Journalists are less likely to show confirmation bias when they choose what studies to cover before knowing the results.