|
This appendix provides a brief summary
of the methodology of the study including
the sample design and the statistical
techniques used in summarizing the data.
It also includes a discussion of sampling
errors, provides the standard errors for
key variables in the study and presents
a simplified methodology for estimating
standard errors. Much of the material
included here has been abstracted from
the technical report provided by the Research
Triangle Institute (RTI), the contractor
who carried out the sampling for and conducted
the seventh National Sample Survey of
Registered Nurses discussed in this report.
The basic sample design used in all seven
cycles of the National Sample Survey of
Registered Nurses is basically the same.
The NSSRN 2000, the seventh in the series,
oversampled minority nurses in order to
allow for more in-depth analysis of this
special population of RNs. Several options
for oversampling were considered. The
State boards of nursing were asked to
provide information on the race/ethnic
background of RNs in order to facilitate
the oversampling. However, this information
was not available on many of the States’
files. Two States, Texas and North Carolina,
did provide race information which was
used to oversample minorities. For States
that were not able to provide race/ethnic
information for nurses on the list, minority
population distribution and minority nurses
distribution by State from the 1996 study
were used to assign larger samples of
nurses to States with both high proportions
of minority populations and high proportions
of minority nurses. This increased both
the number of minority and non-minority
nurses in the sample for those states
relative to the sample sizes for 1996.
The basic design was enhanced by using
sample design optimization methodology
and software developed by Chromy¹
to determine the sample allocation to
the lists that would simultaneously satisfy
variance constraints defined by the 51
States, the minority race groups and the
total US.
Sample Design
The seven surveys carried out to date
all followed the same design developed
by Westat, Inc. under a contract with
the Division of Nursing, BHPr, HRSA in
1975-76. The design approach took into
account two key characteristics of the
sampling frame. First, no single list
of all individuals with licenses to practice
as registered nurses in the United States
exists although lists of those who have
licenses in any one State are available.
Second, a nurse may be licensed in more
than one State.
¹ Chromy, James
R. “Design Optimization with Multiple
Objectives”. American Statistical
Association of the Section on Survey Research
Methods, Arlington, VA., pp A4-199
A sampling frame was required to select
a probability sample of nurses from which
valid inferences could be made to the
target population of all those with current
licenses to practice in the United States.
State Boards of Nursing in the 50 States
and in the District of Columbia (hereafter
also referred to as a State) provided
files containing the name, address, and
license number of every RN currently licensed
in that State. The States were also asked
to provide the race/ethnicity of each
nurse. Texas and North Carolina provided
files containing usable race/ethnicity
data for 5 groups. For sample allocation
and selection, these race/ethnicity groups
in Texas and North Carolina were collapsed
into White and nonWhite. Thus, 53 separate
lists were used: a White and NonWhite
file for Texas and North Carolina and
a separate file for each of the remaining
49 States and the District of Columbia.
These 53 lists constituted a multiple
sampling frame containing all the RNs
licensed in the US. Because many nurses
are licensed in more than one State, their
names could appear in the combined list
more than once. A nested alpha-segment
design was used to properly determine
selection probabilities for nurses appearing
in more than one of the 53 lists.
The target population of this study was
the current RN population of the United
States as of March 2000. RNs were selected
with equal probabilities within States.
Whether RNs fell into the sample depended
on whether their names fell within one
of the alpha-segments or portions of alpha-segments
that were selected for the sample. Approximately
equal-sized alpha-segments were constructed
by partitioning an alphabetically ordered
list of all RN names nationwide into 250
segments with equal (or as nearly equal
as possible) numbers of RNs. An alpha-segment
consisted of all alphabetically adjacent
names falling between set boundaries.
Both national and State-level estimates
were required. While uniform-sampling
rates would have produced the best national
estimates, the resulting sample sizes
for the smallest States would have been
inadequate to support State-level estimates.
Sampling rates were increased in the smaller
States to obtain larger State-level sample
sizes. Planned sampling rates ranged from
less than 1 percent in several of the
States with a large RN population to 15
percent in Wyoming. Planned State sizes
ranged from a sample of over 4,320 RNs
in California to approximately 625 in
Nebraska. While this disproportionate
sampling improved the precision of estimates
in the smaller States, it also reduced
precision of national estimates due to
unequal weighting effects.
Registered nurses were in the sample
on the basis of name, with an RN being
included in the sample if the name of
licensure fell within a specific portion
of the alpha-segments included in the
sample from the RN’s State of licensure.
As stated earlier, an alpha-segment consisted
of all alphabetically adjacent names falling
between set boundaries. The segments were
constructed so that each segment contained
approximately the same number of RNs.
Specifically, the lower boundary of an
alpha-segment was the last name in alphabetical
order of all the names included in that
segment. The membership of the segment
consisted of all names, beginning with
the lower boundary, up to but not including
a name that defined the upper boundary.
The latter name fell into the next alpha-segment.
A planned variation in the size of the
portions of segments was used to accommodate
the differing State sampling rates. The
largest portions used were full alpha-segments
while other sizes were 1/2-, 1/4-, 1/8-,
1/16-, and 1/32-portions. The fractions
indicated the size of the specified alpha-segment
portion relative to the size of the basic
alpha-segment. The sampling rate required
for a given State was achieved using a
combination of these portions of alpha-segments.
