Origin of differences between Serbian and Roma children in social intelligence test's achievement

Abstract

The goal of the presented research is to identify the origin of differences between Serbian (N=166) and Roma children (N=169) of primary school age in achievement on Modified Rosenzweig Test, used as a measure of social intelligence (Roma children had lower results than Serbian children). Results show that these differences can be partly explained with extreme inviolable socio-economic and cultural status of Roma children. Social intelligence test's scores are significantly correlated with socio-economic and cultural status; this correlation is more apparent in Roma than in Serbian sample. Differences between Serbian and Roma children become less apparent when we control the influence of socio-economic and cultural status. Parent's educational level is recognized as the most important indicator of socio-economic and cultural status. Another way to identify the origin of differences was directed to investigation of adequacy of used test as a measure of social intelligence of Roma children. Our assumption that some items make whole test unfair for Roma children is not confirmed. However, results concerning the relation between experience with particular social situations and success in particular test items, and results concerning the different structures of implicite understanding of social intelligence by Serbian and Roma parents, point out that test key favors responses that Serbian people accept as optimal in task situation, while the responses according to Roma implicite understanding of social intelligence are less valued.Cilj prikazanog istrazivanja je identifikovanje izvora razlika u postignucu srpske (N=166) i romske dece (N=169) osnovnoskolskog uzrasta na modifikovanom Rozencvajgovom testu, koriscenom za merenje socijalne inteligencije (romska deca postizu nize rezultate u odnosu na srpsku decu). Rezultati pokazuju da jedan od izvora ovih razlika lezi u izrazito nepovoljnom socio-ekonomskom i kulturnom statusu Roma. Postignuce na testu socijalne inteligencije je u znacajnoj korelaciji sa socio-ekonomskim i kulturnim statusom, a ova korelacija je izrazenija na romskom delu uzorka nego na srpskom. Kada se kontrolise efekat socio-ekonomskog i kulturnog statusa, razlike izmedju srpske i romske dece se smanjuju. Medju pokazateljima socioekonomskog statusa kao najznacajniji se izdvaja obrazovanje roditelja. Drugi pravac traganja za izvorima razlika bio je usmeren na preispitivanje adekvatnosti samog testa za merenje socijalne inteligencije romske dece. Ocekivanje da ce se naci stavke koje test cine nepravednim prema romskoj deci nije potvrdjeno. Medjutim, nalazi o uticaju poznatosti situacija iz testnih stavki na uspesnost u njihovom resavanju i nalazi o razlicitoj strukturi implicitnog razumevanja socijalne inteligencije kod srpskih i romskih roditelja ukazuju na to da kljuc za ocenjivanje odgovora favorizuje ono sto u srpskoj populaciji predstavlja optimalan odgovor u zadatoj situaciji, a manje vrednuje odgovore koji su u skladu sa implicitnim razumevanjem socijalne inteligencije kod romskih roditelja.