Escolar Documentos
Profissional Documentos
Cultura Documentos
Very important
Somewhat important
Somewhat unimportant
Very unimportant
Don't know
Total
Total
44%
31%
13%
7%
4%
1980
Democrat
57%
27%
7%
5%
3%
792
Republican
29%
41%
19%
8%
3%
703
Independent
48%
22%
15%
12%
2%
302
Very important
Somewhat important
Somewhat unimportant
Very unimportant
Don't know
Total
41%
26%
16%
14%
4%
1980
56%
26%
10%
5%
2%
792
23%
24%
23%
26%
3%
703
39%
31%
17%
13%
1%
302
Very important
Somewhat important
Somewhat unimportant
Very unimportant
Don't know
Total
60%
29%
5%
2%
3%
1980
63%
30%
4%
2%
1%
792
60%
30%
6%
2%
2%
703
61%
31%
5%
2%
1%
302
Very important
Somewhat important
Somewhat unimportant
Very unimportant
Don't know
Total
43%
29%
15%
10%
4%
1980
58%
28%
7%
4%
2%
792
21%
34%
25%
17%
3%
703
49%
24%
17%
9%
1%
302
Very important
Somewhat important
Somewhat unimportant
Very unimportant
Don't know
Total
64%
25%
5%
3%
3%
1980
66%
25%
5%
3%
1%
792
66%
25%
5%
3%
2%
703
61%
30%
6%
1%
1%
302
Very important
Somewhat important
Somewhat unimportant
Very unimportant
Don't know
Total
61%
25%
6%
3%
4%
1980
71%
20%
3%
4%
2%
792
51%
34%
9%
4%
2%
703
63%
25%
8%
2%
1%
302
Very important
Somewhat important
Somewhat unimportant
Very unimportant
Don't know
Total
48%
21%
12%
15%
4%
1980
66%
19%
6%
6%
2%
792
29%
21%
19%
28%
2%
703
43%
24%
15%
16%
2%
302
Very important
Somewhat important
Somewhat unimportant
Very unimportant
Don't know
Total
29%
42%
17%
5%
7%
1980
31%
43%
18%
3%
5%
792
27%
47%
16%
6%
4%
703
28%
37%
20%
7%
8%
302
51%
60%
30%
57%
35%
30%
48%
26%
15%
9%
21%
17%
1056
417
388
159
Strongly agree
Somewhat agree
Somewhat disagree
Strongly disagree
Dont know
34%
25%
15%
12%
14%
25%
23%
16%
19%
17%
42%
27%
17%
6%
9%
44%
25%
8%
11%
11%
Total
1056
417
388
159
Strongly agree
Somewhat agree
Somewhat disagree
Strongly disagree
Dont know
26%
29%
16%
15%
13%
22%
32%
14%
17%
15%
29%
31%
19%
14%
8%
35%
26%
14%
15%
11%
Total
1056
417
388
159
Strongly agree
Somewhat agree
Somewhat disagree
Strongly disagree
Dont know
25%
28%
16%
16%
15%
22%
27%
16%
18%
17%
30%
30%
19%
13%
9%
26%
29%
10%
22%
12%
Total
1056
417
388
159
Strongly agree
Somewhat agree
Somewhat disagree
Strongly disagree
Dont know
24%
26%
18%
17%
14%
21%
22%
20%
25%
12%
31%
33%
17%
10%
9%
21%
29%
19%
11%
19%
Total
1056
417
388
159
Strongly agree
Somewhat agree
Somewhat disagree
Strongly disagree
Dont know
13%
21%
24%
29%
13%
14%
22%
24%
26%
14%
12%
20%
28%
33%
7%
13%
23%
21%
29%
14%
Total
1056
417
388
159
Strongly agree
Somewhat agree
Somewhat disagree
Strongly disagree
Dont know
15%
21%
23%
27%
14%
13%
20%
23%
30%
14%
18%
25%
22%
26%
9%
17%
19%
22%
25%
17%
Total
1056
417
388
159
Strongly agree
Somewhat agree
Somewhat disagree
Strongly disagree
Dont know
24%
23%
16%
10%
26%
17%
20%
17%
17%
29%
35%
28%
14%
6%
18%
21%
26%
24%
4%
25%
Total
1056
417
388
159
Strongly agree
Somewhat agree
Somewhat disagree
Strongly disagree
Dont know
24%
25%
18%
7%
25%
15%
27%
21%
10%
27%
38%
25%
15%
4%
18%
23%
24%
23%
4%
27%
Total
1056
417
388
159
Strongly agree
Somewhat agree
Somewhat disagree
Strongly disagree
Dont know
13%
20%
29%
21%
17%
14%
23%
27%
17%
19%
8%
21%
34%
27%
10%
12%
15%
34%
19%
20%
Total
1056
417
388
159
Strongly agree
Somewhat agree
Somewhat disagree
Strongly disagree
Dont know
7%
16%
28%
20%
28%
7%
15%
27%
20%
31%
5%
17%
36%
24%
18%
10%
18%
29%
13%
31%
Total
1056
417
388
159
Strongly agree
Somewhat agree
Somewhat disagree
Strongly disagree
Dont know
8%
15%
20%
24%
33%
7%
19%
20%
16%
38%
9%
9%
22%
38%
21%
9%
12%
24%
22%
33%
Total
1056
417
388
159
Strongly favor
Somewhat favor
Somewhat oppose
Strongly oppose
Total
38%
23%
23%
16%
1056
58%
21%
14%
8%
417
21%
24%
29%
27%
388
31%
20%
32%
17%
159
The calculation of credibility intervals assumes that Y has a binomial distribution conditioned on the
parameter \, i.e., Y|~Bin(n,), where n is the size of our sample. In this setting, Y counts the number
of yes, or 1, observed in the sample, so that the sample mean (y ) is a natural estimate of the true
population proportion . This model is often called the likelihood function, and it is a standard concept
in both the Bayesian and the Classical framework. The Bayesian 1 statistics combines both the prior
distribution and the likelihood function to create a posterior distribution. The posterior distribution
represents our opinion about which are the plausible values for adjusted after observing the sample
data. In reality, the posterior distribution is ones knowledge base updated using the latest survey
information. For the prior and likelihood functions specified here, the posterior distribution is also a
beta distribution ((/y)~(y+a,n-y+b)), but with updated hyper-parameters.
Our credibility interval for is based on this posterior distribution. As mentioned above, these intervals
represent our belief about which are the most plausible values for given our updated knowledge base.
There are different ways to calculate these intervals based on (/y). Since we want only one measure
of precision for all variables in the survey, analogous to what is done within the Classical framework, we
will compute the largest possible credibility interval for any observed sample. The worst case occurs
when we assume that a=1 and b=1 and y=n/2. Using a simple approximation of the posterior by the
normal distribution, the 95% credibility interval is given by, approximately:
For this poll, the Bayesian Credibility Interval was adjusted using standard weighting design effect
1+L=1.3 to account for complex weighting2
Examples of credibility intervals for different base sizes are below. Ipsos does not publish data for base
sizes (sample sizes) below 100.
Sample size
2,000
1,500
1,000
750
500
350
200
100
Credibility intervals
2.5
2.9
3.5
4.1
5.0
6.0
7.9
11.2