Mixed-Methods Approach to Developing the Spanish Form of the Adjustment Scales for Preschool Intervention

• In Event: Advancements in Assessing the Early Educational Experiences of Diverse Children and Families Living in Poverty

Abstract

Objectives
Gaps exist in the availability of tools to accurately assess the socioemotional needs of diverse populations, especially Latino children living in low-income households. In response to this need, we undertook a careful multi-step process to develop a culturally- and linguistically-comparable Spanish version of the Adjustment Scales for Preschool Intervention (ASPI; Lutz, Fantuzzo, & McDermott, 2002). The ASPI is a multisituational assessment originally developed for a predominantly English-speaking urban Head Start population in the Northeast. It permits teacher assessment of overactive and underactive behavior within the context of 24 routine classroom learning and social interactions.

Theoretical framework
We employed a multi-step adaptation and translation process as recommended best practice to develop a measure for a population distinct from the original normative sample (Alegria, Vila, Woo, Canino, Takeuchi et al., 2004; Pena, 2007 (see Figure 2). Our goal was to obtain, as close as possible, three dimensions of equivalence across English and Spanish language forms: semantic, content, and technical equivalence.

Method
The English ASPI form was translated into Spanish by several translators, who represented diverse cultural, linguistic, and educational backgrounds. Translations were reconciled by a panel of experts along several dimensions including grammatical and linguistic correctness, generalizability, and semantic equivalence to ensure that the language would be easily understood by the intended population of teachers. Construct equivalence was balanced against linguistic equivalence. The Spanish ASPI form was then back-translated by two independent bilingual individuals. Discrepancies were reconciled.

Data sources
The Spanish form was piloted with 53 bilingual Spanish speaking Head Start lead teachers who completed both the English and translated Spanish ASPI forms (N = 582 children). Teachers provided feedback via focus groups and written feedback.

Results
Item analysis across the two language forms suggested differences in (a) prevalence (percent of items endorsed by teachers) and (b) relative rank ordering of the least and most commonly endorsed items. Consistency of rankings was tested through Spearman’s rho. Intraclass correlations examined associations between item prevalence across English and Spanish ratings (Shrout & Fleiss, 1979). Whereas prevalence and rank order across language forms were not identical, Spearman’s ρ indicated a significant and positive relationship (ρ = .70, p < .001 and ρ = .68, p < .001 for all pairwise comparisons for the least and most commonly endorsed behaviors, respectively). Table 5 shows the 20 least commonly endorsed items.
Several guiding themes emerged from our qualitative analysis, which informed our translation process and scale refinement. Teachers provided feedback on: (a) developmental appropriateness and cultural relevance, (b) generalizability, (c) linguistic complexity, academic level of vocabulary, and accuracy of the translation; and (d) the intensity of the translated behavior in Spanish versus English.

Significance
This multi-step process represents an effort to expand the availability of tools for use by bilingual Spanish speaking Head Start teachers. Future directions will include additional refinement in response to psychometric analyses, careful examination of distinctions between language forms with regard to teacher language preference and proficiency, and large-scale data collection to resolve dimensionality and examine criterion-related validity.

Authors

• Rebecca Jane Shearer, University of Miami

• Mileidis Gort, The Ohio State University

• Paul A. McDermott, University of Pennsylvania

• Michael L. Lopez, Nat. Center for Latino Child & Family Research