Introduction

The Teacher-Report Scale (TRS) is a measure to evaluate reactive aggression (RA) and proactive aggression (PA) in childhood. Nevertheless, several studies with TRS show correlation between RA and PA ranging from moderate (.41) to high (.83). Therefore, its capacity for reliably distinguishes individuals along the two-factor model is still an issue. This distinction, however, is useful to understand the structure and the function of aggression in childhood. Thus, being highly associated may not be enough to constitute a one-dimensional measure. Further evidences are needed, considering the relevance of a two-dimensional model.

Objectives

Analyze the relevance of the two-dimensional model for TRS in two different samples of children.

Method

Confirmatory factorial analyses (CFA) were used to compare the fit of the model in aggressive children (n= 250, boys= 80%, mean of age= 10, SD= 3.3) and children from general population sample (n= 597, boys= 51%, mean of age= 10, SD= 1.3). For the CFA, robust estimations were used in the RStudio.

Results

For the children’s aggressive samplea moderate correlation was found, and a two-factor solution was clearly supported (r= .48 (p<.001), x²= 16.99 (p<.01), df= 7, CFI= .97, TLI= .93, AIC= 4162.09, BIC= 4231.86, RMSEA= .06, SRMR= .03). For the children from a general population sample, the correlation was higher, and a two-factor solution was not so consistent (r= .79 (p<.001), x²= 45.96 (p<.001), df= 8, CFI= .88, TLI= .78, AIC= 4140.45, BIC= 4185.59, RMSEA= .14, SRMR= .08).

Conclusions

The relevance of a two-dimensional model may lie in the purpose of the measure. When its main objective is to distinguish the structure and function of anexisting aggressive behavior trait, a two-factor solution is more appropriate. When evaluating general population, however, another fit may be more statistically suitable.