WEIGHTING PROCEDURES FOR 2010 MISSISSIPPI POLL:
|
Education (for
age 25 & older): |
Latest Census |
WEIGHT1 (adults/ phone numbers) |
WEIGHT2 (wgt1 * age correction) |
WEIGHT3 (wgt2 * sex correction) |
WEIGHT4 (wgt3 * education correction) |
WEIGHT5 (wgt4 * race correction) |
|
11 years or less |
21.2% |
10.9% |
11.7 |
10.9 |
21.3 |
21.6 |
|
12th grade |
31.3% |
32.3 |
31.1 |
31.0 |
31.3 |
30.6 |
|
Some college |
28.5% |
26.1 |
26.0 |
24.5 |
28.4 |
28.4 |
|
College grad & some graduate school |
19.0% |
30.7 |
31.2 |
33.6 |
19.0 |
19.4 |
|
Sex (for age 18 & older) |
|
|
|
|
|
|
|
Male |
47.5% |
30.0 |
30.0 |
47.5 |
45.6 |
46.1 |
|
Female |
52.5% |
70.0 |
70.0 |
52.5 |
54.4 |
53.9 |
|
Age (for age 18 & older) |
|
|
|
|
|
|
|
18-29 |
23.5 |
5.9 |
23.5 |
23.9 |
26.4 |
26.0 |
|
30-44 |
26.0 |
19.3 |
26.0 |
25.0 |
22.6 |
22.0 |
|
45-59 |
27.0 |
33.3 |
27.0 |
27.8 |
27.7 |
28.0 |
|
60 + |
23.5 |
41.5 |
23.5 |
23.3 |
23.3 |
24.0 |
|
Race (for age 18 & older) (2000 census) |
|
|
|
|
|
|
|
White |
64.8% |
68.8 |
62.0 |
64.5 |
60.3 |
64.8 |
|
Black |
33.1% |
26.5 |
33.6 |
31.2 |
35.8 |
33.1 |
|
Other |
2.1% |
4.7 |
4.4 |
4.3 |
3.9 |
2.1 |
WEIGHT is the final weight variable, which should be used to weight the dataset. It merely multiplies
WEIGHT5 by a constant in order to ensure that the weighted sample size has the same number of
cases as the unweighted sample size.
Special note on the most recent Mississippi Polls. Notice how the increased prevalence of cell phones
without landlines has reduced the numbers of young adults interviewed, so that those under 30 have to
receive a weight of nearly 4 times in the second weighting. Also notice how women today now comprise
70% of the unweighted sample, so that men must be overweighted for the third weighting. As always,
high school dropouts are undersampled, so that they must be nearly double weighted in the fourth
weighting. Finally, comparatively speaking, racial minorities are relatively well sampled, necessitating
only a slight correction in the fifth weighting.