A company decides to test two package designs for canned soup by placing cans with each design at opposite ends of the soup aisle in a random sample of 26 supermarkets. Data were collected on weekly sales for each soup. Differences in sales were symmetric, but not normally distributed. The rank sum of the positive differences was 300, and the rank sum of the negative differences was 51. Is there sufficient evidence to suggest that sales of the soup in the two different packages differ? Use α = 0.05.