The world recently lost Judge Joseph Wapner, a man of integrity who brought evenhandedness and a gentle humor to the proceedings of The People’s Court. Along with judicial wisdom, he could have low tolerance for the many idiots who came through his court, and always found a way to bore right to the core issues. He probably would have had low tolerance for the recent dealings we had with two agrichemical giants over an herbicide ad one of them apparently ran in parts of Illinois. For the sake of anonymity, we can call these two giants, “large company S” and “large company D”. Both companies sell very broad spectrum, multi-component preemergence corn herbicide premix products, and of course both “large company S” and “large company D” believe their products walk on water and are truly superior to all other preemergence corn products. The ad in question, produced by “large company S”, showed combined yield results from a 2016 study conducted by OSU and Purdue that compared weed control and yield from various premergence corn herbicides. As shown in the ad bar chart and wording, the product of “large company S” resulted in a “yield advantage” compared with use of the “large company D” product. “Large company D” asked us to review the ad for accuracy. Upon reviewing the data from the two individual OSU and Purdue studies, we found that there was no significant difference in yield between the two products that walk on water in either study. Nonetheless, someone at “large company S” HQ decided to average yields between studies and create an ad that made it look as though there was higher yield for their product. We do have to respect the ability of “large company S” to fess up when we confronted them with this. Their response was essentially “yup you caught us. We ignored the statistics and just floated it out there anyway figuring that no one would notice”.
In the meantime, it was pointed out to us that while trying to occupy most of the high moral ground here, “large company D” was known to be cutting recommended rates of their product that walks on water in certain areas to be more cost competitive with the “large company S” product. And in the end “large company S” essentially reminded us of the possibly also somewhat dubious approach of “large company D” in selling product and stated “the attorneys for large company D can contact our attorneys if they want to pursue this further”. We believe in the legal world these are known as the “wait – but they did this” and “nananana poo poo – just try to get us” arguments. Both very effective at times we’re sure. However we can’t help thinking that in this case even Judge Wapner’s response might have been, “Bailiff, kick these people in the a** and get them out of my court”. (In the interest of honesty, which is what this blog post is about, we should say we borrowed this last quote from a Doonesbury comic strip. Always wanted to use this somewhere.)
We have been doing this for 25+ years, and have observed numerous instances where companies manipulated data somewhat or ignored statistics to create the story they want. This manipulation can be the only reason why herbicide ads for a given product always show that product as either more effective or providing higher yield than the competitors’ products. In all fairness to companies, it probably doesn’t make much sense to run an ad that shows the opposite. In fact, we had this same discussion with “large company S” a couple of years ago, with regard to another of their ads. The ad also showed a bar chart where their preemergence corn product that walks on water outyielded the competitors’ products in an OSU trial. Checking our data, it was apparent that there was no significant yield difference, but you never would have known this from the ad, where the bar for their product was a different color and substantially taller than the other bars. Because different colors of course do indicate that significant differences occurred regardless of what the statistics showed. Which leads us to ask “large company S” (and everyone else really), if you’re going to ignore the statistics that in fact show whether real differences occurred, why even bother to have independent researchers such as OSU and Purdue conduct the study? Why not just make up the results so that everyone knows your product is truly the best?
We really have better things to do than try to police ads where our data are misrepresented. It’s not a very rewarding activity. We don’t want to say it’s a mistake to ever trust the yield results shown in herbicide ads – no wait – let’s go with that – it’s a mistake to ever trust yield results shown in herbicide ads. And hey – just a reminder – experimental design and statistical analysis procedures exist for a reason. If we ignore them, then all we’re left with are “alternative facts”, right?