Examining gender and ethnic differences in Child Behavior Questionnaire Outcomes: A multivariate approach

• In Event: 3-005 - R-E-S-P-E-C-T Individual Differences: A Methods Approach for Examining Culture and Gender

Integrative Statement

Introduction
As part of a longitudinal study involving temperament and psychopathology conducted in a large metropolitan area of the Mid-Atlantic region of the United States, 779 infants were brought into our laboratory at 4 months of age for temperament. Following the screenings at 4-months, 315 (151 males, 164 females) were selected to continue with the study. These infants were again assessed throughout childhood and into adolescence. At age 5 (mean = 5.2 years; 54% female, 46% male), participants were administered the Child Behavior Questionnaire (CBQ, n = 240). This assessment is a highly differentiated measure of temperament and has been shown to correlate highly with measures of personality and cognition, such as surgency and extraversion, negative affectivity, and effortful control.
The purpose of this paper is to first, provide descriptive information on whether males or females perform differently on outcomes of the CBQ. Secondly, to identify whether ethnicity impacts outcomes from the CBQ. And lastly, to conduct a descriptive discriminant analysis (DDA) to determine how gender and ethnicity uniquely contribute to scores from the CBQ.
Methods
Multivariate analyses are particularly important when dealing with large datasets and longitudinal data. Multivariate analysis of variance (MANOVA) is one way to determine whether differences exist between independent groups on multiple dependent outcomes. This is an omnibus test statistic, therefore, MANOVA cannot tell you where group differences exist (i.e., males vs. females) on statistically significant outcomes.
The current study used MANOVA to assess whether changes in gender (male, female) and ethnicity (African American/Black, Asian, Caucasian, Hispanic, other) have significant effects on CBQ outcomes. MANOVA also yields relationship measures between both CBQ outcomes and gender/ethnicity. It is expected that gender will affect outcomes related to externalized behaviors (activity level, high intensity pleasure, and fear). Ethnicity is hypothesized to be a strong factor in internalized behaviors (discomfort, low intensity pleasure, and fear).
Results
The two-way MANOVA was statistically significant. There was a statistically significant difference in CBQ outcomes based on participants’ gender, F(7, 224) = 2.75, p < .01; Wilk's Λ = .921, partial η2 = .08. Additionally, ethnicity also yielded a significant difference in CBQ outcomes, F(28, 809.07) = 7.76, p < .01; Wilk's Λ = .81, partial η2 = .05. Tests of between-subjects effects indicated statistically significant outcomes between gender and inhibitory control (p = .04), low intensity pleasure (p < .01), and perceptual sensitivity (p = .01). Ethnicity also yielded statistically significant findings with CBQ outcomes, specifically on activity level (p < .01) and fear (p < .01).
The follow up analysis to this study will focus on descriptive discriminant analysis (DDA) to identify which gender (male vs. female) and what ethnicities (African American/Black, Asian, Caucasian, Hispanic, other) contributed to the above results. This robust post hoc analysis evaluates the relevance of the dependent variables in accounting for group differences. Both standardized discriminant functions and structure coefficients will be consulted. Coefficients will identify which variables are the dominant contributors for each dependent variable.

Author

• Kaylee Rae Seddio, University of Maryland- College Park
  Presenting Author