3.2.14 Validation Study of the HOPE Scale: Identifying Gifted Students from Low-Income and Multicultural Families

In the field of gifted education, the underrepresentation of Black, Latinx, and Native American students has been one of the primary concerns among the researchers (Moon & Brighton, 2008). Because of the challenges inherent in identifying gifted students from low-income families and those from minority backgrounds, HOPE Scale is designed to more equitably identify students’ giftedness based on teachers’ observation (Peters & Gentry, 2010; 2012). In the U.S., the scale is proved to be a good measure, however, I aimed to find out whether the scale is still valid in different county as well. In this study, 1,157 students’ data was collected from public elementary schools in South Korea from 55 teachers (216 students were from low-income, and 167 students were from multicultural families). Students’ HOPE Scale score which is composed of academic and social scores was first compared by income and ethnic groups to see whether the true difference exists, then Confirmatory Factor Analysis (CFA) as well as Multi-group CFA were conducted to see construct validity evidence. In order to see the association with other variables (e.g., reading and math scores), Structural Equation Modeling (SEM) is added along with teachers’ interview data (N=6) who shared their opinions and perception regarding gifted students from low-income and multicultural families. The final version of the basic CFA model yielded a chi-square value 361.084 (df=39, p<.001), with the other indices showing a good fit (NFI=.961, CFI=.965, IFI=.965, and RMSEA=.0845). The alpha reliability estimates for the two factors were 0.916 and 0.838 respectively for the academic and social constructs which indicating strong internal consistency. In MCFA, two different income groups showed equal factor loadings, equal indicator intercepts and equal indicator for error variances (X^2=210.610 (df =104, p<.001), NFI=.972, CFI=.986, IFI=.986, and RMSEA=.0683). Similarly, ethnicity groups showed equal factor loadings, equal indicator intercepts, but not equal error variances (X^2=243.830 (df =96, p<.001), NFI=.958, CFI=.974, IFI=.974, and RMSEA=.096). As discussed above, although the measurement itself was a valid tool for equitably identify gifted students with diverse backgrounds, still there was a gap between groups. Students from low-income family yielded with lower scores (M=39.04, SD=11.11) compared with students from regular-income family (M=33.27, SD=11.31) in total HOPE Scale score at t(1155)=6.857, p<.001. Similar pattern was found in ethnic groups that students from multicultural family background scored lower (M=35.78, SD=11.06) than regular Korean students (M=38.33, SD=11.38) at t(1155)=2.685, p<.001. Further findings will be shared in the presentation.

Author(s):