First, calculate the mean:
( 4(0.25)+5(0.44)+2(0.80)+1(5.10) ) / (4+5+2+1) = 9.9 / 12 = 0.825
Next, consider the effect of removal on each of the cost of a postcard, a letter, a large envelope, or a small package.
If a postcard was removed from the receipt, then the new mean would be
( 3(0.25)+5(0.44)+2(0.80)+1(5.10) ) / (3+5+2+1) = 9.65 / 11 = 0.88
If a letter was removed from the receipt, then the new mean would be
( 4(0.25)+4(0.44)+2(0.80)+1(5.10) ) / (4+4+2+1) = 9.46 / 11 = 0.86
If a large envelope was removed from the receipt, then the new mean would be
( 4(0.25)+5(0.44)+1(0.80)+1(5.10) ) / (4+5+1+1) = 9.1 / 11 = 0.83
If a small package was removed from the receipt, then the new mean would be
( 4(0.25)+5(0.44)+2(0.80)+0(5.10) ) / (4+5+2+0) = 4.8 / 11 = 0.44
As you can see, the item which caused the largest change in the mean (or skew) was the small package. That's because the small package was much more expensive than any of the rest of the items. For future problems like this one, try examining the values of your dataset and focusing on the outlier. Hope this helps!