Imagine an experiment with one independent variable: noise level (quiet vs. noisy). If the researcher is using a between-subjects design and wants to have 20 participants per condition, how many participants will he need all together?