From the frame of 250 alpha-segments,
40 alpha-segments were randomly selected.
Although each State had 40 sample segments
(i.e., portions of alpha-segments), the
segments differed in size depending on
the State’s sampling rate. To identify
and account for nurses having multiple
licenses, the alpha-segment portions from
larger States were “nested”
or included, within those from smaller
States. Under this scheme, an RN who was
licensed under the same name in two States
with identical sampling rates was selected
(or not selected) for both States because
the alpha-segments and portions of alpha-segments
that defined sample membership were identical
for both States. However, for two States
that were sampled at different rates,
the alpha-segment portions for the lower
sampling rate (the State with a larger
RN population) were nested within those
of the higher sampling rate (the State
with the smaller RN population). The nested
alpha-segment design permitted the use
of each sample RNs data for State estimates
of each of her/his States of licensure
and also provided appropriate (multiplicity-adjusted)
weights for both State and national estimates.
The nesting was based on how the 40 basic
alpha-segment selections were used to
define the sample for each State. Each
of these alpha-segments, or one of the
fractional portions of it, constituted
one of the 40 sample clusters for each
State. Accordingly, each of the basic
alpha-segments had associated with it
a 1/2-portion selection and 1/4-portion,
1/8-portion, 1/16-portion, and 1/32-portion
selections.
The sampling rate for a particular State
was obtained from some combination of
the alpha-segments and portions. For example,
the 40 complete alpha-segments would have
constituted the sample for States with
a 16 percent sampling rate. Because each
segment contained an expected 0.4 percent
of the State’s RN names, taken together
they contained an expected 40 x 0.4 percent,
or 16 percent, of those names.) The sample
for a State with an 8 percent sampling
rate consisted of the 40 2-portion selections.
A 5 percent sampling rate was achieved
by first randomly dividing the 40 alpha-segments
into two groups, the first containing
30 alpha-segments and the other containing
10, and by using the 1/4-portions from
the first group and 2-portions from the
second group (0.4 x [(30x1/4) + (10x1/2)]
= 5).
The survey design specified precisely
which alpha-segments and portions would
correspond to each of the different sampling
rates used. This design resulted in the
specification of 40 pairs of names for
each of the sampling rates. Each pair
consisted of the names defining the lower
and upper boundaries for one of the alpha-segments
or alpha-segment portions corresponding
to the sampling rate. Thus, the alpha-segment
(portion) was defined by all names from
its lower boundary up to but not including
its upper boundary.
To ensure that current information about
RNs could be obtained, the survey design
called for periodic implementation. A
panel structure for the RN survey allowed
for several of the sample alpha-segments
in the periodic surveys to be systematically
replaced. Under the original survey design,
the 40 sample alpha-segments were randomly
assigned to five panels of eight alpha-segments
each. For each successive survey, a new
panel (consisting of eight new alpha-segments)
was entered into the sample, replacing
one of the five panels that was in the
previous survey. Under this scheme, a
nurse who maintained an active license
in the same State(s) and whose name did
not change could be retained in the sample
for up to five surveys. With the reconstruction
of the alpha-segments in the fourth RN
survey (1988), changes were made so that
exact correspondence of the current segments
to those established initially may no
longer exist; therefore, some nurses may
not have been carried through all five
surveys.
Each of the 51 State Boards of Nursing
provided one or more files that contained
the names of currently licensed RNs. These
files formed the basis of the sampling
frame from which the RNs for each State
were selected. The licensure files provided
by the States were submitted on computer
tape, on diskettes, or on a printed list.
Essentially the same procedure was followed
for sample selection for all States regardless
of which form was submitted. For this
current study, States were also asked
to identify those for whom the State provided
advanced practice nurse (APN) status.
In some cases, these APNs were identified
on separate lists and their APN status
was added to the information on the RN
sampling frame list. In other cases, the
State identified these nurses on the basic
list provided. Once a State provided a
licensure file containing all appropriate
names of individuals with active RN licenses
and meeting all specifications, the required
sample names in that file were selected.
Regardless of the way a State alphabetized
and standardized the names in its files,
the sample names were selected according
to the standards established by the survey
design. That is, sample selections ignored
blanks and punctuation in the last names
(except a dash in hyphenated names) and
ignored titles (e.g.,”sister”).
Whether or not the RN was an APN was not
taken into account in the sample selection.
Table B-1 shows the sampling rates and
sample sizes that were planned and actually
obtained for the 53 State and State by
race lists in the survey. Differences
between planned and actual sampling rates
result from State-specific variation in
the distribution of nurses’ names.
States are priority ordered by sampling
rate size.
The original State frame sizes were adjusted
to account for duplicate licenses within
States and ineligible licenses (i.e.,
frame errors) found in the sample. Duplicates
within States arose primarily from combining
RN and APN lists. Most duplicates were
identified before selecting the sample
and determining the frame size, but a
few were identified after sample selection,
requiring a frame size adjustment. The
ineligible licenses were identified in
the process of reconciling the State and
nurse reported licenses. Cases that could
not be reconciled by RTI were sent to
the State Boards of Nursing for resolution.
No changes in the sampling rates occurred
as a result of the frame adjustments,
so the nesting of the alphabetic clusters
remained the same even though the ordering
by adjusted frame would have changed.
