Circular Knitting Gauge

I've noticed that when I knit socks using a meticulously-knitted test swatch and exacting calculations, they tend to come out much larger than they should.  According to my favorite little book, "The Knitting Answer Book," by Margaret Radcliffe,

this is because I have a different gauge in knitting vs. purling.  I was actually aware that I purl much more tightly than I knit, but I didn't do the mental math on the effect it would have on my circular gauge until I had to re-do a pair of socks for my hubby 3 times before getting it right. 

I've changed to doing test swatches in seed stitch so as to even it out, and that gets pretty close, generally within a quarter of a stitch per inch.  Of course, if the intended recipient is handy, I've just been making socks toe-up and trying them on, with the increases stopping once the right circumference is reached.  Unfortunately, the intended recipient is not always handy, so Ms. Radcliffe suggests an alternative method: sliding the stitches back down the double-pointed needle and just wrapping the yarn loosely behind the swatch, thus knitting all rows without creating a garter stitch.  I'll have to try that. 

I highly recommend this book, by the way.  There are questions I've had about knitting that I could not get answered anywhere else, including what turned out to be different knitting styles, but that's a matter for another post.