I'm somewhat uneasy with his conclusion based on such a small sample. Basically, 40 kids said they preferred the food with the McDonald's logo vs. 25 who did not, a difference of 13 kids. So it would appear that we have a conclusion based on 13 kids.
Some other things said in the story also make me uneasy. The story said "obesity prevention expert Dr. Thomas Robinson" and "He said the study supports efforts to ban or regulate advertising or marketing...". I'm sorry, but that makes me wonder if they set out a pre-conceived conclusion, and set out to obtain something that backed it up- starting out with an "ax to grind" in a sense.
The story said "The food -- taken from the same order -- was wrapped in either McDonald's packaging or unbranded packages in the same color and style.". It would seem to to me that if the idea were to show a difference with a McDonald's logo, it may have been a better idea to give a choice between something with a McDonald's logo vs. something with a made up logo. It would seem to me that this might eliminate the possibility that the preference was for packaging that had more decoration vs. packaging that was "plain", looking at a company logo as decoration. See why I said a made up logo? A comparison between those two decorated choices rather than McDonald's logo decoration vs. plain no logo decoration would make it a contest (however limited) between an existing logo vs. a logo made up for the study. In this case, a significant difference might better show influence of the logo that actually exists in the market and results favoring the made up logo would seem to show influence of just more decoration.
Chowhound
Comic Vine
GameFAQs
GameSpot
Giant Bomb
TechRepublic