NCVS

The Wikipedia page lists six stages of the Northern Cities Vowel Shift, with the following results:

1. cat: [k{t] becomes [ke@t] or [kI@t]
2. cot: [kAt] becomes [kat]
3. caught: [kOt] or [kQt] becomes [kAt]
4. ket: [kEt] becomes [k@t]
5. cut: [kVt] becomes [kOt]

Also the vowel in 'kit' is backed. I don't think I've ever heard anyone with all six stages. Which ones occur in what areas? Are they more advanced in younger speakers? What about social status?

I'm from Michigan and I have the first three stages, i.e.

"cat" [ke@t]
"cot" [kat]
"caught" [kAt]

I ordered a sandwich at a cafeteria in Buffalo, NY a few days ago. The lady behind the counter asked me [anw@t]. I was sure the question had been "And what?" and I replied "Just the sandwich, please". She asked again a couple of times before I realized she was saying "On what?".

Note that the NCVS actually varies from dialect to dialect with it, and has not necessarily taken the form that Wikipedia says it takes. For instance, in my dialect I have:

"cat" : [k_hE{?], [k_he{?], or sporadically [k_hE@?]
"cot" : [k_ha?]
"caught" : [k_hQ?], or sporadically [k_hA?]
"ket" : [k_hE_"?]
"cut" : [k_hV?] (note that this is not rounded at all *but* it is fully backed, unlike what is often transcribed as [V] for NAE dialects), or in more careful speech [k_h7?] (also note that this is fully backed)
"kit" : [k_hI_"?]

You could say that I have all the stages, but most of them are not complete degree-wise. Of all the changes, the one for me which is the most complete is [A] -> [a], which only does not occur when r-coloring (sporadically) or l-coloring is present (in any register) or occasionally in very careful speech. The raising and diphthongization of [{] is pervasive, but in more careful speech it can be quite weak. Also, the centralization of [E] and [I] are pervasive, but they are not complete in the sense of shifting [E] to [3] or [@].

<<cut" : [k_hV?] (note that this is not rounded at all *but* it is fully backed, unlike what is often transcribed as [V] for NAE dialects)>>

Yes, I've noticed that the /V/ in 'cut' for example, is for me much closer to [6] or [@] than [V]. I think this is the most common pronunciation outside of the Carribean and Scotland (perhaps Northern England as well). I really have no distinction between /V/ and /@/; it's just a matter of stress.