x

Objectives. The number of identified gifted children in Slovakia is small and Slovak psychologists need more modern and specialized assessment tools. Test for Identifying Gifted Children in Mathematics in Grades 3–5 (TIM3–5) is standardized in Czechia and has excellent psychometric properties with two parallel and fully equated forms. The study aims to adapt the test in the Slovak language and support its practical use. The authors ran a small pilot study using form A only and compared it to the Czech standardization sample. Sample and settings. The Slovak sample consisted of 169 pupils from four elementary schools and was supplemented by the Czech standardization sample (404 children in form A). The Slovak data were collected in February 2022; informed consent from the parents/legal guardians was obtained before data collection with approx. 40 % drop-out. The original authors of the test provided the Czech data collected in 2015. Hypotheses and analysis. The authors compared descriptives of Slovak and Czech pupils, assessed essential psychometric parameters (reliability and factor validity), and mainly performed measurement invariance and Differential Item Functioning analyses. Results. The psychometric parameters of the Slovak test form were excellent, fully comparable, or even better than the original Czech version. The mean of IRT reliability across grades was .76. However, the test differentiates mainly in above-average children. The confirmatory IRT analyses suggested clear unidimensionality and scalar invariance across the Czech and Slovak samples. However, Slovaks outperformed Czech pupils. The difference was highest in the third grade with 9.6 and 95%CI = [6.7–12.1] points at the T-score scale and lowest in the fifth grade, 3.6 with 95%CI = [0.9–6.3] points. The authors recommend the test for practical use using Czech norms. However, a user should be aware of possible differences in average performance, considering that the Czech norms could be too mild for Slovak children. Limits. Such a massive difference between Czech and Slovak populations is not plausible. The authors provided several explanations based mainly on the sampling procedure and systematic missingness in the Slovak sample, correlated to math ability. The most realistic explanation of the difference is a systematic sampling error in one or both samples. Therefore, the results are of importance for Czech users as well since it might be the case that the norms are too mild also for the Czech pupils. The authors advise Czech test users to interpret the test results with caution (and rather conservatively) until a new norming study is performed.