We then averaged the weekly sales growth rates over a four week win-dow before and after the week in which the seller got his first, second, third, fourth and fth negative feedback;24 and conducted paired t-tests of the null hypothesis of equality of growth rates before and after the negative feedback event