It was, therefore, not necessary to change
the priority ordering as a result of any
changes in relative size.
Weighting Procedures
The probability sample design of the
survey permits the computation of unbiased
estimates of characteristics of the target
population. These estimates are based
on weights that reflect the complex design
and compensate for the potential risk
of nonresponse bias to the extent feasible.
The weights that are assigned to each
sample nurse may be interpreted as the
number of nurses in the target population
that the sample nurse represents. The
weight for an RN is the reciprocal of
the nurse’s probability of selection
in her/his priority State, adjusted to
account for nonresponse.
The weights were computed sequentially
for States A, B, etc., where A was the
highest-priority State, and B the next-highest-priority
State. The weight for State A was the
ratio of the count of licenses in the
sampling frame for State A to the number
of respondents licensed in State A. For
State B, and the remaining States, the
numerator and denominator of this ratio
were adjusted to account for State A and
other higher-priority States. To describe
the basic method, the following terms
are defined:
N(i)
= total number of licenses
for State I
m(i) =
number of respondents for
State I that did not have a license in
a higher-priority State
n(i,j)
= number of respondents
with a license in both State i and State
j [note n(i,i) denotes the
number of eligible respondents with a
license only in State i]
W(i) =
the adjusted weight for eligible respondents
who were assigned to the priority State I.
The weight for State A was computed as
follows:
W(A)
= N(A) / m(A).
For the State B weight, W(B), the numerator
was the total frame count of RNs licensed
in State B, N(B), after removing
the estimated total count of State B nurses
who were also licensed in State A (i.e.,
W(A) n(A,B)). Similarly, the numerator
of W(C) excluded State C nurses who were
also licensed in either State A or State
B (i.e., W(A) n(A,C) + W(B) n(B,C)). That
is, for the State B weight and the State
C weight, the computations were:
W(B) = [N(B) - W(A)
n(A,B)] / m(B)
W(C) = [N(C) - W(A)
n(A,C) - W(B) n(B,C) ] / m(C) .
In either case, the denominator was the
number (m(B) or m(C)) of respondents in
the State not licensed in a higher-priority
State.
In general, the numerator of a State
I weight, W(I), was the total frame count
licensed in State I after
Table B-1. State Sampling Rates and
Sample Sizes (Priority Ordered)
|
State |
Sampling Rate Percentage
Frame Size¹ |
Sampling Rate Percentage
Planned |
Sampling Rate Percentage
Actual |
Sampling Rate Percentage
Actual Sample Size |
| Total |
3,066,554 |
 |
|
54,125 |
| Wyoming |
5,123 |
15.00 |
15.26 |
782 |
| Alaska |
6,629 |
11.00 |
10.44 |
692 |
| North Dakota |
7,694 |
10.00 |
9.66 |
743 |
| Vermont |
7,906 |
9.00 |
8.73 |
690 |
| Delaware |
10,196 |
7.00 |
7.50 |
765 |
| South Dakota |
10,442 |
7.00 |
6.81 |
711 |
| Montana |
10,633 |
7.00 |
7.50 |
798 |
| Idaho |
11,876 |
7.00 |
6.408 |
761 |
| Nevada |
14,173 |
7.00 |
6.216 |
881 |
| Hawaii |
11,248 |
6.00 |
6.383 |
718 |
| New Mexico |
15,556 |
5.00 |
5.08 |
790 |
| North Carolina |
32,929 |
5.00 |
4.41 |
1.451 |
| North Carolina
Minority |
10,456 |
4.50 |
4.065 |
425 |
| Rhode Island |
16,752 |
4.50 |
3.94 |
660 |
| Utah |
17,345 |
4.50 |
4.94 |
857 |
| New Hampshire |
17,207 |
4.00 |
3.70 |
637 |
| District
of Columbia |
19,941 |
4.00 |
3.955 |
788 |
| West Virginia |
21,194 |
4.00 |
3.77 |
798 |
| Maine |
18,216 |
4.00 |
3.82 |
695 |
| Nebraska |
20,830 |
3.00 |
3.20 |
666 |
| Mississippi |
28,343 |
3.00 |
3.25 |
921 |
| Arkansas |
28,649 |
3.00 |
3.02 |
865 |
| Oklahoma |
31,156 |
3.00 |
3.09 |
963 |
| Kansas |
42,840 |
2.50 |
2.71 |
944 |
| South Carolina |
36,136 |
2.50 |
2.42 |
875 |
| Oregon |
35,007 |
2.25 |
2.14 |
749 |
| Iowa |
38,896 |
2.25 |
2.22 |
863 |
| Louisiana |
40,117 |
1.75 |
1.68 |
673 |
| Colorado |
43,371 |
1.75 |
1.93 |
837 |
| Kentucky |
43,750 |
1.75 |
1.67 |
729 |
| Alabama |
44,749 |
1.75 |
1.68 |
754 |
| Arizona |
46,165 |
1.75 |
1.78 |
821 |
| Connecticut |
50,143 |
1.75 |
1.63 |
818 |
| California |
247,562 |
1.75 |
1.625 |
4,022 |
| Minnesota |
59,098 |
1.50 |
1.633 |
965 |
| Maryland |
59,228 |
1.50 |
1.528 |
905 |
| Washington |
61,139 |
1.50 |
1.42 |
871 |
| Georgia |
79,327 |
1.50 |
1.56 |
1,238 |
| New
Jersey |
108,330 |
1.50 |
1.44 |
1,558 |
| New
York |
231,793 |
1.50 |
1.44 |
3,334 |
| Tennessee |
64,805 |
1.25 |
1.20 |
776 |
| Indiana |
74,184 |
1.25 |
1.22 |
903 |
| Virginia |
81,957 |
1.25 |
1.22 |
998 |
| Massachusetts |
105,955 |
1.25 |
1.11 |
1,172 |
| Illinois |
142,828 |
1.25 |
1.262 |
1,809 |
| Wisconsin |
67,415 |
1.125 |
1.19 |
805 |
| Missouri |
71,033 |
1.125 |
1.17 |
828 |
| North Carolina
White |
75,548 |
1.00 |
1.06 |
801 |
| Ohio |
134,915 |
1.00 |
1.06 |
1,432 |
| Florida |
170,108 |
1.00 |
.96 |
1,640 |
| Michigan |
113.753 |
0.90 |
.91 |
1,037 |
| Pennsylvania |
199,252 |
0.90 |
.89 |
1,782 |
| Texas White |
130,656 |
0.85 |
.86 |
1,129 |
| Texas Total |
163,585 |
1.514 |
1.577 |
2,580 |
| North Carolina
Total |
86,004 |
1.381 |
1.426 |
1,226 |
¹/ Adjusted frame
size.
