2021 NSCH Table of State Sample Sizes

Appendix C

2021 National Survey of Children’s Health
Table of State Sample Sizes

Appendix C
2021 NSCH - Estimated State Sample Sizes

Base Sample
Oversampling
Rate 1
'k'
State Sample Sizes
Alabama
4
Alaska
4
Arizona
3
Arkansas
4
California
3
Colorado
3
Connecticut
3
Delaware
4
District of Columbia
6
Florida
3
Georgia
3
Hawaii
2
Idaho
3
Illinois
3
Indiana
3
Iowa
6
Kansas
3
Kentucky
4
Louisiana
3
Maine
5
Maryland
3
Massachusetts
4
Michigan
4
Minnesota
4
Mississippi
4
Missouri
4
Montana
5
Nebraska
5
Nevada
3
New Hampshire
5
New Jersey
3
New Mexico
4
New York
3
North Carolina
3
North Dakota
6
Ohio
3
Oklahoma
4
Oregon
4
Pennsylvania
4
Rhode Island
3
South Carolina
4
South Dakota
5
Tennessee
3
Texas
3
Utah
3
Vermont
6
Virginia
3
Washington
3
West Virginia
5
Wisconsin
4
Wyoming
5
Total
Proportion by Strata
1

Proportion of
Households with
Children from
2019 ACS
'P'
29%
26%
34%
30%
43%
38%
40%
35%
21%
32%
37%
33%
36%
38%
37%
25%
38%
31%
33%
25%
41%
37%
38%
39%
30%
33%
22%
29%
34%
29%
42%
23%
33%
35%
22%
37%
28%
34%
36%
36%
32%
25%
36%
38%
47%
22%
39%
39%
22%
38%
23%

Base
Sample
Sampling
Variance1
‘R’
1.05
1.06
0.99
1.05
1.02
0.98
0.98
1.04
1.03
0.98
1.01
0.99
0.99
0.99
0.98
1.03
0.98
1.03
0.99
0.99
0.99
1.04
1.04
1.03
1.05
1.03
1.01
1.03
0.99
1.01
1.01
1.02
0.98
0.98
1.01
0.97
1.03
1.03
1.03
0.98
1.04
1.02
0.98
1.02
1.01
1.01
0.98
0.98
1.04
1.04
1.04

Total
Sample 2
6200
8100
6000
6700
4900
10000
4500
5600
5000
6200
10500
7200
4300
4900
5100
4000
4600
5300
9000
5300
4400
3700
4300
3100
7400
4500
5600
5800
6300
4200
4500
7400
5800
5300
4500
9900
6300
15500
4100
5000
5700
4800
5200
6600
3800
4400
4300
4300
6100
8000
5800
300,000

Stratum 1
Stratum 2a
Sample
Sample
(Admin Flag
(Admin Flag
for HHLD
for HHLD w/o
w/Children)
Children)
4600
1700
4500
3600
4000
2000
4600
2100
3700
1200
6300
3900
3200
1300
4000
1500
3600
1400
4200
2000
6900
3400
3500
3700
2800
1500
3500
1500
3600
1600
3000
1000
3500
1200
3800
1500
5900
3000
3700
1500
3200
1200
2800
900
3400
850
2400
700
5500
1900
3400
1100
3600
2000
4200
1700
4300
1900
3100
1100
3200
1300
4600
2700
3900
2000
3800
1500
3200
1300
6500
3300
4400
1900
10500
5000
3200
900
3600
1400
4300
1400
3300
1500
3900
1400
4700
1900
2900
900
3000
1400
3100
1200
3000
1300
4000
2100
5900
2200
3800
2000
207,000
93,000
69%
31%

These are the Stratum 1 oversampling rates and the sampling variances for the base production sample only. Colorado, Nebraska,
Oregon, Wisconsin, Georgia, Louisiana, and Ohio all have additional sample that may have different oversampling rates and
sampling variances.
2 Sample sizes include both the base production sample as well as the additional nationwide oversample of young children and state
oversamples in CO, NE, OR, WI, GA, LA, and OH. All sample sizes have been rounded.

Appendix C
Calculations for Base Production State Sample Sizes
The oversampling rate (k) was calculated to maximize the sample from Stratum 1 without increasing the
variance (R) too much beyond that of a proportional stratified design of a similar cost.
(P) is the proportion of households with children based on the 2019 ACS audit 3.
The portion of the sample coming from each stratum was calculated using the oversampling rate (k) and
the proportion of all households in each of the sampling strata (W1 and W2a). Address valid rates, child
prevalence rates, and Screener and Topical returns rates were estimated by state and stratum primarily
using previous years’ response outcomes. A national total sample size of approximately 300,000 was
determined based on budget. This includes the base production sample for all states as well as a national
oversample aimed at getting additional interviews from children aged zero to five and additional sample
in Colorado, Nebraska, Oregon, Wisconsin, Georgia, Louisiana, and Ohio. We expect the base
production sample will produce approximately 600 Topical interviews per state. The expected number of
interviews along with previous years’ response information was used to calculate stratum sample sizes
(n1 and n2a) in each state.
For example, in Alabama 4:
The oversampling rate (k) was recalculated to be 4.0, for an R = 1.05. Of the 2019 ACS households
eligible for NSCH sampling, 20.8% would be put in Stratum 1 (W1) and 53.5% would be put in Stratum
2a (W2a).
The portion of the sample coming from Stratum 1 was estimated to be (k*W1) / (W2a + k*W1) = 60.9%
and the portion of the sample from Stratum 2a was 39.1%.
We can expect a completed Topical from 91.6% (Valid) * 37.3% (Screener) * 78.7% (Children) * 78.7%
(Topical) = 21.2% of addresses in Stratum 1 and 76.4% (Valid) * 43.1% (Screener) * 10.6% (Children) *
82.7% (Topical) = 2.9% of addresses in Stratum 2a.
Since we expect to get 600 Topicals per state, we can get the state sample size (n) by evaluating the
following:
(n * Portion of the sample from Stratum 1 * Stratum 1 completed Topical rate) + (n * Portion of the
sample from Stratum 2a * Stratum 2a completed Topical rate) = 600.
Within rounding, it then follows that n = 600 / (60.9% * 21.2% + 39.1% * 2.9%) = 4300. Now n1 = 4300
* 60.9% = 2600 and n2a = 4300 * 39.1% = 1700. These are the base production sample sizes in each stratum
for Alabama. 5

3

Before mailout, this will be updated with the 2020 ACS.
Note that numbers have been rounded here for all calculations.
5
These numbers do not match what is in the table since this shows only the calculation for Alabama’s base production
sample while the sample sizes in the table include the additional nationwide oversample of young children.
4