Measuring Positive and Negative Affect and Physiological Hyperarousal Among Serbian Youth

Abstract

This study extended previous cross-cultural work regarding the tripartite model of anxiety and depression by developing Serbian translations of the Positive and Negative Affect Scale for Children (PANAS–C), the Physiological Hyperarousal Scale for Children (PH–C), and the Affect and Arousal Scale (AFARS). Characteristics of the scales were examined using 449 students (M age = 12.61 years). Applying item retention criteria established in other studies, PH–C, PANAS–C, and AFARS translations with psychometric properties similar to English-language versions were identified. Preliminary validation of the scales was conducted using a subset of 194 students (M age = 12.37 years) who also completed measures of anxiety and depression. Estimates of reliability, patterns of correlations among scales, and age and gender differences were consistent with previous studies with English-speaking samples. Findings regarding scale validity were mixed, although consistent with existing literature. Serbian translations of the PH–C, PANAS–C, and AFARS mirror the original English-language scales in terms of both strengths and weaknesses.

Notes

When used together, the PANAS–C and PH–C are sometimes referred to as the PH–PANAS–C (e.g., Kiernan et al., 2001).

Results reported in the article are based on the full sample. A reviewer of an earlier version of the article suggested that the sample be split and parallel analyses be conducted. Results from analyses conducted with the randomly split sample are available from the authors. The findings for the split samples were similar to those reported for the entire sample with respect to item–total correlations and alpha coefficients. The results for factor analyses followed the general pattern of the PAF conducted with the full sample, although the magnitude of factor loadings tended to fluctuate more in the split samples. This was likely due to smaller sample sizes in the split sample analyses. Modest communalities for the PAF with the AFARS, in particular, would indicate that using the full sample to maximize size was preferable to split sample analyses (MacCallum, Widaman, Zhang, & Hong, 1999).

The final structure matrices for the PH–PANAS–C and AFARS principal-axis factor analyses requesting a three-factor solution are available from the authors.