A researcher performed an experiment with two groups. She found the difference of the means for each group. Then she combined the groups, chose two new groups, and found the difference between the means of those groups. She repeated this process 200 times. The normal distribution of the difference in the means she found is given below. How great would the difference in means between the first two groups have to be in order to be considered significant?

At least 8
At least 10
At least 9
At least 7