RESUMO
Two ways of increasing the initial interaction among integrated EMR students and their nonretarded classmates in a game-playing situation were examined. Three hundred and four randomly assigned intermediate school students observed classmates about to play a simple game and were asked to select partners and opponents from two pairs of students matched on sex and grade. The experimental pair had 1 EMR student and 1 nonretarded student; the control pair had 2 nonretarded students. The subjects' past knowledge of the pair members served as a covariate in the design to determine whether past knowledge influenced their selections of pair members. Each subject was also provided information regarding the competency level of each pair member in relation to the particular game. Results indicated that the covariate was not a significant factor in the selection process; however, the appended competency statement and the game-playing option were found to be highly significant in influencing selection of the pair members and increasing the selection of the EMR child by his nonretarded classmate.