Inference is needed on the difference between the mean math SAT score of high school students who went to a 10-minute presentation (μ₁, group A) and the mean math SAT score of high school students who did not go to the 10-minute presentation (μ₂, group B). A 95% confidence interval on the difference of two means gave (25.15, 38.69). What would be the impact of reversing the population mean definitions: defining μ₁ as for group B and μ₂ as for group A?