirmatory factor analysis supported a distinction between 2 aspects of autonomy: mutuality (perceptions of parents as encouraging of autonomy) and conflictual dependence (negative representations of "self-in-relation-to-parent" and related feelings of anger and shame). Path analysis findings mostly were consistent with a mediated model predicting personality from autonomy experiences and functional impairment from personality. However, conflictual dependence rather than negative self-feelings directly predicted girls' emotional difficulties. Comparisons of total effects suggested that conflictual dependence may have stronger and more consistent implications for adolescent psychopathology than mutuality. In general, significant relations between conflictual dependence and psychopathology remained significant even after controlling for a measure of negative affectivity (NA).

The Emotion Expression Scale for Children (EESC) is a new self-report scale designed to examine 2 aspects of deficient emotion expression: lack of emotion awareness and lack of motivation to express negative emotion. Validity was assessed using self-report measures of emotion regulation and self- and peer-report of internalizing and externalizing symptoms. Using a community sample of 208 fourth- and fifth-grade children, reliability analyses revealed high internal consistency and moderate test-retest reliability of the EESC. The results provide initial support for concurrent validity for the EESC factors evidenced by relations with measures of emotion management. Associations were found between the EESC and measures of internalizing symptoms.

Evaluated the psychometric properties of a newly created measure of cognitive vulnerability to depression for use with adolescents. Previous measures have shown poor internal consistency reliability and have not completely assessed all hypothesized components of cognitive vulnerability. High school students completed questionnaires assessing cognitive vulnerability to depression, negative life events, depressive symptoms, and general internalizing and externalizing symptoms. The Adolescent Cognitive Style Questionnaire (ACSQ) demonstrated excellent internal consistency reliability and good test-retest reliability. Confirmatory factor analysis showed there were 3 latent factors to the ACSQ. Construct validity was supported by significant correlations with another attributional style questionnaire, as well as with depressive and internalizing symptoms. The interaction of ACSQ with negative events significantly predicted concurrent depressive and internalizing symptoms but not externalizing problems. Last, cognitive vulnerability mediated the gender difference in depressive symptoms. Overall, results suggest that the ACSQ is a highly reliable and valid measure of cognitive vulnerability to depression in adolescence.