(Who would have ticked "yes" to "do you feed your child dessert?" on that survey, because we finished most meals with a bowl of pureed fruit...)

The survey asked for the specific food items the child consumed on the target day, and then the researchers categorized the food by type. It's the only way you could do it, really, because otherwise you have people giving their kids Kool-Aid and ticking off that they gave "juice," or classifying strawberry ice cream as fruit, or whatever.