A researcher is studying the impact of sleep deprivation on cognitive performance. They plan to use an independent measures design, with one group sleeping for 4 hours and another for 8 hours. To ensure the two groups are as comparable as possible at the start of the study, what is the most effective technique they should use?