Docstoc

An Evaluation of Alternative Weights for a Weighted Estimator

Document Sample
An Evaluation of Alternative Weights for a Weighted Estimator Powered By Docstoc
					United States Department of Agriculture Statistical Reporting Service Research Division Staff Report Number AGES811027 Washington, 20250 October 1981 D.C.

An Evaltiorf~';'z Alternati~ W ..

- -"""

::..=-~- -,-,,---::- .--.

.~~ for a We!tJhted
....

Estimatol~

Jack Nealon

An Evaluation of Alternative Weights for a Weighted Estimator. By Jack Nealon; Research Division; Statistical Reporting Service; U.S. Department of Agriculture; Washington, D.C. 20250; October, 1981. Staff Report AGES811027.

ABSTRACT This report evaluates two weights that are being considered as input into a weighted estimator of hog and cattle inventories and the number of farms. The weight based on cropland acreage with a definitional modification is recommended for further research over the weight based on total acres minus continuous waste, woods (excluding grazed woodland), roads and ditches. The cropland weight is preferred because it simplifies the questionnaire and instructions to the enumerators, should be less susceptible to nonsampling errors, reduces the data collection and computer editing requirements, and should provide estimates of similar precision.

* * * * * * * * * * * * * * * * * * * *
* * * * *

* * * * * * * * * * * * This paper was reproduced for limited distribution to the * research community outside the U.S. Department of * Agriculture. The views expressed herein are not necessarily* those of SRS or USDA. * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * CONTENTS Page

SUMMARY ••• INTRODUCTION ESTIHATORS ANALYSIS
DATA COLLECTION REQUIREMENTS ••• NONSAMPLING ERRORS ••••••• ESTIMATES AND RELATIVE ERRORS

1 2 2

4 4
7

CONCLUSIONS AND RECOMMENDATIONS •. REFERENCES .... APPENDIX A APPENDIX B •..• APPENDIX C APPENDIX D

· 10 • • 12 · . 13

14
· . 19

· 21
• 23

SUMMARY Two weights were evaluated in five states during the 1981 June Enumerative Survey (JES) for use in a weighted estimator of hog and cattle inventories and the number of farms. The first weight was based on total acres minus This continuous waste. woods (excluding grazed woodland). roads and ditches.

weight was infested with reporting and recording problems that could result in seriously biased estimates. Although many of these problems can be remedied. The second weight. which was

numerous problems are expected to continue. weight.

based on cropland acreage. inherited many nonsamp1ing errors from the first However. these errors can be minimized in the future by modifying the definition of the cropland weight and by improving the questionnaire design and instructions to the enumerators. The cropland weight is preferred because it: (1) is easier to define,

which will simplify the questionnaire and instructions to the enumerators, (2) should be less subject to nonsampling errors. (3) yields greater data collection savings. (4) greatly reduces the number of computer edits, and (5) should provide estimates of similar precision. Three estimators were evaluated that are based on the cropland weight. These estimators generally provided similar estimates and levels of precision. The Survey Research Section recommends that research be conducted on a larger scale during the 1982 JES so that the modified cropland weight and the three estimators based on this weight can be evaluated thoroughly.

- 1-

INTRODUCTION The Statistical sixteen Reporting Service (SRS) uses an area frame sample in

states to provide weighted

estimates

of hog and cattle inventories estimator as input

and of the number of farms. into multiple proration frame livestock

In addition» estimates

SRS uses a weighted states.

in thirteen weighted A research

The weight or is the total tract conducted in

factor used in the operational by the total farm acres. and Oklahoma

estimator project

acres divided

Indiana» North Carolina Survey revealed states mainly

during the 1976 December

Enumerative

that the operational

weight was biased upward in all three of nonagricultural or waste land on the project

due to an underreporting 1/ One of

entire farm operation. was that other weights SRS collected Kentucky, called the

the recommendations

of the research

be explored. during the 1981 JES in Georgia» The first weight, Staff. which is is

data on two other weights North Carolina weight, and Ohio.

Minnesota,

alternative

was proposed by Methods waste, woods

This weight

based on total acres minus continuous roads and ditches.

(excluding

grazed woodland), that are There-

The term "continuous"

is defined as all parcels as a separate

large enough to draw off on an aerial photograph fore, small parcels waste. defined

field.

of waste within crop fields are not defined is based on cropland acres. Cropland

as "continuous" acreage was

The second weight as land devoted

to crops, hay acreage, of waste, woods fields.

land in summer fallow, idle cropland grazed woodland), roads and

and noncontinuous ditches within The purpose cropland

parcels

(excluding

the cropland

of this report is to compare to their:

the operational,

alternative

and

weights with respect errors»

(1) data collection and relative

requirements,

(2) nonsampling

and (3) estimates

errors.

ESTIMATORS The results from seven estimators will be presented in this report. The

first two estimators»

the open and closed estimators, were examined

which are not dependent (1) to have a two of the

upon any of the three weights» thorough estimators comparison

for two reasons:

of all estimators

available,

and (2) because

based on the cropland

weight depend to a certain

extent on the open

and closed estimators. estimators

The third and fourth estimators and alternative

are the weighted respectively.

based on the operational

weights.

- 2 -

Finally,

the last three estimators

are mainly based on the cropland weight. the problem that

These three estimators the cropland weight

differ only in their method of solving when a farm operation

is undefined

has no cropland. The formulas for each

A brief description

of each estimator

will now be given.

for the "area frame" estimate estimator
(1)

of a total and the variance A.

estimate

are presented

in APPENDIX

OPEN ESTJA~TOR:

This estimator, the enumerator

which is also referred

to as the farm

expansion,

requires

to obtain data on the entire farm for each inside the segment. the enumerator must identify the obtains

farm operation
(2)

with headquarters

CLOSED ESTIMATOR:

To apply this estimator,

each farm operation segment operated

which has any land in the segment. is called a tract.

The land within The enumerator

by a farm operation

data for the tract rather than for the entire farm from each farm operation. This estimator
(3) OPERATIONAL

is commonly

called the tract expansion. This estimator, which is also called

WEIGHTED ESTI~TOR:

the weighted

expansion,

is based on the following

two pieces of information (1) the entire farm

from each farm operation

which has land in the segment:

data for the survey item being estimated,

e.g. number of hogs and pigs on the which serve as a

entire farm, and (2) the total tract and farm acreages, weight to prorate

the entire farm data for a survey item to a tract basis. reports 150 hogs and pigs on the entire tract value is: ;~ farm,

For example,

if a farm operator

10 tract acres and 30 farm acres, then the weighted 50 hogs and pigs. previously,
(4)

*

150

The operational

weight

is always defined.

As mentioned

the operational

weight has been shown to be biased upward. This estimator is identical to the weight

ALTERNATIVE

WEIGHTED ESTI~TOR:

operational is used.

weighted

estimator

in all aspects

except that a different

The alternative

weight

is based on the tract and entire farm data waste, woods (excluding grazed woodland),

for total acres minus continuous roads and ditches.

This weight was always defined This estimator

in the five states even referred to as the

though it can be undefined. alternative weighted

is sometimes

expansion.

- 3 -

(5)

CROPLAND WEIGHTED ESTIMATOR I:

This weighted

estimator

makes use of the is not defined.

weight based on cropland The cropland weight

acres except when the cropland weight when there are no cropland

is undefined

acres on the entire the of

farm for a farm operation. operational weight

lolheneverthe cropland weight the data. weight,

is not defined, disadvantage

is used to prorate

A possible

this estimator

is that the operational whenever

which has been shown to possess weight is undefined. The

an upward bias, is utilized

the cropland

effect this may have on an estimate (1) whether operations the operational is limited weight

depends upon the following is biased when the population

two factors: of farm

only to those with no cropla.nd acres, and (2) how often when the operational weight is used. combines the concepts of of

a survey item is nonzero

(6)

CROPLAND WEIGHTED ESTI~TOR
estimators (1) whenever approach

II:

This estimator

the open and weighted this estimator operation are:

into a single estimator. the cropland weight

The mechanics

is defined

for a farm

the weighted

based on the cropland weight is not defined

is used, and is used.

(2) whenever

the cropland weight

the open estimator differs

(7)

CROPLAND WEIGHTED ESTIMATOR III:

This estimator

from the previous

estimator

in that the closed rather than the open estimator is not defined. the weight

is used when the

cropland weight estimate

That is, the tract data is used in the for a farm operation.

whenever

is 'not defined

ANALYSIS
The analysis will concentrate alternative and cropland weights. on comparisons The comparisons requirements, among the operational, will focus on the follow(2) nonsampling errors,

ing three areas:

(1) data collection and relative

and (3) the estimates

errors.

DATA COLLECTION REQUIRE~NTS:
In order to apply a weighted survey items is collected segments weight and multiplied estimator, entire farm information for the

from each farm operation corresponding

with land in the sampled to each operation. for a specified The

by the weight

is based on the tract and entire farm values

variable. since

If the weight the weighted

is zero, entire farm data does not '-laveto be collected tract value is automatically zero.

- 4 -

The operational

weight

is always greater than zero.

Therefore,

the

enumerator must obtain entire farm data from each farm operation the operational weight. For this five-state

when using obtained

study, the enumerators

entire farm data on hogs, cattle and gross farm sales from 6,859 farm operations. In addition, the total acres in the tract and in the entire farm were obtained in order to generate the weight. weight is that this weight is

from each farm operation An attractive

feature of the alternative

not always greater than zero. tion is classified roads and ditches, enumerator

If all the land in the tract for a farm opera(excluding grazed woodland), is zero. When this occurs, the In

as continuous waste, woods then the alternative weight

does not have to obtain entire farm data for the survey items. had an alternative weight

this study, 163 of the 6,859 operations However, of these 163 operations only a very small reduction of the alternative the additional

equal to zero. Therefore, because

only 32 had hogs and 95 had cattle. efforts is anticipated

in data collection

weight being zero.

This small savings is greatly offset by weight.

time needed to obtain the data for the alternative

Five questions were added to the questionnaire entire farm data from each farm operation these questions

to provide the tract and weight. Some of

for the alternative

were not easy to answer for many operations. (excluding grazed woodland), as continuous

For example,

every parcel of waste, woods

roads and ditches land. Also,

in the tract had to be classified

or noncontinuous

the farm operator had to estimate the total acres in the farm that were continuous Finally, woodland waste, woods (excluding grazed woodland), in the farmstead, the respondent weight. than the alternative reason is that fewer roads and ditches. pasture and grazed

the total farm acreage was asked. Therefore,

feedlots,

burden for the alternative

weight was greater than the operational The cropland weight requires weight for three reasons. and less complicated

less data collection

The first and most important

questions are needed to obtain the cropland weight.

