Imagine that you are testing the relative effectiveness of a new reading program for English-speaking vs. bilingual children using a quasi-experimental design. What are some of the steps you would take to obtain equivalent groups for your experiment?