However, you are getting the margin of error wrong. Indeed, the margin given in the press release (+/-3%) is for percentages close to 50%. For such a small percentage (5%), the margin of error is smaller. (This is a very common mistake.)

In fact, the number of adults that “never eat meat, fish, seafood, or poultry” is 5% +/- 1.4%.

You can use the following formula:

M = 2 * SQRT((p * (1-p))/N)

where M is the margin, p the percentage (in the form 0.05), N the sample size. (Do it for 50% and you’ll get 3%.)

So it really looks like there is a trend. (Even if computing the margin of error for this second assertion would require having some info on the 1994 poll.)

]]>Matt,

That’s an interesting idea. I’d be surprised if a higher percentage of immigrants to the U.S. are vegetarian than are non-immigrants. To my knowledge, India would be the only country, from which the U.S. receives much immigration, that has a higher percentage of people who never eat meat. But, it’s definitely not my area of expertise. Come to think of it, I wonder if immigrants are represented much in these polls, and if not, how that effects the accuracy. It looks like since 1990 there have been about 20 million immigrants into the U.S., which is 6.4% of the population so I’m not sure immigrants could have much of an effect on the sampling.

]]>