Second, the cropland weight will be zero more often than the alternative weight. of zero. In this study, 580 of the 6,859 operations Therefore, had a cropland weight

entire farm data for items such as cattle would not Third, if the cropland

have to be asked from these 580 farm operations. weighted estimator

II or III is used, further savings in data collection

occur when the cropland weight is not defined. - 5 -

For the cropland weighted when the cropland weight used in the estimator. open estimator

estimator

I. therE' an' no additional

savings

is llndefined hecause

thE' 1))1".,11 ionn1 1,.,1eight then is
',.;L';.\' j.

Jlowl'ver.for the cropland

Ited estimator not defined.

II. the Recall with to the

is utili7.,~d\vhen the cropland weight data is not collected Therefore.

that with the open estimator headquarters outside

fn'ffi farm operations

the segment.

it would be unnecessary

obtain entire farm data from farm operat ions with he;ldquarters outside segment which do not have any cropland such operations weighted operation in the five states. on the entire farm.

There were 185

Finally. when dealing with the cropland

estimator

Ill. there is no need to collect entire farm data from any cropland weight because the tract data will be had

with an undefined

used in the estimator an undefined

for these operAtions.

A tot;ll <)1' 712 operations

cropland weight.

Table 1 gives a summar,' of the nUPlher of tracts for

in the five states for which entire farm data would not have to be collected each of the weighted estimators.

Table 1: The Number of Tracts With Zero o. Undefined '.,Jei~hts for Bach Estimator. The Total Number of Tracts is 6,859. OPERATIONAL WE I(~HTED ALTERNATIVE WEIGHTED CROPL AND \-JErCH TED CROPLA~n WT:IGHT"=D II SRO lR5 765 CROPLAND ~fEH-:HTED III 5~O 712 1292

r
Zero Undefined Zero or Undefined The data collection or undefined
0 0 0

163
0

0

---- ---

-

f)
.

Hi)

e,t) 0

-

g;wings provided by the crupland weight when it is zero Of thE' 580 operations with a cropAlso, of

are not as great as they appear.

land weight of zero, 466 had no hogs and 182 did not have any cattle. the 712 operations with an undefined Therefore. weight,

580 and lOR did not have any hogs and estimators

cattle, respectively. have different if operations

although

the three rropland weighted these differences

data collection

requirements,

will not he large tend

with no cropland acres, that is, undefined Nevertheless, illustrate these savings that the cropLmd

cropland weights,

not to have livestock. simplified collection questionnaire

in addition

to the greatly less data

weight requires

than the alternative

weight. requirements for the operational and crop-

A comparison land weights

of the data collection

is difficul t to quantify. to properly On
tIlE'

One add itiona 1 question will bE" needed on

the questionnaire respondent burden.

obtain the cropland weight, which will increase the

other hand. data coll('",ion savings are provided by thi'

- (,

-

cropland weight when the weight undefined.

is zero and are possible when the weight

is

How much savings are actually is zero or undefined,

provided will depend on how often which estimator is used, and the weight. In should not

the cropland weight

sizes of the operations conclusion,

that tend to have a zero or undefined requirements

the data collection

for the cropland weight weight.

be much higher
NONSAMPLING

than the requirements

for the operational

ERRORS:

Each weight particular

is constructed

by obtaining

tract and entire farm data for a were

variable.

Sections A and D of the 1981 JES questionnaire

used to obtain the tract and entire farm data, respectively, three weights. If a weight resulting Sections A and D are shown in APPENDIX is not accurately obtained, B.

for each of the

the weight may be biased,

thereby

in biased estimates. estimates

A research

project

in three states demonstrated weight were biased

that the weighted upward. -

based on the operational weight

1/

Therefore,

the operational

is not satisfactory. weight during of

Numerous the 1981 JES. waste, woods

problems were encountered Probably

with the alternative

the most serious problem was the classification roads and ditches

(excluding grazed woodland),

both in the tract to into one

and in the entire farm. classify waste, woods of two categories considered

In the tract, the enumerator grazed woodland), or noncontinuous

was instructed

(excluding

roads and ditches land.

-- continuous

A parcel of land was

continuous field.

if it was large enough to draw off on an aerial photograph Unfortunately, the wrong category was often coded by coded as

as a separate the enumerator. noncontinuous as continuous

For example, a field of woods was often erroneously

woods, or noncontinuous waste. These mistakes

waste in a crop field misclassified went undetected through the manual to discover edits many

in many instances, of these problems. Improvements instruction manual

and the computer

edits were not designed

could be made to the questionnaire that would reduce the incidence

wording

and the enumerator's

of errors when classifying criterion of

waste at the tract level. waste would decrease

Also, a change in the classification The classification

misclassifications.

criterion when

used

in the five states was difficult classifying

to apply successfully, The recommended

especially

waste on the entire farm.

classification

criterion

would classify

the waste simply as either waste within crop fields Finally, field by field computer Without

(regardless

of size) or all other waste. are needed problems to minimize

edit checks

erroneous to continue

classifications.

all of these changes,

are expected

at the tract level.

- 7 -

When several waste. extend

gathering

data for the alternative apparent.

weight

on the entire the concept photograph,

farm. of continuous does not does in

problems

were readily

First

of all,

that is, large enough itself to the entire

to draw off on the aerial farm in many cases because farm. woods, The

the enumerator

not delineate

each field for the entixe criterion for waste,

reccmmended

modification

the classification Referring divided

etc. should

remedy

this problem.

to Section

D in APPENDIX

B notice

that the total land was Feedlot, Grazed Pasture Woodland).

into three categories: Woodland, and (3)

(1) Cropland, Continuous Waste,

(2) Farmstead, Woods

and Grazed Roads

(Excluding makes

and Ditches.

The sum of the first weight.

two categories The assumption woods,

up the that all in the crop

denominator farmers fields

of the altern3ti~e include

was made

would

noncontinuous

waste,

etc. that were

in the first category.

This assumption,

however,

was not addressed manual. Preliminary with enumerators

on the questionnaire results

or in the enumerator's study in North and Ohio

instruction

from a followup North

Carolina indicated

and discussions

in Minnesota. assumption. indicated

Carolina

that this was not a valid in the North Carolina study

A total of 35 of the 113 respondents that they did not include of cropland of wasteland waste parcels

of waste,

,.,roods,oads r will

and ditches a it

in their estimate continuous parcel

acreage.

Also,

some fanners estimate

include

in the cropland

rather than include

in the continuous An easier in the fields be asked

category. is needed for cropland acrea:~e with regard to waste should should equal

definition in order

to minimize

these problems. all waste.

Th[' farm operator A computer acres check

either

to include to make

or to exclude

also be initiated

sure that the farm cropland

is at least

to the tract cropland Another stead, problem

acres. encountered had to do with Woodland. the second category -- Farm-

Feedlot,

Pasture

and Grazed

The entry was is almost

zero for this always associated

category with

in numerous This

cases

('v en though

a farmstead during

a farm.

error was corrected

the manual

edit when detected. for this is

However, category. biased

157 farm operat ion~;on the clean JES file had a zero value If the farmstead A computer category acreage was classified is needed as waste,

the weight

upward.

l~dit check

to assure

that the farm data tract

for the second data. stead Also,

is greater errors
.1

than or equal

to the corresponding not including

to reduce

caused

by the operator should

the farm-

in this category, is greater

computer

check

be made

that the second

category

than zero.

- 8 -

The alternative

weight was greater

than one on the clean JES file for because a computer check was not weight during

265 of the 6,859 operators.

This occurred

made on the weight being less than or equal to one. was originally greater than one for many of computer

The alternative

other operations checks comparing waste,

but was corrected separately total

the editing process because

land in the tract and farm and continuous To avoid getting weights developed greater

etc. in the tract and farm. edit check should be

than one, a computer

for the alternative

weight. in Minnesota, for another North research study

Review of a random sample of JES questionnaires Carolina revealed categories excluded required naturally enumerators and Ohio from operations that were selected

that much editing was required in Section D. For example,

in the offices

for the three land if refusals are

in North Carolina,

from the count, a total of 36 of 117 completed some office editing quite alarming encountered

JES questionnaires This is and

in one or more of the three categories. to light the difficulty operators

and brings

trying to classify

the total land into the three land in ~1innesota and Ohio. were not

use categories. Discussions popular thought

Similar problems with enumerators

were observed revealed

that these questions

with the enumerators

and farm operators.

Also, several enumerators

that the second and third categories Although nothing concrete

were the major source of the

unpopularity. the estimates obtained

can be said about the effect on that the data in its present farm data

from all this editing,

it should be obvious Therefore,

in the categories

is highly suspect.

form, Section D seems to be ineffective accurately for the alternative weight.

in obtaining

the entire

In summary,

there were many problems can be remedied addressing modifying by:

with the alternative
(1) improving

weight. on

Many of these problems the questionnaire, to the enumerators, ing additional level computer tions except field-level
(2) (3)

the wording

the topic in more detail the definition

in the instructions
(4)

of the weight,

incorporatfield-

tract-level

computer

edit checks, and (5) initiating of waste.

checks on the classification

All these recommendawith ease. The

for the field-level edits would require

edits can be implemented significant changes

in th(' p.dit program, the field by field

which should be avoided,

if possible.

However,

without

- 9 -

checks

on the waste

classifications,
1\ simple

many

errors

are exnected

to continue

with

the alternative that should

weight.

solution

to this diJemma

is to use a weight edit checks.

be more

accurate

and that does not require based acres on cropland

field-level

One such weight The weight nonsampling acreage,

would based

be a weight on rroDland Cropland

acreage. to hay

in this study w~s also subject as land devoted

errors.

acres was defined idle cropland and ditches

to crops, waste,

land

in summer

fallow,

and nonrontinuous within

woods Therefore, with in the one

(excluding

grazed

woodland), inherited

roads

the crop fields. weight waste

the cropland respect

weight

the problems

from the alternative and noncontinuous weight Most

to the misclassification [arm.

of continuous For example,

tract and in the entire

the cropland

exceeded of these

for 439 of the 6,859 operators unacceptable farm. weights were

on the clean JES data by misclassified

file.

caused

waste

in the tract or entire probably waste, farm. in the future is in the be deducted the cropland caused by

Some of the misclassifications

on the entire

farm were

not mentioning included

in the enumer:ltor's manual of cropland with

that noncontinuous for the entire

etc. was

to be

in the estimate problems

acreage

Reporting if cropland tract. when That

the cropland

weight

can he minimized as it currently Fields should the number simplify

acreage

in the entire

farm is defined with

is, all waste cropland. than one. manual'lOd field-level

associated This will

the crop reduce also

estimating

greatly

of times

weight

is greater

This approach \"ill greatly cl1ecks would

will

the questionnaire edits mentioned, that

and enumerator's required since

reduce

the number

of computer

not be needed. should

As previously

a computer the tract

edit check cropland

on cropland

acreage

he implemented

to insure

acres

does not exceed should

the farm cropland less subject

acres.