²/ Since the actual distribution of names
differs for each State from the distribution
derivedfrom the merged States used for
the development of the 250 alpha-segments
some variation occurs between the planned
and actual sampling rates.
Removing the estimated total count of
State I nurses also licensed in higher-priority
States. The denominator, m(I), was the
number of State I respondents not licensed
in a higher-priority State. This weighting
scheme incorporated a nonresponse adjustment
that inflated the respondents’ data
to represent the entire universe. The
adjusted frame total shown in Table B-1
was used in computing the State I weight.
Estimation Procedure
State-level estimates can be computed
using the final set of sampling weights,
Wk (for sample nurse k).
For example, an estimate of the total
number of RNs working in Iowa may be based
on the following indicator variable, Xk:
Xk
= 1 if nurse k
worked in Iowa,
= 0 otherwise.
The desired estimated total may then
be written as
the sum being over all sample nurses.
Estimates of ratios and averages are
obtained as the ratio of estimated totals.
Sampling and Nonsampling
Errors
To the extent that samples are sufficiently
large, relatively precise estimates of
characteristics of the licensed RN population
of the United States can be made because
of the underlying probability structure
of the sample data. Such estimates are,
sometimes, an imperfect approximation
of the truth. Several sources of error
could cause sample estimates to differ
from the corresponding true population
value. These sources of error are commonly
classified into two major categories:
sampling errors and nonsampling errors.
A probability sample such as the one
used in this study is designed so that
estimates of the magnitude of the sampling
error can be computed from the sample
data. Nonsystematic components of nonsampling
error are also reflected in the sampling
error estimates.
Nonsampling Errors
Some sources of error, such as unusable
responses to vague or sensitive questions;
no responses from some nurses; and errors
in coding, scoring, and processing the
data are, to a considerable extent, beyond
the control of the sampling statistician.
They are called “nonsampling errors”
and also occur in cases where there is
a complete enumeration of a target population,
such as the U.S. Census. Among the activities
that were directed at reducing nonsampling
errors to the lowest level feasible for
this survey were careful planning, keeping
nonresponses to the lowest feasible level,
and coding and processing the sample data
carefully.
If nonsampling errors are random, in
the sense that they are independent and
tend to be compensating from one respondent
to another, then they do not cause bias
in estimates of totals, percents, or averages.
Furthermore, the contribution from such
nonsampling errors will automatically
be included in the sampling errors that
are estimated from the sample data.
Although nonsampling errors that are
randomly compensating do not tend to bias
estimates of simple statistics, correlations
or relationships in cross-tabulations
are often decreased by such errors, and
sometimes substantially. Thus, errors
that tend to be compensated in estimates
of simple aggregates or averages may (but
not necessarily will) introduce systematic
errors or biases in measures of relationships
or cross-tabulations.
Nonsampling errors that are systematic
rather than random and compensating are
a source of bias for sample estimates.
Such errors are not reduced by increasing
the size of the sample, and the sample
data do not provide an assessment of the
magnitude of these errors. Systematic
errors are reduced in this study by such
things as careful wording of questionnaire
items, respondent motivation, and well-designed
data-collection and data-management procedures.
However, such errors sometimes occur in
subtle ways and are less subject to design
control than is the case for sampling
errors.
Nonresponse to the survey is one source
of nonsampling error because a characteristic
being estimated may differ, on average,
between respondents and nonrespondents.
For this reason, considerable effort has
been expended in this survey to obtain
a high response rate through such actions
as respondent motivation and follow‑up
procedures. A high response rate reduces
both random and systematic errors. After
taking into account duplicates and frame
errors, the overall response rate to this
survey was 72 percent. State-level response
rates ranged from a little over 60 percent
in the State of Louisiana to 83 percent
in Wisconsin.
Sampling
Errors
Sample survey results are subject to
sampling error. The magnitude of the sampling
error for an estimate, as indicated by
measures of variability such as its variance
or its standard error (the square root
of its variance), provides a basis for
judging the precision of the sample estimates.
Systematic sampling, which was the selection
procedure used in choosing the alpha-segments
for this study, is convenient from certain
practical points of view, including providing
for panel rotation. However, it does not
permit unbiased estimation of the variability
of survey estimates unless some assumptions
are made.
Standard errors were estimated based
upon the assumption that the systematic
sample of 40 alpha-segments is equivalent
to a stratified random sample of two alpha-segments
from each of 20 strata of adjacent alpha-segments.
Ordinarily, this assumption should lead
to overestimates of the sampling error
for systematic sampling, but in this case
(with alpha-segments as the sampling units)
the magnitude of the overestimate is believed
to be trivial.
Regarding the sample as consisting of
20 pairs of alpha-segments thus obtaining
20 degrees of freedom) for variance estimation,
the probability is approximately .95 that
the statistic of interest differs from
the value of the population characteristic
that it estimates by not more than 2.086
standard deviations.
Specifically, a 95 percent confidence interval
for an estimated statistic 
takes the form
where 
is the estimated standard error
for 
Direct Variance Estimation
The direct computation of the sampling
variance used the jackknife variance estimation
procedure with 20 replicates of the sample.
Each replicate was based on 19 pairs of
alpha-segments and 1 alpha-segment from
the 20th pair. The actual respondent count
in the included segments for a particular
replicate was approximately 39/40ths of
the full respondent sample and was weighted
to represent the full population.
Variance estimates using the jackknife
approach require the computation of a
set of weights for the full sample and
a set for each replicate using the established
weight computation procedure i.e., 20
additional sets of weights). For the replicates,
the weights were based on the number of
responding nurses from the 39 segments
associated with each replicate. Having
20 sets of weights permits construction
of 20 replicate estimates to compare with
the estimate produced from all of the
data; each replicate estimate is based
on about 39/40ths of the data.
This procedure was performed 20 times,
once for each pair of alpha-segments.
The variance estimate is computed using
the following procedure. Define the following:
= an estimated total
for replicate I associated with alpha-segment
pair I, and
= an estimated total
obtained over the full sample.
The variance of 
Var 
is estimated by computing
If the estimate of interest is a ratio
of two estimated totals (e.g., the proportion
of RNs resident in Florida between 25
and 29 years old to the total number of
RNs resident in Florida), the variance
estimate for the estimated ratio would
be of the following form:
Following the example, the  and
 measurements
would be full sample and replicate estimates,
respectively, of the number of RNs resident
in Florida who were 25 to 29 years old,
while  and
would be the corresponding estimates of
the total number of RNs resident in Florida.
The variance of any other statistic, simple
or complex, can be similarly estimated
by computing the statistic for each replicate.
The jackknife variance estimator can
use either the full sample estimate,  or
the average of the replicate estimates.
While usually little difference exists
between the two estimates, RTI used the
estimator,  which
tends to provide more conservative estimates
of variance
Direct estimates of the variance were
computed for a variety of variables. These
variables were chosen not only due to
their importance, but also to represent
the range of expected design effects.
The average of these design effects (on
a State-by-State basis) provides the basis
for the variance estimate for variables
not included in the set for which direct
variance estimates were computed. Direct
estimates of the standard error (the square
root of the variance) are presented for
a selected set of variables in Table B-2.
Table B-3 shows the estimated State population
of nurses and the standard error of these
population totals.
Design Effects and
Generalized Variances
The generalized variance is a model-based
approximation of the sampling variance
estimate, which is less computationally
complex than the direct variance estimator
but is also less accurate. The generalized
variance equations use the national-level
or State-level estimates of the design
effect and, for some estimates, the coefficient
of variation (CV) to estimate the sampling
variance. The design effect, F, for an
estimated proportion 
is determined by taking the ratio of the
estimated sampling variance,
obtained by the jackknife method, to the
sampling variance of the 
simple random sample of the same
size. For the
proportion, 
this is given by
where n is the unweighted number of respondents
used to determine the denominator of 
Direct estimates of the design effect
were computed for a set of variables for
each State. The averages of the design
effects were then computed for each State
and the nation. These average design effects
can be used in formulas for estimating
generalized variances or standard errors.
This procedure uses average design effects
for a class of estimates instead of calculating
direct estimates (with a resulting economy
in time and costs), at the sacrifice generally
of some accuracy in the variance estimates.
A generalized standard error estimate
for an estimated proportion, 
for a State or for the United States,
is provided by the equation:
 (1)
where n is the number of survey respondents
used to determine the estimate  .
The multiplier F, the median²
design effect, depends upon the State
for which the estimated proportion was
generated. The median design effects are
on Table B-4
Generalized estimates of standard errors
can also be computed for estimated numbers
(or totals) of
Table
B-2. Estimates and Standard Errors (S.E.)
For Selected Variables or U.S. Registered
Nurse Population
| Number
of Nurses |
2,696,540 |
6,348 |
|
|
| Basic
Nursing Education |
|
|
|
|
| Diploma |
799,354 |
7,694 |
29.64 |
0.3100 |
| Associate
Degree |
1,087,602 |
11,559 |
40.33 |
0.3900 |
| Baccalaureate
Degree |
791,004 |
8,687 |
29.33 |
0.3100 |
| Master
Degree |
10,282 |
828 |
0.38 |
0.0300 |
| Doctorate
(N.D.) |
525 |
211 |
0.02 |
0.0100 |
| Not
Reported |
7,773 |
1,251 |
0.29 |
0.0500 |
| Employed
in Nursing |
|
|
|
|
| Yes |
2,201,834 |
9,663 |
81.65 |
0.3200 |
| No |
494,727 |
8,766 |
18.35 |
0.3200 |
| Racial/Ethnic
Background |
|
|
|
|
| Hispanic |
54,861 |
6,368 |
2.03 |
0.2400 |
| American
Indian/Alaska Native Alone (Non-Hispanic) |
13,040 |
1,264 |
0.48 |
0.0500 |
| Asian
Alone (Non-Hispanic) |
93,415 |
15,565 |
3.46 |
0.5800 |
| Black/African
American Alone (Non-Hispanic) |
133,041 |
15,373 |
4.93 |
0.5700 |
| Native
Hawaiian/other Pacific Islander
Alone (Non-Hispanic) |
6,475 |
960 |
0.24 |
0.0400 |
| White/Alone
(Non-Hispanic) |
2,333,896 |
20,970 |
86.55 |
0.8265 |
| Two
or More Races (Non-Hispanic) |
32,536 |
2,127 |
1.21 |
0.0800 |
| Race
Missing (Non-Hispanic) |
10,808 |
1,605 |
0.40 |
0.0600 |
| Not
Reported |
18,468 |
1,579 |
0.68 |
0.0600 |
| Employed
Status in 1996 |
|
|
|
|
| Employed
in Nursing FT |
1,576,675 |
13,973 |
58.47 |
0.4790 |
| Employed
in Nursing PT |
625,139 |
8 |
23.18 |
0.3200 |
| Not
Employed in Nursing |
494,727 |
8,766 |
18.35 |
0.3200 |
| Graduation
Year |
|
|
|
|
| Before
1961 |
233,583 |
5,003 |
8.66 |
0.1900 |
| 1961
to 1965 |
156,895 |
3,744 |
5.82 |
0.1400 |
| 1966
to 1970 |
199,732 |
3,615 |
7.41 |
0.1400 |
| 1971
to 1975 |
288,607 |
6,004 |
10.70 |
0.2200 |
| 1976
to 1980 |
370,937 |
6,668 |
13.76 |
0.2600 |
| 1981
to 1985 |
374,872 |
5,975 |
13.90 |
0.2200 |
| 1986
to 1990 |
332,627 |
4,472 |
12.34 |
0.1600 |
| After
1990 |
730,466 |
10,755 |
27.09 |
0.3800 |
| Not
Reported |
8,823 |
942 |
0,33 |
0.0300 |
| Employment
Setting (For nurses employed in
nursing) |
|
|
|
|
| Hospital |
1,300,323 |
13,009 |
59.06 |
0.4400 |
| Nursing
Home Extended Care |
152,894 |
5,758 |
6.94 |
0.2600 |
| Nursing
Education |
46,655 |
1,973 |
2.12 |
0.0900 |
| Public
Health Community Health |
282,618 |
5,519 |
12.84 |
0.2400 |
| Student
Health |
83,269 |
3,755 |
3.78 |
0.1800 |
| Occupational
Health |
36,395 |
1,950 |
1.65 |
0.0900 |
| Ambulatory
Care/Not Owned |
203,346 |
3,234 |
9.24 |
0.1600 |
| Owned/Operated
Ambulatory Care |
5,978 |
801 |
0.27 |
0.0400 |
| Other |
18,033 |
1,250 |
0.82 |
0.0600 |
| Not
Reported |
9,651 |
1,067 |
0.44 |
0.0500 |
| Type
of Position (For nurses employed
in nursing) |
|
|
|
|
| Administrator/Assistant
Administrator |
124,461 |
4,285 |
5.65 |
0.2000 |
| Consultant |
24,712 |
1,515 |
1.12 |
0.0700 |
| Supervisor |
78,295 |
3,057 |
3.73 |
0.1514 |
| Instructor |
61,641 |
2,605 |
2.80 |
0.1200 |
| Head
Nurse or Assistant |
105,803 |
3,562 |
4.81 |
0.1600 |
| Staff
or General Duty |
1,357,349 |
14,180 |
61.65 |
0.4900 |
| Practitioner/Midwife |
67,882 |
6,772 |
3.08 |
0.3000 |
| Clinical
Specialist |
40,833 |
1,753 |
1.86 |
0.0800 |
| Nurse
Clinical |
30,396 |
1,754 |
1.19 |
0.0680 |
| Certified
Nursing Anesthetist |
24,314 |
1,553 |
1.10 |
0.0700 |
| Research |
16,118 |
1,264 |
0.73 |
0.0600 |
| Private
Duty |
10,592 |
842 |
0.48 |
0.0400 |
| Informatic
Nurse |
8,406 |
892 |
0.38 |
0.0400 |
| Other |
216,047 |
5,563 |
9.81 |
0.2600 |
| Home
Health |
3,153 |
664 |
0.14 |
0.0300 |
| Survey/Auditors
Regulators |
5,096 |
635 |
0.23 |
0.0300 |
| Not
Reported |
23,747 |
1,639 |
1.12 |
0.0700 |
|
2,422 |
568 |
0.09 |
0.0222 |
| Highest
Nursing Education |
|
|
|
|
| Diploma |
601,704 |
7,787 |
22.31 |
0.3000 |
| Associate
Degree |
925,516 |
9,211 |
34.32 |
0.3200 |
| Baccalaureate |
880,996 |
9,997 |
32.67 |
0.3700 |
| Masters |
257,812 |
7,989 |
9.56 |
0.2900 |
| Doctorate |
17,256 |
1,274 |
0.64 |
0.0500 |
| Other |
7,682 |
966 |
0.28 |
0.0400 |
| Not
Reported |
5,573 |
987 |
0.21 |
0.0400 |
| Age
of Nurse |
|
|
|
|
| <25 |
66,482 |
3,001 |
2.46 |
0.1100 |
| 25
to 29 |
176,777 |
4,002 |
6.56 |
0.1500 |
| 30
to 34 |
248,375 |
4,924 |
9.21 |
0.1800 |
| 35
to 39 |
360,030 |
5,601 |
13.35 |
0.2000 |
| 40
to 44 |
464,425 |
8,576 |
17.22 |
0.3300 |
| 45
to 49 |
464,539 |
6,203 |
17.23 |
0.2200 |
| 50
to 54 |
342,415 |
5,903 |
12.70 |
0.2300 |
| 55
to 59 |
238,129 |
5,326 |
8.83 |
0.1900 |
| 60
to 64 |
156,061 |
3,374 |
5.79 |
0.1200 |
| >=65 |
154,467 |
4,420 |
5.73 |
0.1600 |
| Not
Reported |
23,861 |
1,570 |
0.92 |
0.0600 |
| Marital
Status and Children |
|
|
|
|
| Married
Child < 6 |
206,078 |
4,397 |
7.64 |
0.1600 |
| Married
Child > 6 |
783,573 |
10,691 |
29.06 |
0.3900 |
| Married
Child < 6 and > 6 |
204,053 |
5,397 |
7.57 |
0.2000 |
| Married
No Children |
720,077 |
8,923 |
26.70 |
0.3000 |
| Married
Child Unknown |
14,703 |
1,145 |
0.55 |
0.0400 |
| Wid/Sep/Div
Child < 6 |
11,973 |
894 |
0.44 |
0.0300 |
| Wid/Sep/Div
Child > 6 |
176,743 |
5,690 |
6.55 |
0.2100 |
| Wid/Sep/Div
Child All |
19,281 |
1,070 |
0.72 |
0.0400 |
| Wid/Sep/Div
No Children |
271,170 |
6,557 |
10.06 |
0.2500 |
| Wid/Sep/Div
Child UK/Refused |
3,728 |
612 |
0.14 |
0.0200 |
| Never
Married |
251,484 |
5,537 |
9.83 |
0.2154 |
| Not
Reported |
17,680 |
1,296 |
0.66 |
0.0500 |
| Mean
Gross Annual Salary for Full-Time
RNs |
46,782 |
117 |
|
|
| Mean
Scheduled Hours Per Year |
1,747 |
5 |
|
|
| Mean
Hours Worked in Week Beginning March
22, 2000 |
38 |
0.1 |
|
|
Table B-3. Direct Estimates
of State Nurse Population, Standard Error,
and Coefficient of Variation by State,
2000
| State |
2000
Estimated State Nurse Population |
Standard
Error |
Coefficient
of Variation (in Percent) |
| United States |
2,696,540 |
6,348 |
0.24 |
| Alabama |
41,513 |
570 |
1.37 |
| Alaska |
5,900 |
240 |
4.06 |
| Arizona |
42,658 |
858 |
2.01 |
| Arkansas |
23,291 |
472 |
2.03 |
| California |
226,352 |
1,606 |
.71 |
| Colorado |
40,084 |
625 |
1.56 |
| Connecticut |
41,767 |
760 |
1.82 |
| Delaware |
8,605 |
493 |
5.73 |
| District
of Columbia |
10,307 |
765 |
7.42 |
| Florida |
158,722 |
2,340 |
1.47 |
| Georgia |
67,958 |
1,112 |
1.64 |
| Hawaii |
10,228 |
506 |
4.95 |
| Idaho |
10,069 |
371 |
3.69 |
| Illinois |
126,166 |
1,608 |
1.27 |
| Indiana |
60,888 |
1,055 |
1.73 |
| Iowa |
35,089 |
537 |
1.53 |
| Kansas |
29,134 |
740 |
2.54 |
| Kentucky |
39,470 |
808 |
2.05 |
| Louisiana |
40,661 |
704 |
1.73 |
| Maine |
15,793 |
314 |
1.99 |
| Maryland |
51,456 |
957 |
1.86 |
| Massachusetts |
91,628 |
1,373 |
1.50 |
| Michigan |
100,769 |
1,159 |
1.15 |
| Minnesota |
54,920 |
573 |
1.04 |
| Mississippi |
24,874 |
515 |
2.07 |
| Missouri |
62,403 |
1,064 |
1.70 |
| Montana |
9,299 |
276 |
2.97 |
| Nebraska |
18,550 |
398 |
2.15 |
| Nevada |
12,940 |
361 |
2.79 |
| New Hampshire |
13,281 |
548 |
4.13 |
| New Jersey |
87,979 |
1,919 |
2.18 |
| New Mexico |
13,723 |
342 |
2.50 |
| New York |
197,532 |
1,740 |
0.88 |
| North Carolina |
83,016 |
1,097 |
1.32 |
| North Dakota |
7,661 |
277 |
3.62 |
| Ohio |
121,722 |
1,080 |
0.89 |
| Oklahoma |
27,083 |
625 |
2.31 |
| Oregon |
30,369 |
617 |
2.03 |
| Pennsylvania |
165,989 |
1,921 |
1.16 |
| Rhode Island |
13,690 |
381 |
2.79 |
| South Carolina |
32,539 |
721 |
2.22 |
| South Dakota |
9,587 |
222 |
2.32 |
| Tennessee |
55,947 |
956 |
1.71 |
| Texas |
150,251 |
1,147 |
0.76 |
| Utah |
15,648 |
254 |
1.62 |
| Vermont |
6,901 |
300 |
4.35 |
| Virginia |
66,466 |
1,183 |
1.78 |
| Washington |
54,771 |
704 |
1.29 |
| West Virginia |
17,725 |
456 |
2.57 |
| Wisconsin |
58,658 |
1,032 |
1.76 |
| Wyoming |
4,508 |
186 |
4.13 |
²The
median design effect was based on all
design effects for estimates of proportions
computed on selected variables. Using
a median instead of mean value avoids
the effects of extreme estimates of standard
errors, which can occur for some relatively
rare attributes. In prior years, an average
(mean) design effect was computed for
selected variables. Given that the distribution
of design effects is skewed to the right,
it is expected that the true median be
less than the true mean.
RNs in
a State, 
with a particular characteristic (such
as those employed in hospitals). The estimate
is a subtotal of the estimate 
the estimated total of RNs working and/or
living in the State. The standard error
and coefficient of variation of 
(represented by 
were determined for the nation and for
each State. The following explanation
is made simpler by defining the relative
variance of an estimate as the square
of its coefficient of variation.
Then the relative variance of the
ratio of 
to 
(called 
can be calculated as:
where F is the design effect for the
State of interest and n is the number
of respondents to the survey (i.e., the
number in the sample that were weighted
to obtain the estimate 
Then we can approximate the relative
variance of  denoted
u sing
This approximation is based on the first-order
Taylor series approximation to the variance
of a product and the assumption of zero
correlation between the estimate of ratio
and the denominator of the ratio.
Finally, the variance of  can
be estimated by multiplying by the relative
variance above by the square of the estimate.
The standard error of 
is thus estimated as
(2)
The standard error of an estimated percentage
for a region of the United States depends
upon a linear combination of the variance
of the same estimated percentages for
the States making up that particular region.
The estimated proportion for the region
is
here h is the number of States in region
R, and 
and 
are estimates for a particular State.
The formula used to approximate the standard
error of an estimated proportion for a
region is
 (3)
where  represents
the standard error of the estimated proportion
for the States and the standard errors
are estimated from equation (1) or from
direct estimation.
The direct standard error for an estimated
number for a region of the United States
also depends upon a linear combination
of the variance of the same estimated
numbers for the States that make up the
region. The formula used is
(4)
where the standard error 
of the estimated number 
is available either from the direct
procedures or from equation (2)
Table B-4. Median Design Effects for
Percentages Estimated from the Seventh
National Sample Survey of Registered Nurses,
2000
| State |
Median Design Effect |
| United States |
1.66 |
| Alabama |
1.10 |
| Alaska |
1.03 |
| Arizona |
1.02 |
| Arkansas |
0.99 |
| California |
1.16 |
| Colorado |
1.01 |
| Connecticut |
1.02 |
| Delaware |
1.12 |
| District
of Columbia |
0.98 |
| Florida |
1.17 |
| Georgia |
0.99 |
| Hawaii |
1.04 |
| Idaho |
1.01 |
| Illinois |
1.02 |
| Indiana |
1.02 |
| Iowa |
0.99 |
| Kansas |
1.08 |
| Kentucky |
0.98 |
| Louisiana |
1.03 |
| Maine |
0.96 |
| Maryland |
1.13 |
| Massachusetts |
1.06 |
| Michigan |
1.08 |
| Minnesota |
0.98 |
| Mississippi |
1.02 |
| Missouri |
1.11 |
| Montana |
1.00 |
| Nebraska |
0.97 |
| Nevada |
1.01 |
| New Hampshire |
1.03 |
| New Jersey |
1.03 |
| New Mexico |
0.99 |
| New York |
1.04 |
| North Carolina |
1.15 |
| North Dakota |
1.03 |
| Ohio |
1.06 |
| Oklahoma |
1.00 |
| Oregon |
1.05 |
| Pennsylvania |
1.05 |
| Rhode Island |
1.00 |
| South Carolina |
0.97 |
| South Dakota |
0.97 |
| Tennessee |
1.03 |
| Texas |
1.50 |
| Utah |
1.03 |
| Vermont |
1.07 |
| Virginia |
1.14 |
| Washington |
1.11 |
| West Virginia |
0.94 |
| Wisconsin |
0.98 |
| Wyoming |
0.97 |
|
