I saw a TV commercial the other day for some prescription drug that softens nose hair or whatever. For those outside the USA, yes, drug companies are allowed to advertise prescription drugs directly to consumers here. Anyway, the ad said that it's not a weight loss drug, but people taking it lost "on average up to 12 pounds."
My question for you is what on EARTH does "on average up to" mean? Keep in mind that companies must be able to back up their claims with clinical studies so this must be some actual statistic from the study. To me, "on average" is clearly a mean and "up to" is clearly a maximum, so "on average up to" sounds self-contradictory.
I have a theory, but I'm curious to read what my fellow nerds think.
And as an aside, I posted this question on facebook about a week ago. The comments are almost entirely about how companies advertise, the problems of advertising to consumers, etc. It appears that even my normally smart friends drew a blank on the statistic and really didn't care. Personally, I always look/listen for numbers in news and advertising and try to check them for sanity.
Perhaps there were lots of studies, and 12 pounds is the average of the maximum weight loss measured in each study? Or perhaps the maximum average weight loss?
Not clear, is it?
Nigel (UK)
Hi David,
Your question reminds me of a person I worked with a few years ago. During design, we would need to provide the client with estimated construction costs. He would always refer to it as "probable estimated construction costs." We went back and forth on what this meant, so he decided to change it to "approximate probable estimated construction costs". I guess he wanted to make sure that no one could hold him to his estimate.
Bill
Smithville, NJ
Perhaps "on average" just meant typically, generally true ...
Lose on average up to 12 pounds = Lose up to 12 pounds (except, when it is not)
This gives more wiggle room to avoid false advertisement
Probably the average of those who lost weight.
(07-18-2019 03:43 PM)David Hayden Wrote: [ -> ]I saw a TV commercial the other day for some prescription drug that softens nose hair or whatever. For those outside the USA, yes, drug companies are allowed to advertise prescription drugs directly to consumers here. Anyway, the ad said that it's not a weight loss drug, but people taking it lost "on average up to 12 pounds."
My question for you is what on EARTH does "on average up to" mean? Keep in mind that companies must be able to back up their claims with clinical studies so this must be some actual statistic from the study. To me, "on average" is clearly a mean and "up to" is clearly a maximum, so "on average up to" sounds self-contradictory.
I have a theory, but I'm curious to read what my fellow nerds think.
And as an aside, I posted this question on facebook about a week ago. The comments are almost entirely about how companies advertise, the problems of advertising to consumers, etc. It appears that even my normally smart friends drew a blank on the statistic and really didn't care. Personally, I always look/listen for numbers in news and advertising and try to check them for sanity.
In the USA, it's only illegal to make false claims if you get caught.
Even if you do get caught, the Federal Trade Commission (FTC) will fine you a lot less (on average
) than you made by the false advertising so it pays to make those false claims.
This sounds suspiciously like ISPs advertising download speeds of "up to X Mb/s or more," which sounds good but means nothing...
(07-18-2019 04:16 PM)Nigel (UK) Wrote: [ -> ]Perhaps there were lots of studies, and 12 pounds is the average of the maximum weight loss measured in each study? Or perhaps the maximum average weight loss?
Not clear, is it?
Nigel (UK)
My thoughts were somewhat similar, but I was think of multiple weight measurements per person within one study rather than multiple studies. Perhaps they did one study but weighted each participant daily. Then "on average up to 12 pounds" is the average of the maximum weight loss for each individual.
As you said, not clear.
I tend to ignore claims based on descriptive statistics when the 'claim' omits any of the following: high, low, mean, median & mode.
BEST!
SlideRule
Of the ten percent of subjects who lost weight, the average weight loss was 12 pounds. Of course the other 90 percent gained an average of 20 pounds.
Perhaps 12 pounds was the upper end of a bin containing the median.