Question

A researcher studies alcohol's effect on reaction time and finds no difference between people who consumed 1 versus 2 beverages. She included 40 participants randomly selected from the dining hall on campus, with a range in age from 18 to 38 years and a range in weight from 100 to 325 pounds. She decides to rerun her study with 30 women from 18 to 22 years of age and who are all of normal body weight and finds there is a statistical difference in reaction times between those who consumed 1 versus 2 beverages. Why might her results have changed

Answer 1

Her results changed because the power increased due to the fact that she reduced variability in her data as a result of using a sample that had lesser variability.

Explanation:

She is trying to find if there is difference between people who consumed 1 versus 2 beverages.

Now initially she surveyed 40 people between ages 18 to 38 years and of huge weights and found out there was no difference. However, she decided to run the survey again on only women aged 18 to 22 years with normal body weight.

Since she used a lesser sample size and surveyed only women, it means there will be lesser variability in the result as the sample is smaller and streamlined to only women.

Thus, her result changed because the power increased due to the fact that she reduced variability in her data as a result of using a sample that had lesser variability.