In analyzing a survey, I get the crosstab shown below. How do I go about
computing a margin of error for this survey? Do I compute it for the grand
total, for each cell individually individually and take the highest? I'm
assuming that what every p and n I use, the margin of error is
p +/- 1.96 * sqrt(pq/n)
Number Undecided Yes Total
Group A 13 18 54 85
15.3% 21.2% 63.5% 100%
Group B 1 6 21 28
3.6% 21.4% 75.0% 100%
Group C 4 1 12 17
23.5% 5.9% 70.6% 100%
Group D 5 2 5 12
41.7% 16.7% 41.7% 100%
Total 23 27 92 142
16.2% 19.0% 64.8% 100%
Ronny Richardson
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
