Interesting example of the huge gap between what people claim to do and what folk actually do. Of course, the difference is not always so obvious or so large as it is here but some kind of difference there is likely to be.
Note also how embarrassing it can be when you take market research literally and assume that people know what they do, how they do it and that they can tell market researchers/therapists/themselves so.
Think about it next time your considering a survey or the data it generates.
Free gift: demonstration that the web isn't really built on cats but mice