The problematic nature of culture-fair testing: Training effect differences among Czech and Roma children
Abstrakt
The performance in intelligence tests is considered to be influenced by the environment, however some of the tests are called culture-fair (CF). These tests endeavour to eliminate culture influence as far as possible.
One of the most popular tests are Raven's Progressive Matrices (RPM). Unfortunately some authors use CF tests to cultural comparison, which can lead to misleading results.
In this preliminary study, we have verified the potential of RPM to measure native culture-independent cognitive ability using two groups of children coming from different socio-cultural backgrounds (Czech and Roma). This potential was evaluated using a dynamic testing approach (i.e., pretest, training phase, and retest).
Czech children were more successful both at pretest and retest. In contrast, the level of improvement (i.e., retest minus pretest) was significantly higher in the group of Roma children.
Thus, an innate cognitive inferiority cannot explain the poorer results of the Roma children; rather, the results seem to reflect other factors, such as socio-cultural background. In summary, these preliminary data indicate that RPM success is determined by multiple factors.
Thus the performance on intelligence tests, including CF tests, must be considered within scope of specific socio-cultural background of the participants.