In conclusion, and record-

the proposed ing problems EST 1!lATES

cropland

weight

be much

to reporting

than the alternativ2
fftf','

weight.

ANn RF:IAT IVI\

liC;:

Numerous these

hog and cattle

survey

items were

selected farm while

for testing. others have

Some of data only selected.

items have data on the tract and entire farm. APPENDIX C contains survey

on the entire Twenty-two

a list of the survey items were evaluated.

items

hog and eighteen

cattle

Ten of the hog farm data, Finally,

items and eleven while twelve

of the rattle

items examined

had tract

and entire farm data.

hog and seven cattle number of farms with

items only had entire gross value

the variable, included

of sales of at least

$1,000, was

in the analysis.

- 10 -

The number of segments and the number of farm operations the JES are also shown in APPENDIX C.

in each state for for each and

The results of the analysis

state and the five states combined are shown in APPENDIX D. relative errors from each of the seven estimators

The estimates

are displayed

in Tables D-l The results for

through D-7 for survey items with tract and entire farm data.

survey items that only have entire farm data are shown in Tables D-8 through D-14 for the five estimators. The analysis dealing with hog items in North

Carolina was done with all the data and then with the hog data from seven large hog operations the results. deleted since these operations The estimates and relative had an overwhelming influence on do

errors for the five states combined

not include the hog data from these seven operations. Remember that the estimates shown [or the alternative stated previously, and cropland weights and may very well be

are affected by the reporting problems biased. For example,

the open estimate for the number of farms was always less alternative and cropland weighted estimates. The largest

than the operational, differences between occurred

in the three southeastern estimates

states where the percent difference The

the open and weighted

ranged from 11.0 to 19.7 percent.

possibility Therefore,

exists that biases in the weights have contributed detailed comparisons among the estimates

to the differences.

and relative errors are not To this end, only a few general and relative errors. for hog in

worth pursuing until these problems are resolved. comments will be made with regard to the estimates The estimates generated by the alternative

and cropland weights

and cattle items and the number of farms were generally very reasonable comparison Carolina. to the operational Large differences weighted estimates

in all states except North in North Carolina because the

existed for hog numbers Without

of several very large hog operations. hog estimates operational by the alternative estimates.

these atypical operations,

and cropland weights were much closer to the

weighted

The relative errors from the operational, approaches

alternative

and cropland In most

were also very similar except for hogs in North Carolina. weight had the lowest relative

cases, the operational weight the highest.

error and the cropland (estimators 5, 6

The three cropland weighted in most cases.

estimators

and 7) gave very similar results

This is largely due to the

fact that the value of the survey items was often zero when the cropland weight was undefined. Over 600 of the 712 operations with an undefined crop-

land weight had a zero value for many of the survey items, e.g. sows and boars no longer used for breeding. Table 2 provides a summary of the performance the three weights for the five states combined. - 11 of the estimators based on from the

The hog inventories

seven large hog operations impact on the hog results. average relative errors

in North Carolina

were excluded

because

of their is the

Shown in this table for each estimator

for hog and cattle

items with tract and farm data and

with only farm data and the relative The relative

error for the number of farms estimate. estimators are very similar to

errors by the cropland weighted and alternative estimators.

the operational

Table 2: Average Relative Error for Groupings of the Hog and Cattle Items and the Relative Error for the Number of Farms for the Five States Combined

------------~-----------------------VARIABLES and Farm Data: Tract -- --- --10 hog items 11 cattle items Farm Data: ---12 hog items 7 cattle items Number of farms 8.7 5.0 8.7 5.1 3.9 5.4 8.9 5.5 8.8 5.4 OPERATIONAl, WEIGHTED RELATIVE ERROR ALTERNATIVE WEIGHTED

(%)
I CROPLAND WEIGHTED II III

12.4 5.2 2.1

12.3 5.3 2.1

12.8 5.6 2.2

12.8 5.7 2.2

CONCLUSIONS The alternative operational Although persist. A cropland questionnaire to nonsampling reduces weight

AND RECOMMENDATTONS as a replacement errors during for the

that is being considered and recording

weight had many reporting

the 1981 JES. to

many of these errors can be corrected,

problems

are expected

weight would be much easier to define, would simplify to the enumerators, additional

the

and the instructions errors, guarantees

should be less susceptible savings, greatly

data collection

the amount of computer

editing,

and should provide

levels of precision of the and

that are similar cropland recording

to the alternative

weight.

A change in the definition reporting

weight used

in this study will be made to minimize the questionnaire

errors and to simplify defining

and enumerator's

manual.

This change involves acreage

the cropland

weight only in terms of crop1-and and noncontinuous waste, woods the 1981 JES. results

and not in terms of cropland grazed woodland),

acreage

(excluding

roads and ditches estimators

as was done during

The three cropland with respect

weighted

generally

gave very similar

to the estimates the redefined

and the precision cropland

of the estimates.

In order to weighted with

fully evaluate estimators

weight and the three cropland research

under a wider range of conditions, weight on a larger scale during

should be conducted

this cropland

the 1982 JES.

- 12 -

REFERENCES

Hill, George and Farrar, Martha, "Impact of Nonsampling Tract Survey Indications," of Agriculture, December, Statistical 1977. Reporting

Errors on Weighted

Service, U.S. Department

- 13 -

APPENDIX A
The formulas at thl' state level for the area frame estimate of the total the estimated varianct' of this est imate will be presented in this appendix each of the seven e:,t imators eva luatL'd,

and for

(1) OPEN ESTIMATOR:
The sample expressed as: s Y l L: i=l estimate of the total, namely Y , 1 for the open estimator can be

p.
I

~

j=l

r, . 1J L: k=l

s v: ' 1]
0

p.

~

r.o 1J
y

k

): L: 1=1 j=l

k=l

'"

e ..

~1

v " ij k'

where

y: k 1J
0

the

expanded and i

stratum e. ~J
0

data th

for land

a survey use for stratum, each

item

in

till'

k

th

segment,

jth

paper

the h tel

expansion . th

factor

segment

in the

j th

paper

stratum

in

1ant I use

stratum, replicates in the jth paper stratum in the

r ..

~J

the ,th ~

number of sample 1an d use stratum,
l)

p. ~
s

t h e numb er the
g, 'k

I

paper land

strata use strata

. h in tel

. th

1 ;1[1l

I use

stratum,

number

ell

in

the

sLlte,

~J L:

t=l
Oifg.
O

y ijk£

if

gijk 0,

> 0,

k

1J

g, 'k

~J

the number nf farm operations with stratum j and land usp stratum i, value .)f a survey item, , 0 operat Ion f .or t h e ,{..hi' operat t and i the e.g.

headquarters

in

segment

k,

paper

y, . '- 0
1

J

KA...

of hogs, for the entire farm t. on In tlle~. 1 th segment,] ,th paper stratum

number

th

land

llse

stratum.

- 14 -

The estimated

variance

of Y

l

can be expressed

as:

s
1:

p. 1
}:

r

ij

):

i=l j=l k=l

(1 __ 1_) e .. 11 {vy " 1J ( :'1 (1 __ 1_)
r ..

-

-~..

}2,

1J •

1J

1J v .. k 1: ~ r k=l ij

r ..

~

Tha t is

v ..

- 1J'

is

the mean

of the expanded

data

for a

th survey item in the jth paper stratum and i land use stratum. formula makes use of the paper strata to calculate the estimate and includes the finite population correction factor.

This variance of the variance

(2) CLOSED ESTlMA70R:
The sample estimator is: s Y 2
1:

estimate

of the total, namely

Y , 2

for the closed

p. 1
1:

r .. 1J
l:

s y: .k 1J
)~

p.
1

r. 1j
1:

L

i=l j=l k=l

i=l j=l k=l

e .. y. ok' 1J 1J the open estimator,

where

y:. , 1Jk

e .., s, p. and r, are defined 1J 1 1j f. 'k IJ
L

the same as ~ith

t

Y"k IJ

l=l

ijkl

if f. Ok > 0, IJ = 0,
0f

0 if f" k IJ
f"k IJ th e num b er stratum

' tract farm operatIons th and i land use stratum,

th I in tle k segment,

J,th paper

the "tract" the tract, paper

value of a survey item, e.g. number of beef cows in th th ,th for the l farm operation in the k segment, J an d
1

stratum

.th

land use stratum.

- 15 -

The V(Y1) .
(3)

formula

for

the

estimated

variance

of

Y2 is

t h(~ same

in

notation

as

OPERATIONAL WEIGHTEDf:STlMATOR: -- - ----- ----The sample s estimate roo LJ
(1

t

the

total, s

that roo
/:

is

Y 3' is

given

by:

p.
\'

1

P.

Y

J

Z

L:
k=l po,

i=l eo
0 ,

j=l r.o,

Y" "k 11

I

,:

1

1.1

eoo

1=1 j=J and

k=l are

1J Yij k'
defined as in the firs t two estimators,

where

Y~jk'

1J

1J
0

s,

1

r i jk

Yook1 IJ

f.

IJ
L

k
dij

Yijk

£=1

k£ v ijkl
00

if

f ij k

>

0,

0 if
the

f

1)

k

0:

0,
t

weight

Ln the.c

h

farm

operation

in

the

~h k- segment,

th jth paper stratum and i land use stratum. This weight is the ratio of tl1,,-' total c,cres in the tract In the total acres in the entire farm. The sum of the weights O\ll'r all segments for: a given unbiased The (4) formula for tht: farm operation of is the equal to one. This insures that

i3 is

an

t'stimate pstimated

population of Y3 is

tnta1. the same in notation as V(Y ). 1

variance

ALTERNATIVE WEIGHTF:LI STIMATOR: E The sample s estimat,,· po
01

the

total, s

Y , 4

is:

Y

4

),

1 L:

r

0

"

11

p.
I

1

roo k=l

IJ
, ,~
(-:J. ••

i=l

j=l

k=l

v Ii k

L

1=1 j=l

1)

Y 1J - ••

k' -

where Yijk

I:

f"k
L

IJ
h.
if

t'=l

1J
0

kP Y ijld:' if ' k 0,

f

Ok 1J
0

>

0,

0

f

1J
00

- 16

-

bijU

the ratio of the tract and entire farm figures for the variable, total acres minus continuous waste, woods (excluding grazed wood1aQd), roads and ditches for the [ farm operation in th segment, J,th paper stratum an d 1 ,th 1an d use stratum, th e k
th

and all other variables

are as previously of Y

defined. 4 uses the same notation as

The formula for the estimated variance

V(Y1)· (5) CROPLAND WEIGHTED ESTIMATOR I:
The formula s
L

for Y P. 1
L

s

can be expressed as: s Y~ 'k 1J
L

r, ,

1J

Pi
L

rij
L

L

e

i=l j=l k=l f"k 1J
L

i=l j=l k=l

ij Y ijk'

where Y"k
1J

[=1

d. 'U 1J

Yijk£ 0,

if f. 'k > 0, 1J

1J ° if f"k
cijU aijU if c ijU if c ijU

is defined, is not defined,

the ratio of the tract cropland acres to the entire farm
th tract farm operation i n th e k th 1 d segment, J.th paper stratum an d i an use stratum, nth cropland acres for the ~

and all other variables The formula estimates.

are as previously

defined.

for the V(Y ) is of the same form as the previous variance 5

(6)

CROPLAND WEIGHTED ESTIMATOR II:
The formula for Y 6 is: s Y 6
=

Pi
L

rij
L

s

Pi
L

r

ij
L

L

i=l j=l k=l

Yfjk= i:1 j=l k=l

e

ij Y ijk '

- 17 -

f.lJ'k
,..

where

f=l

Y~J ki f, 'k
lJ

if
,) ,

f

ijk

>

0,

0 if

=

cij 1<1 y ij kt y,lJ'1<1

if
if

cijl<1
C

is

defined, has

ij kl

is not defined and thE' farm operation in thl' segment, headquarters is not defined and the farm operation not have headquarters in the segment, previously defined, 6 is of the same form as

°
and all other for The formula var iance
(7)

iI'

cij ki as

does

vari<1blL"s are the estimated fur the

variance first
III:

of Y

the

estimate

given

estimator.

CROPLAND

WEIGHTED

ESTLMATOR

The formula s

for p,
l

Y7 can be given r, ,
l]

by: p,
l

s

r, .
lJ

l:
i=l

l:
j=l

:s:

k=l

v:1J'k .

:s:

L

i=l

j=l

I e, , y, 'k' lJ lJ k=l

f"k lJ

L:
where y" k

lJ

£.=1

Y ij kf f"k lJ

U

if

f"

lJ

k

>

0,

°

if

0,

y,lJ'1<1

I~
other

c ij 1<1 y,lJ'kf t ij k£. if

if
.

cijk£ is are not as

is

defined,

c ij k.t ll'~

defined, previously as the defined, estimated variance for Y given for

and all

variab

the

The V(Y ) has the 7 first estimator.

SilTnl'notation

l

- 18 -

APPENDIX SECTION A - ACREAGES

B BOUNDARY Acres

OF FIELDS AND CROPS INSIDE BLUE TRACT

How many acres are inside this blue tract boundary drawn on the photo (or map)? Now I would like to ask about each field inside this blue tract boundary FIELD NUMBER ... 1 1. TOTAL ACRES IN FIELD
831

I------•
4

and its use in 1981

2

3

•
831

•
831

•
831

•

2. CROP OR LAND USE (SDecifv) 3a. CONTINUOUS WASTE, WOODS (excluding qrazed woodland) ROADS, DITCHES, etc. 3b. GRAZED WOODLAND, EMPTY FARMSTEADS WITH ASSOCIATED FEEDLOTS 3c. WOODS, WASTE, ROADS, DITCHES, etc. (non-continuous parcels within field boundaries) 4. OCCUPIED 5. PASTURE FARMSTEAD OR DWELLING Not in crop rotation Used only for pasture

•
1830 830

•
830

•
830

•
•
829

•
629 829

•
829

•

•
843 842 856
NO YES

•
842 856
NO YES

•
842 856

•
842 856
NO YES

Permanent Cropland

• •
0

•
0

• •
•

•
NO YES

6. TWO CROPS PLANTED

IN THIS FIELD for har-

0

• • •

0

• •

844

844 61

844

844 61

vest this year or two uses of the same crop?
61

7. ACRES LEFT TO BE PLANTED? 8. ACRES IRRIGATED AND TO BE IRRIGATED?

-

-

•
•

61

6. __ 550

6 __ 550 553 540

9. SPRING WHEAT OTHER THAN DURUM--Planted
553

10. DURUM WHEAT 11. WINTER WHEAT 12. 13. 14. 15. OATS 16. 17. BARLEY 18. 19. CORN 20. RYE

-

Planted and to be planted ~~n!e~ ____________ For Grain
547 540 541----

• • •
•

• • • • •

54i-_....!-547

•
•

~I~n!e..!! ~nE .!o_b~ .e.1~n.!~___ For Grain ~I'!...n!e~E.n.£l!~ ~ £!~n.!~ ___ For Grain ~~n.!e~ ~n.£l ~ ~ £I~'!!~ ~ For Grain ~I!n.!eil ~I!.d_t~ ~ey!a!!~~ For Grain PLANTED Use Ac res __ ___

---_.!..548 533

--_....!_548 533 - 534

---_..!-534 535

_....!_-

• • 535 --_....!_• ------ 536 536 • • 530 530 • -------- -----~-531 531 • •
•
653 653

• • • 6__ 6__ • • 550 550 • • 553 553 • • 540 540 5<lr - _!-- 541- - -'"--• • 547 547 ---_!-_- ---.--- 548 548 • • 533 533 • ------ ---_!-_534 534 • • 535 535 • ---_!.._------ 536 536 • • 530 530

---_...!_531

---_..!_531

• • • • • • • •
• •

• • • • •
•

23. OTHER USES OF GRAINS Acres abandoned 24. HAY 25.

cut for hay, silage etc.

and ALFALFA Cut ALFALFA and to be OTHER HAY cut

MIXTURES Kind Acres

•
65

• • • •
•

653

653

-

65

26. SOYBEANS 27. TOBACCO 33 34 35 36 SUGAR BEETS SUNFLOWERS FLAXSEED POTATOES

600

Planted and to be planted
67

Class (SDecifv Planted and to be planted

)

691 697

Planted and to be Dlanted
693

PI<1ntedand to be Dlanted
552

Planted and to be Dlanted Acres planted or in use FALLOW - - Idle all during 1981

38. OTHER CROPS 39. LAND IN SUMMER 39a. IDLE CROPLAND

-..•847 857

• • • • • • •
• • •

-

65

-

65_ 600 67_ 691 697 693 552

600 67_ 691 697 693 552

600 67_ 691

--R47 857

• • • • • • •

697 693 1552

..•....
R47 857

• • •

--847 857

• • • • •

• •

- 19 -

SECTION D - ACRES OPERATED

Refer to Face Page for Type of Operation Individually Partnership or Joint Managed land ..... . . ~) Go to item 1.

D· Go to item@

1. Now I would like to ask you about the total acres you operate under this land anangement. Include all cropland, woodland, pastureland and wasteland. How many acres do you:

a. Ow;1 ?
b. Rent from others? c. Rent to others? , , ,

""" ,.,, ,.,,

"

. . .

I
I

tol

•

to2

• • •

leos

Then the total land you operate is (items a + b - c) This total land operated consists of how many: Cropland Acres Farmstead, Feedlot, Pasture, & Grazed Woodland C.ontinuous Waste, Woods (Excluding grazed woodland) Roads, Ditches, etc.

L
@Now How many acres do you operate as a hired manager? ..•.....

O 7 _ ~

(Go to Section E)

I would like to ask you about the total acres you operate as a hired manager. ,., ,. ,, .

I,-------

eo.

(Complete codes 903, 906 and 907 above, then go to Section E.)

- 20

-

APPENDIX

C

The following hog and cattle items, which have tract and farm data, were used to compare the estimates and relative errors seven estimators: LABEL SURVEY ITEM

from the

HI
H2

H3

H4
H5

H6
H7 H8

H9

HlO
Cl
C2 C3

Sows, gilts and young gilts bred and to be bred Boars and young males for breeding Hogs and pigs for breeding Sows and boars no longer bred Market and home use hogs and pigs under 60 1bs. Market and home use hogs and pigs between 60 and 119 1bs. Market and home use hogs and pigs between 120 and 179 1bs. Market and home use hogs and pigs over 180 lbs. Hogs and pigs not for breeding Total hogs and pigs Beef cows Milk cows Cows that have calved Bulls weighing at least 500 Ibs. Heifers weighing at least 500 1bs. Heifers weighing at least 500 lbs. Other heifers weighing at least 500 Total heifers weighing at least 500 Steers weighing 500 lbs. or more Heifers, steers and bulls weighing Total cattle and calves

C4
C5 C6

C7
C8

for beef cow replacement for milk cow replacement 1bs. Ibs. less than 500 1bs.

C9
ClO

C11

The following hog and cattle items and the number of farms, which only have information for the entire farm, were used to compare from the five estimators: LABEL H11 H12 SURVEY ITEM

the results

H13 H14 H1s
H16

H17
H18

H19
H20 H21 H22

Farrowing intentions for the next quarter Farrowing intentions for the second quarter Farrowing intentions for the next six months Sows and gilts farrowed during the previous quarter Sows and gilts farrowed during the second previous quarter Sows and gilts farrowed during the past six months Pigs from the previous quarter Pigs from the second previous quarter Pigs from the last six months Hog and weaned pig deaths during the previous quarter Hog and weaned pig deaths during the second previous quarter Hog and weaned pig deaths during the last six months

- 21 -

LABEL C12 C13
C14

SURVEY

ITEM

CIS
C16 C17 C18
Fl

Cattle and calves on feed for s]a~ghter COWti and heifers remaining to calve in 1981 Calves born since January 1, 19H] Calf crop for 1981 Cattle deaths since January 1, 1981 Calf deaths since January 1, 1981 Total cattle and calf deaths since January 1, 1981 Numh!'} of farms (gross sales of
,It

least $1000)

The Number

of Segments

and Farm Operations Number of Segments 436 338 343 391 324 1,832

for Each State Number
of

STATE
---.----

Farm 0 erations 1,046 1,630 1,470 1,511 1 202 6,859

Georgia Kentucky Minnesota North Carolina Ohio TOTAL ----- ----- --- ---

- 22 -

Table D-1:

Estimates and Relative Errors for Each Estimator for Selected Survey Items in Georgia.

OPEN SURVEY ITEM ~ HI H2 H3 H4 H5 H6 H7 H8 H9
I
N W

CLOSED R.E.
(%)

Est. (000) 262.7 21.7 284.4 12.6 696.7 529.3 261.4 99.0 1599.1 1883.5 746.6 59.8 806.4 50.5 107.8 18.5 58.2 184.5 60.5 575.3 1677.3

I

OPERATIONAL WEIGHTED Est. (000) 212.8 19.0 231.8 17.5 599.7 422.9 276.4 168.8 1485.5 1717.3 883.1 126.8 1009.9 57.9 132.1 38.3 60.1 230.5 70.2 682.7 2051.1

Est. I R.E. (000) (%) 199.3 18.8 218.1 14.7 561.1 337.2 234.4 93.8 1241.0 1459.1 891.2 93.6 984.8 60.1 93.9 63.8 63.4 221.0 74.9 742.0 2082.9 20.4 19.8 20.2 36.3 19.2 26.7 35.7 40.8 21.7 21.0 12.8 47.4 12.6 12.6 16.2 32.7 40.8 16.6 35.1 12.4 11.3

I

ALTERNATIVE WEIGHTED Est. (000) 202.7 18.0 220.7 15.5 582.5 415.5 265.4 163.9 1442.7 1663.4 827.9 137.4 965.3 56.3 118.6 42.0 56.4 217.0 74.0 652.7 1965.2 IR.E.
(%)

R.E.
(%)

CROPLAND T..TEIGHTET) I Est. R.E. (000) (%)

I

CROPLAND WEIGHTED II Est. R.E. (000) (%)

I

CROPLAND WEIr,}lTED III Est. Ill.E. (000) (%) 191.9 17.6 209.5 15.2 541.1 410.1 264.0 158.0 1388.4 1597.9 854.7 120.7 975.4 56.6 113.8 44.8 68.4 227.0 95.0 677 .0 2031.1 12.3 10.6 11.9 33.2 12.1 16.9 25.0 31.3 14.5 13.7 8.8 32.3 8.9 8.7 11.0 31.1 28.8 12.6 26.3 8.7 8.9

26.1 20.7 25.5 40.3 22.3 34.8 33.7 47.8 25.1 25.1 13.1 58.0 13.1 11.1 23.0 55.0 36.7 19.3 35.3 13.0 13.0

13.1 10.5 12.8 39.1 11.8 16.2 21.7 27.4 13.5 13.0 7.2 30.5 7.4 7.7 10.3 29.8 23.8 10.1 18.6 7.3 6.9

13.3 11.1 13.0 32.5 12.5 15.5 22.1 28.0 13.6 13.1 7.9 29.6 8.0 8.2 10.0 28.9 24.9 10.6 22.6 7.7 7.5

191. 7 17.5 209.2 15.1 539.3 415.5 263.4 154.3 1387.5 1596.7 854.5 120.8 975.3 56.8 113.9 40.0 60.0 213.9 78.9 672.9 1997.9

12.3 10.7 12.0 33.3 12.1 16.7 25.1 30.7 14.4 13.7 8.9 32.2 8.9 8.5 10.7 29.1 22.4 11.0 26.3 8.6 8.5

191. 7 17.5 209.2 14.4 533.3 418.8 264.8 129.8 1361.0 1570.2 783.0 120.9 903.9 52.5 104.9 34.8 49.8 189.5 74.0 613.7 1833.7

12.3 10.7 12.0 34.7 12.2 16.8 25.0 30.8 14.6 13.8 9.0 32.2 9.2 8.9 11.5 31.1 23.9 11.8 27.3 8.8 8.9

HI0 Cl C2 C3 C4 C5 C6 C7 C8 C9 C10 Cll

Table D-2:

Estimates and Relative Errors for Each Estimator for Selected Survey Items in Kentucky.

OPEN SURVEY ITEM ~ HI H2 H3 H4 H5 H6 H7 H8 H9 H10
I
I

CLOSED R.E. (%) 20.3 ]8.8 20.1 45.0 23.5 31.9 33.6 36.4 24.7 23.S
10.0

Est. (000) 139.2
L O. 'i

I

OPERATIONAL WEIGHTED Est. (000)
I
I

ALTERNATIVE WEIGHTED Est. (000) 170.9 l:i.5
i

CROPLAND v.'EIGlITED
T

Est . (000)

I

R.E. (%) 19.7 17.1 19.3 71. 3 22.5 30.3 36.8 33.5 24.0 22.6
h. LI

I R.E. (%)
12.1 11.4 12.0 35.1 13.2 18.0 18.7 21.8 13.8 13.2

I R.E. (%)
12. 'i
L1.I-1

Est. (000) 171. 7 14.0 185.7 3.6 480.5 211. 6 164.0 105.5 965.1 1150.8 I
,~ J-

I

R.E. (%) 14.6 13.8 14.5 35.0 16.3 21.3 21.9 27.9 17.7 17.0

CROPLAND WEIGHTED II Est. R.E. (000) (%)

I

CROPLAND WEIGHTED III Est. , R.E. (000) (%) 170.0 13.8 183.8 3.6 480.0 213.7 167.4 107.5 972 .1 1155.9 ~
')

i

149.7 5.0 345.7 159.5 78.1 48.6 636.8 78f1. S
' " 1-14 I

I

I

13/~ .2 11. .1 145.2 1.1 386.1 182.0 139.0 66.1 774.3 919.S
1Oc1-. 1
'

173.7
1 3. 5

170.2
l3.8

14.7
13.9

14.8 13.9 14.6 35.2 16.3 21.1 21. 5 27.4 17.6 l6.9
"

I

187.3 3.8 462.1 205.0 155.5 86.9 913.3 1l00. 5
- -I

184.4 3.5 464.4 206.4 154.6 86.0 914.9 1099.3

12.3 32.9 13.6 18.4 19.5 22.7 14.3 13.7

I

184.1 3.6 479.2 211.3 161. 2 105.4 960.8 1144.8
';'~
/

14.6 35.2 16.3 21. 3 22.3 27.9 17.8 17 .1
"

,!

C1 C2 C3 C4 C5 C6 C7 C8 C9 CI0 Cll
I

.

1

.,

-

,
12. 1 4.7 5.0 8.4 23.7 17.3 9.5 17.2 5.2 4.4 239.4 1286.3 76.4 138.6 109.5 60.1 308.8 266.9 930.7 2869.1 11.9 4.7 5.2 8.4 24.0 17.4 9.5 17.3 5.3 4.5

C

223.7 1070.8 65.5 119.2 141. 6 30.5 291. 3 184.0 782.1 2393.8

17.9 8.2 9.2 13.6 49.0 30.2 24.0 27 .6 8.4 7.9

226.7 1308. 7 80.2 138.5 93.8 42.0 274.3 251.7 936.9 2851.9

16.2 6.2 10.3 11.8 29.4 21.4 12.0 21. 7 6.5 5.5

234.1 1253.8 74.0 136.2 111.9 60.0 308.1 268.3 902.8 2807.0

i

227.2 1121.7 64.6 134.9 108.9 61.0 304.8 200.5 792.0 2483.6

11.8 5.6 5.8 11.0 28.4 21.9 11.7 14.4 6.3 5.6

227.4 1l02.2 63.0 131.0 107.8 61.1 299.9 189.0 780.8 2434.9

11.8 5.6 5.8 11. 3 28.6 21.9 11.9 14.9 6.4 5.6

227.2 1120.4 64.3 134.9 109.6 61.1 305.7 217 .0 785.0 2492.4

11.8 5.6 5.8 11.0 28.2 21.9 11.7 17.2 6.3 5.6

Table D-3:

Estimates and Relative Errors for Each Estimator for Selected Survey Items in Minnesota.

OPEN SURVEY ITEM HI H2 H3 H4 H5 H6 H7 H8 H9 H10
V1 I

CLOSED R.E.
%

OPERATIONAL WEIGHTED

CROPLAND WEIGHTED I Est. 000 648.0 12.7 14.1 12.7 29.9 14.0 12.3 14.5 18.1 12.5 12.3 10.2 8.2 6.3 10.3 11.8 9.0 13.7 7.4 10.4 6.7 5.9 13.7 12.6 29.4 14.0 1084.7 766.2 454.9 4239.4 4925.1 622.0 915.7 1538.7 57.8 95.9 360.3 285.6 741. 8 741.0 1083.6 4161.9 12.4 14.6 18.1 12.5 12.4 9.9 8.3 6.3 10.2 11.3 9.2 13.6 7.3 10.3 6.4 5.8 43.7 691. 7 26.5 1907.7 1082.5 758.2 453.3 4228.2 4919.9 612.5 937.1 1549.6 59.4 92.9 354.6 291. 5 739.0 754.8 1090.0 4192.7

CROPLAND WEIGHTED II

CROPLAND WEIGHTED III

15.3 14.6 15.2 19.1 1791. 4 981. 8 749.8 406.3 3948.4 4579.0 452.6 972.1 1424.8 54.2 77 .3 381.5 257.9 716.8 528.9 941.0 3665.6 29.5 18.6 20.8 20.0 23.2 18.0 17.9 13.6 11.6 8.6 14.4 15.6 12.6 21.8 11.0 15.9 8.2 8.1 871. 0 659.3 390.7 3548.1 4144.6 510.5 831.6 1342.2 50.6 104.9 313.0 319.3 737.1 573.1 907.6 3610.6 30.3 17.4 17.9 18.8 22.4 16.4 16.0 11.0 11. 7 7.7 13.1 20.2 12.6 24.7 12.2 17.0 7.7 7.3 1055.6 746.7 438.7 4120.5 4787.3 610.4 878.0 1488.4 55.6 94.1 345.2 279.0 718.3 721.1 1045.2 4028.5 13.7 12.7 29.2 14.1 12.4 14.6 18.2 12.5 12.4 9.9 8.3 6.2 10.1 11.4 9.1 13.6 7.3 10.2 6.3 5.7

14.1 691. 7 26.5 1907.6 1082.2 757.7 453.0 4226.9 4918.6 613.8 937.3 1551.0 59.5 93.1 354.8 291.7 739.6 755.6 1091.0 4196.7 12.7 29.9 14.0 12.3 14.5 18.1 12.5 12.3 10.1 8.2 6.3 10.3 11.8 9.0 13.7 7.4 10.4 6.7 5.9

43.7 691.7 26.5 1907.8 1082.6 758.2 453.4 4228.4 4920.1 613.7 937.3 1551.0 59.5 93.1 354.8 291. 7 739.5 755.6 1091. 0 4196.7

14.1 12.7 29.9 14.0 12.3 14.5 18.1 12.5 12.3 10.1 8.2 6.3 10.3 11.8 9.0 13.7 7.4 10.4 6.7 5.9

tv

C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11

Table D-4:

Estimates and Relative Errors for Each Estimator for Selected Survey Items in North Carolina.

OPEN SURVEY ITEM HI H2 H3 H4 H5 H6 H7 H8 H9 H10 C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 Cll Est.
(000)

CLOSED R.E.
(%;)

~

I

OPERATIONAL WEIGHTED Est.

ALTERNATIVE WEIGHTED Est.
(000)

CROPLA.'ID WEIGHTED
1

~~~O)~%~. (000) I R.E. I (%;)
154.4 14.8 169.2 9.9 797.8 475.4 398.1 836.5 2517.7 2686.9 473.7 123.6 597.3 38.0 80.2 75.3 16.2 171. 7 41. 7 373.6 1222.3 33.0 22.4 32.0 31.0 45.1 57.1 66.7 72.4 47.5 44.9 11.4 30.3. 11. 5 9.4 18.7 28.8 30.3 16.2 21.8 11.2 10.3 483.3 43.1 526.4 11.0 1416.2 1068.2 921. 5 758.4 4175.2 4701.6 434.7 192.3 627.0 39.1 77 .2 77.7 23.9 178.7 40.7 394.2 1279.7 39.6 42.4 39.8 24.3 34.4 41.8 48.0 55.6 42.7 42.0 8.3 23.0 9.5 8.1 11.8 23.1 25.9 12.0 14.1 8.5 8.6

I ~%~.
23.7 17.1 23.1 24.9 31.8 38.8 46.4 56.9 39.4 37.2 9.0 22.8 9.6 8.1 12.7 23.3 26.5 12.0 14.5 8.8 8.8

Est.
(000)

I R.E.
(%;)

CROPLAND WEIGHTED II Est. R.E.
(000)

I

(%;)

CROPLAND WEIGHTED III Est. ~.E. (000) (%)

I

90.1 11. 7 101.8 9.5 217.8 96.1 96.6 23.9 444.0 545.8

25.1 16.6 23.7 36.1 22.6 23.1 37.5 44.9 18.3 18.5 14.5 30.3 13.8 11.2 18.1 31.9 32.4 18.0 26.1 13.0 13.0

302.6 24.8 327.4 11.8 1057.0 695.7 570.0 400.1 2735.2 3062.7 442.7 185.6 628.3 38.9 79.9 75.5 23.6 179.0 40.3 393.8 1280.4

494.1 43.8 538.0 13.0 1441. 3 1071. 0 939.3 762.5 4227 .1 4765.1 377.6 175.2 552.8 36.5 80.6 63.2 12.0 155.7 31.6 356.2 1132.9

38.9 41.7 39.1 26.5 34.0 41.8 47.2 55.3 42.2 41.6 9.0 24.2 10.5 9.5 13.9 24.2 22.8 12.4 15.4 10.0 9.6

206.5 20.4 226.9 12.2 529.8 339.3 216.7 93.4 1191.4 1418.3 379.3 175.3 554.6 35.6 83.6 63.2 10.9 157.8 31.2 348.7 1127.8

19.8 15.2 19.2 27.7 20.1 25.2 25.1 29.2 21.0 20.5 9.7 24.2 10.8 9.8 14.5 24.2 24.3 12.5 15.7 10.4 9.9

248.5 23.1 271.6 13.8 990.3 619.4 487.4 903.2 3014.1 3285.7 392 .1 175.3 567.4 37.3 82.4 63.2 13.9 159.5 32.7 357.0 1153.8

22.6 16.9 22.0 26.7 35.5 44.5 54.9 67.1 39.7 36.8 8.8 24.2 10.3 9.5 13.6 24.2 24.2 12.2 15.2 9.9 9.4

!

368.2 152.6 520.8 32.3 72.9 76.2 16.4 165.4 40.6 301.6 1060.7

Table D-5:

Estimates and Relative Errors for Each Estimator for Selected Hog Survey Items in North Carolina with Seven Large Hog Operations Excluded. ALTERNATIVE WEIGHTED Est. (000) 158.6 17.0 175.7 9.8 418.9 251.1 155.7 65.6 901. 0 1076.7 CROPLAND \ffiIGHTED I R.E. Est. (000) (%) CROPLAND \ffi IGHTED II Est. R.E. (000) (%) CROPLAND WEIGHTED III Es t . JR. E . (000) (%) 169.9 18.2 188.1 11.7 440.1 263.3 164.2 68.2 947.4 1135.6 16.3 13.2 15.7 28.7 1'.7 20.5 22.9 29.7 16.2 15.6

"-

OPEN SURVEY ITEM s~ H1 H2 H3 H4 H5 H6 H7 H8 H9 HI0 Est. (000) 90.1 11.7 101. 8 9.5 217.8 96.1 96.6 23.9 444.0 545.8

CLOSED Est. (000) 91.1 10.9 102.0 7.7 245.2 140.5 98.3 26.7 518.3 620.3

I R.E. (%)
25.1 16.6 23.7 36.1 22.6 23.1 37.5 44.9 18.3 18.5

OPERATIONAL WEIGHTED Est. (000) 165.9 17.7 183.7 9.1 437.9 275.1 158.3 70.7 951.1 1134.8
(%)

T

R.E.

(%)
25.0 16.3 23.7 33.4 22.4 24.8 31.8 35.6 19.0 18.9

I R.E.
16.6 12.7 16.0 24.5 16.0 21.1 21.6 31.4 16.5 16.0

1 (%)

R.E.

I

I

16.2 12.9 15.6 26.0 17.6 19.2 21.8 28.0 15.4 15.0

168.4 18.0 186.4 10.9 440.4 262.5 163.8 68.2 945.9 1132.3

16.3 13.2 15.7 28.4 17.7 20.5 23.0 29.7 16.2 15.7

167.3 18.0 185.3 11.7 422.8 254.5 160.4 68.2 917.6 1102.9

16.5 13.2 15.8 28.7 17.9 20.9 23.5 29.7 16.5 16.0

Table

D-6:

Estimates

and Relative

Errors

for Each Estimator

for Selected

Survey

Items

in Ohio.

OPEN SURVEY ITEM ~ HI H2 H3 H4 H5 H6 H7 H8 H9 HID C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 Est.
(000 )

CLOSED R.E. (%) Es

I

(000)

t.,
5.4

OPERATIONAL WEIGHTED Est. (000) 195.7 13.4 209.1 5.6 672.7 371. 5 252.4 163.2 1465.3 1674.4 344.2

ALTERNATIVE WEIGHTED Est. (000) 192.6 13.4 205.9 5.3 652.3 370.4 243.1 160.3 1431.5 1637.4 352.9 404.3 757.1 38.0 54.8 166.1 87.3 308.1 242.1 494.3 1839.6

R.E • (%) 29.0 24.0 28.3 51.0 26.5 31.2 38.3 39.8 22.5 22.5 13.0 16.8 11.3 15.1 17.6 16.7 19.9 11.4 30.3 11.8 9.5

I R.E. (%)
17.6 14.9 17.4 29.7 17.5 16.5 18.7 25.2 15.4 15.2 9.7

I

R.E. (%)

CROPLAND WF.:IGHTED I Est. R.E. (000) (%)

CROPLAND
l-lE IGHTED

I

II Est. (000) 200.9 13.8 214.7
L..9

1

R.E.
(%)

CRnPLAim WEIGQTED III Est. 'R..E. (000) (%)

I

I

200.9 14.4 215.3 2.7 600.9 451. 5 326.6 171.1 1552.7 1768.0 300.7 462.7 763.3 38.2 51. 8 179.5 177.2 408.5 365.2 479.1 2054.3

27 .4 23.0 27 .0
43.7

161. 8 12.8 174.6

17.8 15.2 17.5 28.9 17.3 17.0 18.9 26.0 15.3 15.1 9.5 11.5 7.5 9.8 13.7 13.6 29.4 11. 7 15.6 9.5 7.4 I

201. a 13.8 214.8 4.9 673.1 377 .6 250.0 163.2 1468.7 1683.5 335.9 397.5 733.4 37.4 57.3 159.4 81. 5 298.2 235.4 483.0 1787.4

18.8 16.3 18.6 30.0 17.8 17.4 19.1 26.3 15.7 15.6 10.6 12.1
8.0

18.8 16.3 18.6
'3() • ()

200.9 13.8 214.7 4.9 673.1 377 .6 249.4 164.2 1469.3 1683.9 330.9 397.5 728.4 37.1 55.1 159.5 83.1 297.7 237.7 479.1 1779.9

18.8 16.3 18.6 30.0 17.8 17.4 19.1 26.1 15.7 15.6 10.7 12.1
8.0

24.1 28.9 28.5 53.1 23.5 23.5 14.9 18.0 12.6 13.2 25.8 20.7 58.1 28.4 32.3 12.3 15.5

389.7 262.4 200.5 70.3 928.3 1102.9 356.2 475.2 831.4 37.2 43.5 156.8 69.0 269.3 204.8 480.9 1823.6

673.1 377 .6 249.4 163.0 1468.0 1682.7 315.0 397.5

17.8 17.4 19.1 26.3 15.7 15.6 10.7 12.1
o IJ
•. .L

407.') 11.2 751. 7 36.4 54.7 166.5 84.9 306.1 238.6 488.9 1821. 8 7.4 9.6 14.0 13.3 27.8 11.2 15.3 9.6 7.3

." I..LL~'"+
')

I.

1

11.5 18.5 13.8 30.5 12.0 15.8 10.0 7.8

37.1 53.6 159.3 81. 3 294.2 234.5 466.3 1744.5

11.6 19.5 13.8 30.6 12.2 15.9 10.1 7.9

11.6 19.1 13.8 30.0 12.1 15.9 10.1 7.8

I

Cll

Table D-7: Estimates and Relative Errors for Each Estimator for Selected Survey Items in the Five States Combined (Excluding Hog Data from Seven Hog Operations in North Carolina). CROPLAND WEIGHTED I Est. R.E. (000) en CROPLAND HEIGHTED II Est. R.E.
(000)

OPEN ~ SURVEY ITEM HI H2 H3 H4 H5 H6 H7 H8 H9 HID
N \0

CLOSED R.E.
(%)

Est.
(000 )

I

Est . (000) 1147. 9 88.5 1236.4 48.2 3189.9 1793.0 1331. 4 647.6 7010.1 8246.5 3313.7 1750.7 5064.4 266.1 460.9 702.6 509.9 1673.5 1146.2 3441.0 11591. 2

I

OPERATIONAL WEIGHTED Est. (000) 1373.0 105.6 1478.7 61.5 4026.4 2330.1 1589.3 928.4 8935.6 10414.3 3292.1 1838.7 5130.8 263.0 494.2 739.6 507.8 1741.7 1338.9 3513.8 11988.1

ALTERNATIVE WEIGHTED Est. (000) 1367.2 105.2 1472.4 59.7 4026.1 2328.1 1584.9 930.7 8929.5 10401.9 3292.4 1882.3 5174.7 267.5 487.8 753.4 513.5 1754.8 1364. 2 3555.1 12116.2

R .E •
(%)

I ~%~.
7.1 6.6 7.0 17.2 7.7 7.5 8.9 11.3 7.1 6.9 3.5 5.9 3.0 3.6 4.8 7.0 9.5 4.4 7.1 3.2 2.9

I

R.E.
(%)

I

I

(%)

CROPLAND WEIGHTED III Es t • fl. E • (000) (%)

I

1283.7 98.1 1381.8 49.0 3652.<i 2218.2 1512.4 749.0 8181.0 9562.8 2715.2 1870.9 4586.0 240.7 429.0 797.3 540.2 1766.6 1179.2 3079.2 11.0851.7

11.4 9.9 11.2 17.8 11.1 14.0 13.4 18.8 11.1 11.0 5.9 8.4 4.8 5.4 8.6 12.0 22.2 9.2 13.2 4.7 5.0

9.7 8.5 9.5 18.3 10.4 11.6 13.4 15.8 10.1 9.8 4.9 8.2 4.2 5.5 7.6 8.9 16.6 6.6 11.4 4.3 3.8

7.2 6.8 7.1 16.1 7.8 7.4 8.9 11.4 7.1 7.0 3.6 5.9 3.1 3.7 4.8 7.0 9.8 4.4 7.2 3.2 2.9

1380.8 107.0 1487.8 60.9 4040.9 234~.8 1599.3 944.6 8995.5 10483.2 3075.0 1857.9 4932.9 254.7 479.5 726.1 506.1 1711. 7 1301. 2 3394.0 11594.5

7.3 7.0 7.2 16.5 7.9 7.6 9.2 11.6 7.2 7.1 4.0 6.0 3.4 4.0 5.6 7.3 10.0 4.7 7.2 3.6 3.2

1378.1 106.9 1484.9 61.0 4016.0 2344.4 1593.5 919.5 8934.4 10419.3 2965.8 1858.3 4824.1 247.6 466.2 719.9 494.9 1680.9 1284.4 3300.5 11337.5

7.3 7.0 7.2 16.6 7.9 7.6 9.2 11.6 7.3 7.1 4.1 6.0 3.4 4.1 5.8 7.4 10.2 4.8 7.3 3.6

1380.7 107.1 1487.9 61.8 4042.0 2347.2 1603.3 951. 3 9005.7 10493.5 3084.6 1858.0 4942.6 254.7 479.3 731. 9 518.2 1729.4 1337.9 3389.1

7.3 7.0 7.2 16.4 7.9 7.6 9.1 11.6 7.2 7.1 4.0 6.0 3.4 4.0 5.7 7.3 10.2 4.8 7.3 3.6 3.3

C1 C2 C3 C4 C5 C6 C7 C8 C9 CI0 C11

3.3 111653.8

Table D-8:

Estimates and Relative Errors for Each Estimator for Selected Survey Items in Georgia.

OPEN SURVEY ITEM ~ H11 H12 H13 H14 H15 H16 H17 H18 H19 H2O H21 H22 C12 C13 C14 CIS C16 C1l CI8
Fl

OPERATIONAL WEIGHTED R.E. (%) 32.0 23.7 27.2 24.1 30.6 26.6 24.4 30.6 26.7 26.5 28.8 26.2
-

ALTERNATIVE WEIGHTED Est. (000) 96.4
I

Est. (000) 147.7 112.5 260.2 108.2 147.8 256.0 809.7 1053.8 1863.5 35.0 55.7 90.7 9.8 254.8 519.1 773.9 15.0 30.8 45.8 53.3

I

Est. (000) 101.9 99.3
\

I

R.E. (%) 16.6
~~.4
'') I

I

R.E. (%) 16.4 12.6 13.8 12.8 14.2 13.1 12.2 14.0 12.7 66.3 54.8 59.9 36.9 10.0 8.4 8.1 12.9 13.9 12.3 5.4
I

CROPLAND HEIGHTED I Est. R.E. (%) (000)

I

CROPLAND WEIGHTED III .R.E. Est. (000) (%)

I

90.6 93.6 184.2 92.0 109.4 201.4 666.9 770.5 1437.4 83.9 102.8 186.7 11.6 325.0 589.8 914.8 16.9 33.5 50.4 63.0

14.9 12.3 12.8 12.3 13.5 12.4 12.1 13.6 12.3 70.2 57.9 63.3 48.8 10.5 9.6 9.0 14.6 14.4 12.8 5.5 I

90.4 93.2 183.h 91.1 108.4 199.5 662.2 761.4 1423.6 83.5 103.9 187.4 10.7 311. 7 539.6 851. 2 14.1 30.8 44.9 59.3

14.9 12.3 12.13 12.4 13.6 12.5 12.2 13.8 12.4 70.6 58.1 63.4 52.0 11.2 9.8 9.4 15.9 15.5 13.9 5.7

I

96.6 193.0 96.8 114.4 211.2 700.2 800.6 1500.8 78.4 96.1
174.5

201.2 100.5 122.7 223.3 728.3 855.6 1584.0 81.0 98.4 179.4

13.8 13.0 14.7 13.4 12.4 14.6 13.0 66.6 55.5 60.4 34.4 9.1 8.0 7.5 12.7 12.6 11.4 5.2

I

i

71.9 15.2 13.8 13.2 18.1 18.3 16.3 6.5

I

13.5 358.7 589.5 948.2 17.1 34.2 51.3 63.2

13.4 340.8 570.7 911. 5 15.4 32.5 47.9 62.1

Table D-9:

Estimates and Relative Errors for Each Estimator for Selected Survey Items in Kentucky.

OPEN SURVEY ITEM ~ Hll Hl2 H13 H14 HIS H16 Hl7 H18 H19 H2O
I
W

OPERATIONAL WEIGHTED R.E. (%) 22.0 23.0 21.1 23.7 24.3 22.4 24.5 24.6 22.9 42.6 36.1 34.5 59.2 10.1 8.7 8.3 16.5 10.1 9.8 5.2 Est. (000) 78.9 73.0 151.9 66.0 64.2 130.1 473.4 443.7 917.0 9.5 20.5 30.0 17.8 384.5 852.5 1237.0 15.2 45.9 61.1 99.7

ALTERNATIVE WEIGHTED Est. (000) 76.9 72.1 149.0 66.2 62.8 129.0 474.8 434.7 909.5 9.4 20.7 30.1 18.1 394.0 878.4 1272.4 15.4 47.5 62.9 101.4

Est. (000) 69.9 54.3 124.2 50.8 53.9 104.7 364.9 367.2 732.1 10.1 17.3 27.3 13.9 300.5 735.1 1035.6 10.2 36.9 47.2 84.7

I

I

R.E. (%) 13.9 13.0 12.5 13.2 14.7 12.9 13.5 15.2 13.3 28.7 22.9 19.8 40.8 8.2 4.9 4.9 9.9 6.9 6.5 4.0

I

R.E.
(%) 14.1 13.3 12.8 13.4 15.1 13.2 13.8 15.6 13.7 29.6 23.2 20.3 40.3 8.1 5.0 4.9 10.1 6.9 6.5 4.0

CROPLAND WEIGHTED I Est. R.E. (000) (%)

I

CROPLAND WEIGHTED II R.E. Est. (%) (000)

I

78.2 71.7 149.9 66.7 64.6 131.3 486.0 450.7 936.7 10.7 20.4 31.0 15.3 356.5 744.2 1100.7 15.3 43.4 58.7 95.8

15.1 15.1 14.5 15.0 16.8 15.2 15.8 18.0 16.3 30.9 24.6 22.5 45.6 9.7 5.5 5.6 13.1 8.5 8.6 4.4

76.8 71.5 148.3 66.5 64.6 131.1 484.6 450.7 935.3 10.5 20.0 30.5 15.3 351.8 728.6 1080.4 14.6 42.2 56.9 94.0

15.3 15.1 14.7 15.0 16.8 15.2 15.8 18.0 16.3 31.5 25.0 23.0 45.6 9.9 5.5 5.6 13.6 8.6 8.8 4.4

H21 H22 C12 Cl3 C14 CIS C16 C17 Cl8
Fl

""'

Table D-10:

Estimates and Relative Errors for Each Estimator for Selected Survey Items in ~innesota.

OPEN SURVEY ITEM Hll H12 H13 H14 HIS HI6 H17 HI8 H19 H2O
w
N

OPERATIONAL WEIGHTED Est. (000) R.E. (%) 13.6 13.1 12.'1 13.2 13.5 13.0 14.2 15.1 14.3 22.7 21.5
?" _.1..4 •

ALTERNATIVE WEIGHTED Est. (000) 308.3 285.3 593.7 300.1 258.3 558.5 2335.7 1954.7 4290.4 63.4
Ell. 2

CROPLAND HEIGHTEn I Est. (000 309.5
:287.Q

CROPLA..~D WE rr~HTED II Est. (000

I

281. 3 "-'" -.1
_ / j.

18.5
l':! • .c

299.6 :2 76.(, 576. ; 290.8 251. 5 542.3 2263.5 1903.9 4167.4 60.4 78.1 138.5 653.9 586.3 921.3 1507.7 20.2 84.7 104.9 107.8

.~ "

13.6 13.1
L~. 8

13.7 12.9
12.8

309.5 287.9 ')97.3 304.3 261. 0 565.3 2363.1 1973.5 4336.6 63.8 82.9 146.7 673.5 616.1 956.5 1572.6 21.6 90.4 112.0 110.9

13.7 ]2.9 12.H 13.1 13.5 13.0 14.0
15.0

554 . .4

:8.5

')97. 3 304.3 261. 0 565.3 2363.1 1973.5 4336.6 63.8 82.9 1411.7 673.5 615.7 955.4 1571.2 21.6 90.4 112.0 1l0.9

275.2 252.8 528.0 2155.0 1882.5 4037.5 58.5 69.6 128.1
--,

18.5 20.1 18.8 19.2 21.6 20.0 24.7 25.8 24.9 24.3 10.4 8.8 8.8 17.2 11.9 11.0 5.5

13.2 13.5 13.0 14.1 15.0 14.2 22.3 21.2 21.1 13.5 7.0 7.0 6.1 10.6 9.1 8.3 3.8

13.1 13.5 13.0 14.0 15.0 14.1 22.3 21.3
21.0

14.1 22.3 21.3 21.[)
,

H21 H22 C12 C13 C14 CIS C16 C17 C18
Fl

144.n 670.b 608.0 947.9 1555.9 20.7 87.9 108.7 109.8

!

466.5 632.8 811.7 1444.5 20.0 79.8 99.9 104.5

13. ") 7.0 7.0 6.1 10.5 9.1 8.3 3.8

13.4 7.2 7.1 6.3 10.4 9.2 8.3 3.9

13.'" 7.2 7.1 6.3 10.4 9.2 8.3 3.9

Table D-ll:

Estimates and Relative Errors for Each Estimator for Selected Survey Items in North Carolina.

OPEN SURVEY ITEM ~ Hll H12 H13 H14 H15 H16 H17 H18 H19 H2O H21 H22 C12 C13 C14 CIS C16 C17 C18 Fl Est. (000) 41. 7 34.9 76.6 36.5 50.6 87.1 262.0 347.1 609.1 5.5 14.9 20.4 0.0 204.7 282.1 486.8 5.2 11.9 17.1 91.4

OPERATIONAL WEIGHTED R.E. (%) 28.9 21. 7 23.5 25.0 35.2 28.7 25.2 41. 8 32.5 35.3 48.9 39.4 71.3 15.8 14.9 14.1 27 .1 25.2 21. 7 6.4 Est. (000) 222.5 218.4 440.9 204.7 231. 9 436.7 1655.2 1794.2 3449.5 52.5 136.7 189.2 6.3 256.7 327.0 583.7 6.2 19.1 25.3 109.3

ALTERNATIVE WEIGHTED Est. (000) 133.0 127.1 260.1 115.7 140.0 255.7 950.7 1072.0 2022.7 43.9 132.4 176.3 6.0 252.2 333.2 585.4 6.0 18.9 24.9 107.6

I

I

R.E. (%) 42.5 43.2 42.8 45.5 41.9 43.5 45.5 43.7 44.5 40.1 55.6 48.2 38.7 12.6 9.1 9.8 18.3 18.6 17.0 5.4

I

R.E. (%) 24.7 25.2 24.8 24.7 27.9 26.2 26.6 31.0 28.8 42.8 57.3 51.3 36.8 12.6 9.7 10.0 17.5 17.7 16.1 5.4

CROPLAND WEIGHTED I Est. R.E. (000) (%)

I

CROPLAND WEIGHTED II Est. R.E. (%) (000)

I

226.1 221. a 447.1 208.9 234.0 442.8 1686.8 1818.7 3505.5 56.4 136.1 192.6 4.2 240.7 288.1 528.7 4.5 18.2 22.7 106.9

41.9 42.7 42.3 44.7 41. 5 42.9 44.7 43.1 43.8 37.9 55.8 47.4 30.4 13.3 10.8 11.1 17.3 22.9 19.9 5.7

85.6 78.2 163.8 73.2 79.9 153.2 557.7 551. 7 1109.4 18.7 35.9 54.6 4.2 234.9 291. 9 526.8 4.3 17.3 21.6 104.0

18.5 16.4 16.9 18.0 17.9 17.4 19.1 18.8 18.5 29.4 27.4 25.6 30.4 13.8 11.0 11.4 17.5 23.7 20.6 5.8

Table D-l2:

Estimates and Relative Errors for Each Estimator for Selected Hog Survey Items in North Carolina with Seven Large Hog Operations Excluded. CROPLAND WEIGHTED I Est. R.E. (000) (% 77 .0 73.8 150.7 67.5 74.0 141. 5 503.4 514.1 1017.5 34.0 1n.4
50.4

OPEN SURVEY ITEM Hll H12 H13 H14 Hl5 H16 HI? H18 H19
w
.j:-.

OPERATIONAL WEIGHTED R.E. (%) 28.9 21. 7 23.5 25.0 35.2 28.7 25.2 41. 8 32.5 35.3
48.9
•.••• ,,\ 1

ALTERNATIVE WEIGHTED Est.
(000)

Est. (000) 41. 7 34.9 76.6 36.5 50.6 87.1 262.0 347.1 609.1 5.5 14.9
20.4

Est.
(000)

R.E. (%) 17.8 16.4 16.5 17.0 18.3 17.0 17.6 18.2 17.4 26.4 27.0
2.3.8

R.E. (%) 17 .1 16.0 16.0 17.7 16.8 16.5 18.5 18.4 17 .9 27 .5 31.1
25.R

CROPLAND WEIGHTED II Est. R.E. (000) %) 77.2 72.6 149.7 66.5 73.5 140.0 495.9 511.0 1006.9 34.0 17.0 51.0 17.4 16.6 16.4 18.1 17.5 17.2 18.9 18.9 18.5 28.4 29.7 26.1

76.1 72.7 148.8 65.4 74.0 139.4 490.9 501.3 992.1 35.3 13.1
'0 '-f'..Ja / ....•

72.8 68.4 141.2 63.1 68.9 132.0 477 .6 477.3 954.9 32.3 13.9
46,7

17 .5 16.3 16.3 18.0 17.4 17.0 18.8 18.8 18.3 28.4 30.3
26.3

H2O
H2l H22

)'1.4

Table D-13:

Estimates and Relative Errors for Each Estimator for Selected Survey Items in Ohio.

OPEN SURVEY ITEM ~ H11 H12 H13 H14 HIS H16 H17 H18 H19 H2O H21 H22 C12 Cl3 C14 C15 C16 C17 CIa Fl Est.
(000)

OPERATIONAL WEIGHTED R.E. (%) 29.2 27.4 27.2 26.2 27.3 26.4 26.8 30.5 28.2 29.7 24.3 25.4 49.3 15.3 12.3 12.7 21. 3 17.8 16.2 5.4 Est. (000) 100.7 90.8 191. 4 93.5 88.6 182.0 761.9 653.5 1415.4 13.5 23.6 37.1 238.9 347.3 391. 4 738.7 12.5 29.6 42.2 88.8

ALTERNATIVE WEIGHTED Est. (000) 98.0 89.1 187.2 90.8 87.1 177.9 741.7 642.4 1384.1 13.2 23.3 36.5 244.1 347.8 396.8 744.6 12.5 28.6 41.1 89.3

I

I

R.E. (%) 19.6 18.2 18.4 17.9 18.1 17.7 18.5 18.7 18.3 19.2 22.6 20.4 22.7 9.6 8.3 7.9 14.0 11.3 10.4 4.3

I

R.E. (%) 19.7 17.9 18.3 17.9 18.0 17.6 18.6 18.7 18.3 19.6 23.2 20.9 23.6 9.7 8.2 7.9 14.6 11.6 10.6 4.4

CROPLAND WEIGHTFD I Est. R.E. (000) (%)

I

CROPLAND WEIGHTED 1I Est. R.E. (000) (%)

I

107.9 71.8 179.7 76.2 81.3 157.5 652.9 649.3 1302.3 19.1 28.2 47.3 425.9 341.1 388.9 730.0 12.1 38.9 51.0 82.4

102.1 93.4 195.5 93.7 92.2 185.9 768.5 682.7 1451.2 13.0 23.0 36.0 236.7 326.6 384.3 710.9 12.5 28.5 41.0 88.8

20.8 18.6 19.4 18.8 18.8 18.5 19.5 20.2 19.5 20.2 23.8 21. 5 24.0 10.0 9.1 8.4 14.7 11.9 11.0 4.7

102.0 93.4 195.5 93.7 92.1 185.8 768.5 682.2 1450.7 13.0 23.0 36.0 237.2 322.5 366.6 689.1 12.0 27.6 39.6 88.1

20.8 18.6 19.4 18.8 18.8 18.5 19.5 20.2 19.5 20.2 23.8 21.5 23.9 10.1 9.0 8.4 15.1 12.0 11.2 4.6

Table D-14: Estimates and Relative Errors for Each Estimator for Selected Survey Items in the Five States Combined (Excluding Hog Data from Seven Hog Operations in North Carolina). OPERATIONAL WEIGHTED R.E. (%) 12.3 11. 7 11. 5 11. 4 12.8 11.6 11. 9 13.5 12.3 14.4 14.6 13.8 26.1 5.9 4.9 4.9 8.9 7.0 6.4 2.6 Est.
(000 )

OPEN SURVEY ITEM H11 H12 H13 H14 HIS H16 H17 H18 H19 H2O
w

ALTERNATIVE WEIGHTED Est. (000) 652.5 611.6 1264.0 617.1 591. 6 1208.6 4730.1 4309.6 9039.7 253.6 178.3 431. 9 952.3 1942.8 3127.1 5069.8 70.1 215.4 285.5 470.2

~

Est. (000) 648.6 546.7 1195.2 546.8 586.4 1133.3 4244.6 4299.9 8544.5 185.6 128.1 313.8 916.5 1734.0 2736.9 4470.8 62.7 198.3 261. 0 416.2

I

I

R.E. (%) 7.8 7.2 7.2 7.5 7.5 7.2 8.0 8.2 7.9 22.8 31.5 26.1 11.1 4.0 3.2 3.1 5.6 5.0 4.5 2.1

I

R.E. (%) 7.9 7.3 7.3 7.6 7.5 7.3 8.1 8.2 8.0 22.3 30.4 25.4 11.3 4.0 3.3 3.1 5.6 5.0 4.5 2.1

CROPLAND WEIGHTED I Est. R.E. (%) (000)

I

CROPLAND WEIGHTED II R.E. Est. (000 ) (%)

I

657.2 612.3 1269.5 616.1 600.9 1217.0 4718.0 4358.0 9075.9 255.9 177 .5 433.4 930.3 1933.5 3081. 8 5015.3 71.3 213.5 284.8 468.8

657.4 620.4 1277.7 624.2 601.2 1225.4 4788.0 4391.5 9179.5 263.1 187.8 450.9 941.1 1864.5 2961.8 4826.3 70.9 214.0 284.8 465.4

R.O
7.4 7.4 7.6 7.5 7.4 8.2 8.3 8.0 24.1 32.5 27.4 11.4 4.3 3.6 3.4 6.2 5.4 4.9 2.2

655.9 618.6 1274.4 622.1 599.6 1221. 6 4774.3 4378.7 9153.1 263.8 187.8 451.6 940.9 1836.8 2883.2 4720.0 66.6 208.4 274.9 456.3

R.O
7.4 7.4 7.7 7.5 7.4 8.2 8.3 8.1 24.3 32.5 27 .4 11.4 4.4 3.6 3.5 6.3 5.5 5.0 2.2

en

"
Z

'"

H21 H22 C12 C13 C14 CIS C16 C17 C18 Fl

c:

"
0 < m

~
E
m

..•
z
"tl

~

..•
Z Z
0

" ...
i'5

.•.
I

m

~
to>

:;;

g: .0 ~

''"f


				
DOCUMENT INFO