Identifying Hispanic Gifted Children: A Screening

Stength of Correlations

Voelker and Orton (1993) stated that the strength of a correlation depends greatly on the phenomenon being investigated. When correlations are composed of scores obtained by subjective means, such as the T.A.R.G.R.T. R teacher rating checklist, then lower correlations (e.g., .50 to .60) may be indicative of a strong relationship (Kubiszyn & Borich, 1996). Based on the interpretive guide for correlation coefficients from Schiavetti and Metz (1997) the following correlation coefficients for interpretation were based on alpha significance levels of .05 and .01.

Outcomes

There was a moderate correlation with alpha at the 0.05 level for the following: (a) Stanford Reading and Intellectual Characteristics (r=0.443) with r^sup ^^ of 0.196 indicating a shared variance of 20%, (b) Stanford Reading and Total Checklist (r-0.352) with r^sup 2^ of 0.124 indicating a shared variance of 12%, and (c) Stanford Math and General Characteristics (r=0.369) with r^sup 2^ of 0.136 with a shared variance of 14%. There was a strong relationship with alpha at the 0.01 level for the following: (a) Stanford Reading and Academic Characteristics (r=0.535) with r^sup 2^ of 0.286 with a shared variance of 29%, (b) Stanford Math and Intellectual Characteristics (r=0.660) with r^sup 2^ of 0.436 with a shared variance of 44%, (c) Stanford Math and Motivation Characteristics (r=0.465) with r^sup 2^ of 0.216 with a shared variance of 22%, (d) Stanford Math and Leadership Characteristics (r=0.600) with r^sup 2^ of 0.360 with a shared variance of 36%, (e) Stanford Math and Academic Characteristics (r=0.739) with r^sup 2^ of 0.546 with a shared variance of 55%, and (f) Stanford Math and Total Checklist (r=0.615) with r^sup 2^ of 0.378 with a shared variance of 38%. These results are reported in Table One.

Conclusions and Discussion

This study asked the relative question of the effectiveness of standardized tests measuring giftcdness with Hispanic children. It was felt that other indicators of performance, i.e., teacher ratings, could supplement existing procedures in identifying Hispanic students for gifted programs. It was specifically hypothesized that teacher ratings of behaviors would be useful in screening for giftedness ofHispanic students. Thus, the purpose of this study was to investigate the relationship between standardized test scores and teacher ratings of behaviors in a group of Hispanic children in a small, rural school district.

Fourteen Pearson Product Correlations were calculated. Of these 14 comparisons involving Stanford Reading and Stanford Math scores and the T.A.R.G.E.T. checklist nine correlations were significant at the .05 or .01 level of significance (9/14=64%). Of these, three comparisons were found to be moderately significant (3/14=21%), while six were found to be highly significant (6/14=43%).

Moderate relationships were found between the Reading and the Total teacher checklist scores and the Math and the General teacher checklist scores indicating that general characteristics and averaged checklist scores (across the entire checklist) are not strongly sensitive to discriminate behaviors. Peterson (2000) found that teachers are the gatekeepers for selective programs in schools, particularly, when students are culturally different from them. Thus, global or averaged scores may not be sensitive when teachers are rating children from other cultural and linguistic backgrounds.