on-line dialect survey (B. Vaux)

[James A. Landau]

In a message dated 10/10/02 9:11:42 AM Eastern Daylight Time,
Dalecoye at AOL.COM writes:

> answer c and d are indentical....  Here's
>  the question
>
>  1. aunt
>       a. [] as in "ah" (9.94%)
>       b. [] as in "ant" (75.16%)
>       c. [] as in "caught" (2.24%)
>       d. I have the same vowel in "ah", "caught", and "aunt" (2.24%)
>       e. I pronounce it the same as "ain't" (0.64%)
>       f. I use [/] when referring to the general concept of an aunt, but []
>  when referring to a specific person by name. (7.21%)
>       g. I use [] when referring to the general concept of an aunt, but [/]
>  when referring to a specific person by name. (1.76%)
>       h. other (0.80%)
>      (624 respondents)

Also the arithmetic is suspicious.  Adding up all the percentages from a thru
h, inclduing both c and d, you get 99.99% (the last .01% is round-off error).
 That implies that c and d were tabulated separately and both were included
in the count.  Why then do c and d have exactly the same response (2.24%,
which is 14 people)?  Did 28 people answer EITHER c or d and were allocated
evenly to what turns out to be two identical answers?  Or did c and d each
just happen to have 14 respondents?

      - Jim Landau
        systems engineer
        FAA Technical Center (ACB-510/BCI)
        Atlantic City Int'l Airport NJ 08405 USA