A week ago I started a small quality control project where I measured twenty “1 kg” Albert Heijn (AH) potato bags in order to assess whether or not AH is systematically underfilling them, as some social media posts had implied. Surely you have mulled over what the results would look like, and it is now time for the big reveal. Every bag was measured three times with kitchen scale A, and three times with kitchen scale B. These measurements were highly consistent and differed by a few grams at most; therefore I will present the weight for each bag averaged across the six measurement occasions. Also, I have subtracted 5 grams from each average to account for the plastic wrapper, whose weight I determined separately. Here is a JASP raincloud plot of the results:
The results are clear from the descriptive plot and do not require statistical inference for a proper understanding:
- The AH clearly overfills the bags. The median weight is 1023 grams, and none of the 20 bags weighs less than 1017 grams. Perhaps the overfilling strategy is used to prevent complaints, reputation loss, and fines for violating the legal norms. Based on these measurements I speculate that the target value for filling the 1 kg bags is set either to 1020 or to 1025 grams.
- The bags are filled with great precision, as the weights are tightly clustered around the median of 1023 g. Such high precision may have been achieved by first filling smaller containers (hoppers), determining their weight, and then combining their contents in order to get as close as possible to the target weight. This process is known as multihead combinatorial weighing.
- The data show a single outlier and it is cause for some concern. Yes, the outlier overfills at 1082 g, but it also suggests that the system does in fact glitch — and that it may possibly also glitch in the opposite direction, leading to a substantial underfilling of the bag. Presently I think it is more likely that the filling system tolerates the positive glitches and filters out the negative glitches, but without further information (e.g., knowledge of the filling routine, or additional measurements) we cannot be certain.
In sum, after measuring twenty bags of potatoes, we can only be impressed with the precision with which the bags are filled. More importantly perhaps, these data should nip any potato riot in the bud. As it turns out, the AH is not ripping off the Dutch by selling them short on potatoes. It is rather the other way around: at the AH, you get about 2.3% more potatoes than what’s on the label!
Authors
Eric-Jan Wagenmakers (CEO) and the JASP Services BV Team
