Publication details

Preschooler´s Ability Rating Scale (PARS): Searching for the most appropriate form of parent rating scale for assessing giftedness



Year of publication 2016
Type Appeared in Conference without Proceedings
MU Faculty or unit

Faculty of Social Studies

Description Although according to Pfeiffer (2002) rating scales are the second most used type of diagnostic tools intended for identification of giftedness besides intelligence tests, they are often criticized due to the lack of validity. Some studies (eg. Miller, Manhal, & Mee, 1991) suggest that parents are able to assess the ability of their children only to some degree of accuracy. It appears that distortion in the assessment may also be caused by variables related to the evaluator. The aim of the study is to verify, if selecting the appropriate form of rating scale could reduce this distortion. For this purpose Preschooler's Ability Rating Scale (PARS) was developed. This method is designed to identify gifted children aged 4-6 years by their parents and consists of two parts. The first one focuses on the development of the child - it detects in which age period specific skills and particular behavior related to giftedness were first observed. The second part is similar to classic rating scales. During the verification of the method (N=263) part of rated children were also assessed by intelligence tests WJ IE II COG (n=90) and UNIT (n=94) to get objective view on their abilities. To capture potentially confounding variable related to the evaluator the FMPS questionnaire of perfectionism was administered to some parents (n=144). Exploratory factor analysis (maximum-likelihood method with geomin rotation) proved three factor structure of the method. The first factor consists of 4 items from developmental part, which are focused on the reading and counting. The second one is comprised by the remaining 8 items of the developmental part, probably due to of their same formal appearance. The same applies in the case of the third factor, which consists of all 12 items of the rating part. Regression and correlation analysis proved that developmental part (expressed by two first factors) has higher concurrent validity with intelligence tests than the rating part (third factor). The strongest relation was proved with the first (reading and counting) factor (eg. r=.44 for the total IQ of WJ). On the contrary unlike the developmental part the third factor correlates with total score of FMPS questionnaire (r=.21). Thus it appears, that classical form of rating scales could be more affected by other variables related to the evaluator than developmental form. These conclusions could have significant implications for further development of rating scales in this area.
Related projects: