●
●
Narrow range—This class consists of (1) statistics that estimate a population attribute, for example, the number of persons in a particular income group, and (2) statistics for which the measure for a single individual during the reference period used in data collection is usually either O or 1 and, on occasion, may take on the value 2 or very rarely 3,
Medium range—This class consists of other statistics for which the measure for a single individual during the ref-erence period used in data collection will rarely lie outside the range O to 5.
Wide range—This class consists of statistics for which the measure for a single individual during the reference period used in data collection can range from O to a number in excess of 5, for example, the number of days of bed dis-ability.
In addition to classi&ing variables according to whether they are narrow-, medium-, or wide-range, statistics in the
sur-mates presented in M report. These charts represent standard errors of NHIS data. They should be used in preference to the charts that have appeared in all previous Series 10 publications.
Rule 1.
Rule 2.
Rule 3.
Rule 4.
Estimates of aggregates: Approximate relative stand-ard errors for estimates of aggregates such as the num-ber of persons with a given characteristic are obtained from appropriate curves, figures I–V. The number of persons in the total U.S. population or in an age-sex-race class of the total population is adjusted to oftlcial U.S. Bureau of the Census figures and is not subject to sampling error.
Estimates ofpercents in a percent distn”bution: Rela-tive standard errors for percents in a percent distribu-tion of a total are obtained from appropriate curves, figures VI-X. For values that do not fall on one of the curves presented in the chart, visual interpolation will provide a satisfactory approximation.
Estimates of rates where the numerator i$ a subclass of the denominator: This rule applies for prevalence rates or where a unit of the numerator occurs, with few exceptions, only once in the year for any one unit in the denominator. For example, hI computing the rate of visual impairments per 1,000 population, the nu-merator consisting of persons with the impairment is a subclass of the denominator, which includes all persons in the population. Such rates, if converted to rates per 100, may be treated as though they were percents and the relative standard errors obtained from the percent charts for population estimates. Rates per 1,000, or on any other base, must first be converted to rates per 100; then the percent chart will provide the relative standard error per 100.
Estimates of rates where the numerator is not a sub-class of the denominator: This mle applies where a unit of the numerator often occurs more than once for any one unit in the denominator. For example, in the computation of the number of persons injured per 100 currently employed persons per year, it is possible that a person in the denominator could have sustained more than one of the injuries included in the numerator.
Approximate relative standard errors for rates of this kind may be computed as follows:
(a)
(b)
Where the denominator is the total U.S. popula-tion or includes all persons in one or more of the age-sex-race groups of the total population, the relative error of the rate is equivalent to the rela-tive error of the numerator, which can be obtained directly from the appropriate chart.
In other cases the relative standard error of the
50 40 30
20
10 9 8 7
6
5 4 3
2
1
0.9 0.8 0.7 0.6 0.5 0.4 0.30.2
0.1
~ ~1 I II I I I II
[ ,
-H+ - + – + -H- - i+ - ————t——~ + –-H- +-1. . + . , ,
, , , , ! . [ ,,
I ~ — — —- - --- + - - - + — ~ =Fw 1 1
R-— — – – - - - --‘
---I
I I I I I 17I 1 I J 1 , I , 1 - n— r 77 - - - rrr- -
, ,
1,I
I
I I I I I I [1111111111 I I [ I I I I I I 111111111 I I I I 1-I!!! !!!l! !!!! I I I I I I 111111 II , I , , I , I I ,,
11111111111 ml-i-l I I I I I 1111111111111 I
60 50 40 30
20
10 9
; 6 5 4 3
2
:.9 0.8 0.7 0.6 0.5 0.4 0.3
0.2
0.1
A 2 3 456789A 2 3 456789A 2 3 456789A 2 3 456789A 2 3 456789A
10 100 1,000 10,000 100,000 1,000,000
Size of estimate (in thouaanda)
NOTE This curve representa estimates of relative standard errors bssedOTI 8 quarters of data collection for narrow range estimates of population characteristic or narrow range estimatea of aggregates using a 12-month reference period.
EXAMPLE OF USE OF CHART An estimate of 10000000 Persons with annual familv income of $15000 or mora, or 10,000000 persons who were hospitalized 1 or more times in tha previous year (on scale at bottom of chart) has a ralative standard error of 1.3 Percent (read from scale at left aide of cha~), or a standard error of 130,000 (1.3 percent of 10,000,000).
50 40 30
20
10 9 8 7 8 5 4 3
2
1
0.9 0.8 0.7 0.6 0.5 0.4 0.30.2
0.1
A 2 3 456789A 2 3 456789A 2 3 456789A 2 3 456789A 2 3 456789A
100
% 70 60 50 40 30
20
6 5 4 3
2
;.9 0.8 0.7 0.6 0.5 0.4 0.3
0.2
0.1
100
$
60 50 40 30
20
1$1
8 : 5 4 32
1
::
0.6 0.5 0.4 0.3
0,2
0.1
A 2 3 456789A 2 3 456789A 2 3 456789A 2 3 456789A 2 3 456789A
10 100 1,000 10,000 100,000 1,000,000
size of estimste (in thoussnds)
NOTE The curve related to hospital days is based on 4 quarters of data collection for wide range estimates of aggregates using a 12-month reference period; the cuwe for population characteristic is based on 4 quarters of data collection for narrow range estimates of aggregatea.
EXAMPLE OF USE OF CHART An estimate of 10,000,CKXJ daya of hospitalization in the previous year (on scale at bottom of chart) has arelative standarderror of 7.8 parcent (read from curve A on scale at Iaft side of chart), or a standard error of 780,000 (7.8 percant of 10,000,000). An estimata of 1,000,000 persona with 1 hospital episode or more (curve P) haa a reletive standard error of 5.7 parcant.
60 50 40 30
20
10 98 7 6 5 4 3
2
1 0.9 0.8 0.7 0.6 0.5 0.4 0.3
0.2
0.1
A 2 3 4 56789A 2 3 4 56789A 2 3 456789A 2 3 456789A 23
($
:;
50 40 30
20
10 9 8 : 5 4 3
2
;.9 H 0.6 0.5 0.4 0.3
0.2
0.1 4 56789A
50 40 30
20
10 9 8 7 6 5 4 3
2
0.:
0.8 0.7 0.6 0.5 0.4 0.3
0.2
0,1
I I 1
lrr,~
I I I I I I I I1111 !! 111 \ I I i 1 111111!1I I I B I II I I I I I I 111111111111 I I I I I 1111111~
I ,
I I I II i 11111 I
I 1
, , I
,
I I I I I I I I II
I
I I
II I I I I I I1111111111 I I I I I I I [1111 !11111 1 I I I I I 111111111111 \ I I I I !!!!!!!!!~
I I I
I
I i I I I 11111 (11[1111 I II
1 1 I 1 I I I I 111 [11 III I I1 “---:::— – – 1 “ - - --- 1-— – – - - 7--1---::::
90 80
% 50 40 30
20
10 9 8 7 6 5 4 3
2
;.9 0.8 M 0.5 0.4 0.3
0.2
0.1
A 2 3 456789A 2 3 456789A 2 3 456789A 2 3 456789A 2 3 456789A
10 100 1,000 10,000 100,000 1,000,000
Size of estimate (in thousancfs)
NOTE: The curves relsted to short-stay hospital daya and diachargaa ara basad on 4 quartera of data collection for wide and nerrow ranga estimates of aggregates using a 6.month rafarence period; tha curve for population chsracteristica is baaed on 4 quartara of data collation for nsrrow ranga estimatea of aggregatea.
EXAMPLE OF USE OF CHART An estimata of 10,000,000 hospital days (on scale st bottom of chart) has a relstive standard error of 10.2 percant (raad from cuwe,4 on scale at left aide of chart), or a standard error of 1,020,000 (1 0.2 percant of10,000,000). An estimate of 1,000,000 discharges from short-stay hoapitala (curve B) haa a relativa standsrd error of 7.4 percent, An estimate of 1,000,000 persons in the Northeast Region (curve P) has a relative standard error of 5.7 percent.
100 ,90 80 70 60 50 40
30
20
10 9 8 7 6
5
4
3
2
0.:
0.8 0.7 0.6 0.5 0.4
0.3
0.2
0.1
100 90 80 70 60 50 40
30
20
I II 111111111 I I I I I
I I I I I111111[ I I I I I I )
10
9 8 7 6 5 4
3
2
0.;
0.8 0.7 0.6 0.5 0.4
0.3
0.2
0.1
A 2 3 4 56789A 2 3 4 56789A 2 3 4 56789A-””
1 10 100
Estimated percent
NOTE These curves reprasant estimatas of relativa standard errors of percents bf population characteristics based on 8 quartera of deta collection for nsrrow range estimates. Base of percent is shown on curves in millions.
EXAMPLE OF USE OF CHART An estimate of 20 percent (onscale at bottom of chart) based on an estimate of 10,000,000 has a relative standard error of 2.7 percent (read from scale at the left side of chart), the point at which the curve for a base of 10,000,000 intersects the vertical line for 20 percent. The standard errorin percentage pointa is equal to 20 percent X 2.7 percent or 0.54 percentage points.
Figure V1. Relative standard errors of percents of population characteristics based on 8 quartars of data collection
Rule 5. Estimates of dl~erence between two statistics (mean, where Xl is the estimate for class 1,X2 is the estimate
rate, total, and so forth): The standard error of a dif- fer class
2,and
Vxland
Vxz are therelative errors of
100 .90 80 70 60 50 40
30
20
10 9 8 7 6 5 4
3
2
0.6 0.5 0.4 0<3
0.2
0.1
m 1 1 17
Iu _
\
1 ! I
1 I I I I I 111111111111 I I I I I I I
100 90 80 70 60 50 40
1
L
30
20
10 9 8 7
H
I I I I I I I I I I1111,I 654
I
I ,,,,,,,,,” :;
3i
I I I I I I I I 1 I 1 1 11 1111
I 1 I I I I I I I II 1 IIH+
2
0.:
0.8 0.7 0.6 0.5 0.4 0.3
0.2
0.1
A 2 3 4 56769A 2 3 4 56789A 2 3 4 56769A
1 10 100
Estimated percent
NOTE: These curves represent estimates of relative standard errors of percents of population characteristics based on 4 quarters of data collection for narrow range estimatea. 8ase of percent is ahown on cuwea in millions.
EXAMPLE OF USE OF CHARV An estimata of 20 percent (on scale at bottom of chart) based on an estimate of 10,000,000 haa a relative standard error of 3.6 percant (read from scale at the left side of chart), the point at which tha curve for a base of 10,000,000 intersects the vertical line for 20 percent. The standard error in percentage points is equal to 20 percant X 3.6 percent or 0.72 percentage points.
Figure VII, Relative standard errors of percents of population characteristics based on 4 quartars of data collation
I 100
.90 80
100 90 80
70 60
70 60 60
,,,
I 1 1 1 1 i 1I 1 1 1},,
I I I 1111111111 I I I I I I I I II IIIJ.
50
40
, ,, ,
1 1 1 t 1 ! 1 I I 140
30 30
20 20
10 9 8 7 6 5
1 I 1I Ir111111 I I
I
I I I I I I I I 109 8 7 6 5
! ! 50 \ , I i 100 &
4 4
3 3
2 2
0.:
0.8 0.7 0.6
0.:
0.8
M
0.70.60.5
H-!-l
0.50.4 0.4
0.3 0.3
0.2 0.2
0.1 ni
A 2 3 4 56789A 2
“.,
3 4 56789A 2 3 4 56789A
1 10 100
Estimated percent
NOTE These curves represent estimstes of relative standard errors of percents of hospital discharges baeed on 8 quarters of data collection for narrow range data using a 6-month reference period. Baae of percant is shown on curve in millione.
EXAMPLE OF USE OF CHART An estimate of 20 percent (on scale st bottom of chart) based on an estimate of 10,000,000 discharges has a relative standard error of 3.6 percent (read from scale at the left side of chart), the point at which the curve for the base of 10,000,000 intersects the vertical line for 20 percent.
The standard error in percentage points is equal to 20 percent X 3.6 percent or 0.72 percentage points.
Figure Vlll. Relative etandard errors of percents of ehort-etay hospital discharge
100 90 80 70 60 50 40
30
20
10 9 8 7 6 5 4
3
2
0.;
0.8
0.7 0.6 0.5 0.4 0.30.2
0.1
100 90 80 70 60
4
I
20
10 9 8 7 6 5 4
3
2 50 40
30
I
I1111111(
1 , ! ,
I I I I I
1111
I I I I I I I I I I 11111111 I 1 I I I N 1 1 I
I t I 1 I 1 t 1 , , I, }
EEM 1111
II
,,
Ill
o.4r 0.8 0.7
I
0.20.60.50.40.30.1A 2 3 4 56789A 2 3 4 58789A 2 3 4 58789A
1
10 100
Estimated percent
NOTE These curves represent estimates of relative standard errors of percent of 12 months of short-stay hospital days based on 8 quartera of data collection In wide rsnge estimates. Ease of percent is shown on curva in millions.
EXAMPLE OF USE OF CHART An estimata of 20 percent (on scale at bottom of chart) based on an estimate of 10,000,000 has a relative standsrd arror of 12.2 percent (read from scsle at the left side ofchart), the point at which the curve for the base of 10,000,000 intersects thevertical line for 20 percent. The standard error in percentage points is equal to 20 percent X 12.2 percent or 2.4 percentage points.
Figure IX. Relative standard errors of percents of 12-month short-etey hospital days based on 8 quarters of data collection
Baae OH 100
90 80 70 60 50 40
30
20
10 9 8 7 6 5 4
3
2
0.;
0.8 0.7 0.6 0.5 0.4
0.3
0.2
0.1
-m-f
[ ,
1 , I , I r 1r , ,I, i 100 90 80 70 60 60 40 3020
10 9 8 7 6 5 4
3
2
0.:
0.8 0.7 0.6 0.5 0.4
0.3
02
(-II
A 2 3 4 56789A 2 3 4 56789A 2 3 4 56789A”
1 10 100
Estimated percent
NOTE These curves represent estimates of relative standard errors of percent of 12 months of short-stay hospital days based on 4 quarters of data collection for wide range estimates. 8aae of parcent is shown on cuwe in millions.
EXAMPLE OF USE OF CHART An eetimate of 20 percent (on scala at bottom of chart) based on an estimate of 10,000,000 has a relative standard error of 17.0 percent (read from scale at left side of chart), the point at which the cunfa for the base at 10,000,000 intersects the vertical line for 20 percent. The standard error in percentage points is equal to 20 percent X 17.0 percent or 3.4 percentage points.
Figure X. Relative standard arrors of percents of 12-month short-stay hospital days based on 4 quarters of data collection