Cultivating Strength-Based Assessment: Psychometric Evaluation of the Social-Emotional, Evidence-Based Developmental Strengths (SEEDS) Youth Self-Report with Ethnically Diverse Grade School Youth

ABSTRACT

Distillation methodology as applied to youth mental health evidence-based approaches has allowed for identifying technique commonalities across hundreds of protocols and programs. In distillation methodology, “practice elements” refer to discrete clinical strategies (e.g. time out, relaxation) that are used as part of a larger intervention plan, such as a manualized treatment program. Assessments that measure children’s use of practice elements derived from the evidence-base exist, but typically target single areas and do not provide comprehensive simultaneous screening across numerous practice elements. The Social-Emotional Evidence-Based Developmental Strengths (SEEDS) is a new assessment tool, which provides broad screening across numerous important practice element domains. The SEEDS has demonstrated various forms of reliability and validity, but only with a Korean language version with Korean samples of children and adolescents. The current study examined the extent to which the SEEDS was able to perform as a reliable and valid measure with an ethnically diverse U.S. sample of fourth through sixth graders (N = 402, mean age = 10.63, SD = .86). Confirmatory factor analyses pointed to a three higher-order factor with both 13- and 12-lower-order factor structure fitting the data. Collectively across all subscales, scales, and the Total score, internal consistency reliabilities mostly fell in the acceptable to excellent ranges. Additionally, all of the convergence patterns between the SEEDS first- and higher-order subscales, scales, and the Total score with the Strengths and Difficulties Questionnaire (SDQ) subscales and Total score were significantly correlated in the expected directions. Implications and suggestions for future research are discussed.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ 4/5 items are anchored this way; one of the five items for this scale is scored in the secondary way of scoring SEEDS items (i.e., 0 = Almost Never to 3 Very Often)..

² We imposed an inequality constraint to solve the issue of negative error variance. However, this did not change the goodness of fit values reported below..

³ Per the direct request from the SEEDS developer, we cannot report standard deviations at this time owing to copyright issues..

⁴ Although no pairwise combinations against the SDQ Prosocial scale were statistically significant at the scale level, there were a few pairwise differences at the subscale level. Given no a priori reasons for hypothesizing these differences and the early stage of this research, these results are not presented here but available upon request.: