Study: Food in McDonald’s Wrapper Tastes Better to Kids
August 7th, 2007Via: CNN:
Anything made by McDonald’s tastes better, preschoolers said in a study that powerfully demonstrates how advertising can trick the taste buds of young children.
In comparing identical McDonald’s foods in name-brand and plain wrappers, the unmarked foods always lost.
Even carrots, milk and apple juice tasted better to the kids when they were wrapped in the familiar packaging of the Golden Arches.
The study had youngsters sample identical McDonald’s foods in name-brand and unmarked wrappers. The unmarked foods always lost the taste test.
“You see a McDonald’s label and kids start salivating,” said Diane Levin, a childhood development specialist who campaigns against advertising to kids. She had no role in the research.
Levin said it was “the first study I know of that has shown so simply and clearly what’s going on with (marketing to) young children.”
Study author Dr. Tom Robinson said the kids’ perception of taste was “physically altered by the branding.” The Stanford University researcher said it was remarkable how children so young were already so influenced by advertising.
Related: Fast Food ‘As Addictive As Heroin’
It’s like the bottled water scam. Aquafina being pressured to admit the source is Ayre Massachusetts. This water must be good for me because its wrapped in a plastic bottle.
Advertising is an isidious barf weapon. Too bad all of that super size me crap is so appealling to the kids. Next thing you know there will be a scandal re McDonald’s similar to the ongoing Catholic church thang. Poor kids. I don’t know what kind of brainwashing I was on the receiving end of but shiite! It wasn’t about food preferences fer crying out loud. For what its worth I think kids “these days” should and will need to know how to grow their own and studies should be pointing out whether kids and/or adults for that matter have a freaking clue as to how to feed themselves without a wrapper around it.
Oh sister P says the quarter pounder goes for $18 in Iceland.
Who the F*ck would pay that kind of money for ANYTHING from McDonalds??
So what we have here is a study with a small sample size (63 kids, lower-income only), no controls and bad methodology.
McDonald’s food has lots of fat and sugar, things that kids are wired to crave. They associate the flavor of the food with the brand.
This effect has been well established:
http://www.sciam.com/article.cfm?articleID=000B582E-E271-13E3-A27183414B7F0000
This has nothing to do with advertising. Thanks, CNN, for another lovely load of bullshit.
“The Biology of Belief” – Bruce Lipton
There is a self reinforcing expectation at play here. Advertising has much to do with it. The first layer in the program is the science of flavour enhancement. The image of McDonalds, its logo, its colours are deeply associated with the smell and flavour that has been developed by the company. The brand which is McDonalds carries its smell with it, to see the wrapper within the subconscious is to insinuate the smell and flavour of McDonalds. The smell and flavour is uniquely McDonalds, there is no mistaking it, this particular smell and flavour has been developed to attract and cause addiction. The wrapper alone, the logo alone, insinuates a smell and flavour that signals the response of addiction.
@remrof: n=63 is not necessarily inadequate for any individual comparison (e.g. for a 77% probability the standard error for the estimator is is +/- .05%, so that the confidence interval would lie conclusively beyond 50%) although for mid-range probabilities (54%) it would be inadequate. The only issue here is multiple comparisons, but with strong results that are obviously correlated, that might not be a problem in this particular study. I don’t know what you mean by having “no controls”: the plain-wrapper item constitutes the control. As for the population consisting of only low-income kids, I’m assuming that’s the target population, and obviously any nutritional epidemiologist would not generalize beyond that. In short, I don’t see anything obviously wrong with this study design (although there could of course be more subtle flaws).
Re: the article you linked to, yes expectations are related to taste, but who sets up those expectations?
I’ll try again.
This is a study without a control group! The effect of advertising on kid’s food preference (the hypothesis it purports to test) has no control, because there is no measure of how much advertising exposure these children have had.
In order to determine whether it was the flavor of the food or the ads that influenced the kid’s perception of the taste, you would need a control group: kids who haven’t been exposed to the advertising, but who have been exposed to the food.
If the study wants to measure the effect of advertising on the kids, then the independent variable should be the ads, not the wrapper.
The methodology itself–feeding food perceived to be from McDonald’s to kids in order to gauge their susceptibility to advertising–is bad.
The question is not who sets up the expectations, but what: the ads, or the flavor?
Did anyone notice what the food item perception most influenced by manipulation of the brand logo was? The french fries, easily the greasest, most salty food that McDonald’s sells.
There is so much confirmation bias at work here it boggles the mind.
http://www.foxnews.com/story/0,2933,292301,00.html
It’s about the kids.
No.
This has nothing to do with the flavour of the food.
All the samples were the same, but in two types of packaging.
The research was testing one variable only: logo versus no logo. Same food, different wrappers. Simple. Valid.
Are the kids’ responses to a change in this variable alone different in a statistically significant way? Yes.
The study was measuring the effect of a logo, not advertising. People looking at the results can infer wahtever they like about the effect of advertising versus flavour, but that doesn’t invalidate this research or its methodology.
No, that’s not the question being addressed here at all. You are just trying to change the subject. The simple fact is these kids’ taste perceptions are influenced by the presence of a logo.
Que? There is no “conformation bias”, just logo versus no logo. You seem rather confused.
I’ll agree that this study alone cannot prove that there is a causal connection between advertising and flavor perception. The study was not even designed to test an association between the two (though it suggests a followup study that does exactly that).
Like all scientific studies, this one was designed to test a single testable hypothesis: low-income children will prefer food wrapped in a McDonald’s logo to identical food that is in a plain wrapper. There are no obvious flaws with the study design in answering that single hypothesis. This doesn’t prevent overinterpretation of the results, but that is an entirely separate issue.
I agree with tochigi: you’re either confused about how science works or intentionally changing the subject.
Kevin, please forgive the long post, but I think this will clear things up. Last one, promise.
http://en.wikipedia.org/wiki/Hypothetico-deductive_method
This is the sort of hypothesis I had in mind. Here it is applied to the study at hand:
1. Kids like McDonald’s, I guess
2. Kids like McDonald’s because of ads
3. As slomo beautifully put it: “Low-income children will prefer food wrapped in a McDonald’s logo to identical food that is in a plain wrapper.” You know the experiment.
4. There was a corroboration, which I contend is irrelevant because it was an inappropriate experiment.
The hypothesis that motivated the experiment (2.) and the hypothesis (experimental prediction) that they probably put at the top of the study paper (3.) are different things. I disagreed with the experimental prediction that was deduced (3.). That’s why I was so critical of the experimental design in my second post: because I though it was inappropriate to the larger hypothesis (2.).
So when I said:
remrof:
I meant that it had a small sample size, no controls and bad methodology for an experiment designed to test 2.
You understood me to mean it had a small sample size, no controls and bad methodology for an experiment designed to test 3. As you said, it doesn’t, of course!
Which is what (understandably) motivated these comments:
tochigi:
slomo:
That’s all true, but it’s a little obvious isn’t it? Come on guys, gimme a little credit!
Phew! Glad we got that worked out.
@tochigi
By “confirmation bias” I mean confirmation bias in interpretation, i.e. the conclusion drawn and the CNN article.
@slomo
By “small sample size,” I mean that the sample size isn’t big enough to justify generalizing the results to a larger population like the article did, not that the results aren’t statistically valid (i.e. sufficiently non-random). I think ~30 samples is the cutoff for that?
So you see folks, this was all an exercise in accidental equivocation. No hard feelings, eh?
“By “small sample size,†I mean that the sample size isn’t big enough to justify generalizing the results to a larger population like the article did, not that the results aren’t statistically valid (i.e. sufficiently non-random). I think ~30 samples is the cutoff for that?”
Actually, being in the business of designing and analyzing data from studies (albeit not nutritional studies) I have never been a fan of rubrics such as “d samples is the cutoff” where d always varies according to which introductory textbook you choose. n=10 could be just fine if the samples were truly random and the standard error is small relative to the required sensitivity. Of course, if the investigators just grabbed a convenience sample, then there could be all kinds of issues related to the conditions under which the subjects were selected. But this is not a sample size issue at all.
I don’t think the study is poorly designed. Overinterpreted perhaps, but not poorly designed. But going back to my first comment, here’s the issue: if all it takes is a label to influence blind taste preferences, then what is the true psychological mechanism? It could perhaps just be prior association, but it also could very well be advertising. A study designed to test the latter would necessarily be an observational study (due to feasibility and IRB issues), so a “control group” would not even be possible (if by “control” you mean the rigorous definition used in clinical trials). The best you can hope for is a low-exposure group, assessed by questionnaire. The latter type of study would have all kinds of confounding issues, so in fact the original study is more convincing to me.