Gambar halaman
PDF
ePub

thirteenth. This decrease was offset by raising the sampling rate for the intermediate size mills from 10 percent to 20 percent. Better variance and cost data than were available for 1949 indicated that these revisions would scarcely alter the lumber sampling errors,

but would appreciably lower the total cost of the sawmill canvass.

individual industry (4-digit), and multi-units were arrayed roughly by the employment of their largest plant, within the sample classes. The sample companies were then chosen by beginning with a random start and continuing with every second, fifth, tenth, etc. company thereafter, depending upon the specified rate.

The sample so obtained for 1949 was extensively revised for the 1950 annual survey, except for the multi-unit and supplementary 1947 BOASI portion of the sample. Important changes were made for independents, new companies, and sawmills,

For independents from the 1947 Census list, new sampling rates were specified on an individual industry basis, taking account of the relative importance of the industry, its 1947 size distribution, and preliminary average estimates of the 1949 sampling variances. The pattern of sampling rates used was similar to that for the 1949 Survey, and again decreased with decreasing establishment size.

No changes were made for 1950 in the specifications for the multi-units because they accounted for such a minor part of the sample below the certainty class that changes in their sampling rates would have had little effect on the reliability of the estimates. Nether were any changes made in the sample from the supplementary 1947 BOASI list, in this case, because of cost and related administrative factors.

An important problem encounted in conducting a sample survey is the provision of adequate controls on the sample to insure that all establishments chosen, and only those establishments, are included in the survey tabulations or are properly accounted for if they are excluded. In the annual surveys, the sample control operations consisted of a verification that all establishments selected for the sample from the many sources were accounted for by one of several possibilities:

Changes in the composition of the panel were minimized. Where the 1949 and 1950 sampling rates were the same for an industry, no changes were made. Where the 1950 rates were lower than those for 1949, a random systematic sub-sample was drawn from the 1949 list. Where the 1950 rates were higher than those for 1949, more units were selected at random from each designated industry-employment cell, and added to the 1949 list. The sample was reduced for about 140 industries, expanded for about 90, and kept intact for 230. In total, nearly the same numbers of establishments were added and deleted.

(1) A report was received and tabulated. (2) The establishment was identified as “Out

of business''or "Inactive. (3) The establishment was classified as

"Non-manufacturing" or "Out of
scope. "

On the other hand, it

it was necessary to verify that establishments not selected for the sample were excluded from the tabulations. (In this class of cases were included not only mailing and collection errors but also the whole group of “successor" problems.)

The sample of companies that began operations after 1947 was completely redrawn for 1950. The rates were based on the revised rates for independents from the 1947 Census lists, but were, in general, one notch higher on the sampling scale. Fifty percent of the new companies were selected from the industry-employment cells for which a 20 percent rate was used in sampling the Census independents, 20 percent of the new companies were selected from cells for which the corresponding Census rate was 10 percent, etc. The new companies were sampled at these relatively higher rates in order to reduce their disproportionately high contributions to the sampling variances.

Although partial controls on the sample were provided at several stages in the processing of the 1949 Annual Survey, the main sample control operation was performed as part of the final tabulation review and consisted only of the elimination of significant biases resulting from errors in coverage. The problem of sample control was further complicated for the 1950 Annual Survey because of the numerous changes

which were made to improve the sample design in important industries. The sample control operation was again restricted to the elimination of significant biases.

The Eastern sawmill sample also was revised for 1950. A subsample of the 1949 area segments was selected reducing the rate to onetwentieth, compared with the 1949 rate of one

[blocks in formation]

The sums of the squares of the inflated differences between the 1949 and 1947 figures for each sample establishment, except the certainty establishments, were used as approximations for the variances of the estimated totals.

E. Inventories, fuels and electric energy:

B. Lumber estimates (industry 2421):

The difference estimating formula was not used for this industry. Accordingly, the sums of the squares of the weighted deviations around the sample mean in each sample class cell were used as the estimate of variance for production and stocks of rough lumber; the sums of squares of the weighted 1949 figures were used for all other items.

The variances of these items were derived for a random systematic sample of 20 industry groups by calculating the sums of squares of the weighted figures for each sample establishment. The ratios of the standard errors of these items to their corresponding value added errors were then computed, and the simple (unweighted) mean of these ratios found. Standard errors for all the inventories, fuels, and electric energy items were

derived by multiplying the corresponding value-added-by-manufacture standard error by the mean ratio. F. Size tables:

Most of the sample estimates in the size tables

difference estimates for which the variances were computed as described above for the general statistics. For scattered cells, where only the 1950 data were used in deriving the estimates for a size class, the variance approximation consisted of the sums of squares of the weighted 1950 figures.

are

C. Expenditures for plant and equipment:

The sums of squares of the weighted 1949 figures for each sample establishment were used as the estimate of variance.

173

APPENDIX

С

PROVISIONS OF LAW RELATING TO MANUFACTURES CENSUSES AND SURVEYS

(From Fifteenth Census Act, approved June 18, 1929 ; 46 Stat. 21, 18 U.S.C. 210, 211, 217]

[ocr errors]

SECTION 17.—That the Director of the Census be, and any person violating the provisions of this secand he is hereby, authorized and directed to collect tion by refusing or willfully neglecting to answer any and publish, for every second year after 1927, sta- of said questions shall be guilty of a misdemeanor, tistics of manufacturing industries; and the director and upon conviction thereof shall be fined not exis hereby authorized to prepare such schedules as in ceeding $500, or imprisoned for a period not exceedhis judgment may be necessary.

ing 60 days, or both so fined and imprisoned, and SECTION 10.—That it shall be the duty of every

any person violating the provisions of this section by

willfully giving answers that are false shall be fined owner, official, agent, person in charge, or assistant

not exceeding $10,000 or imprisoned for a period not to the person in charge, of any company, business,

exceeding one year, or both. institution, establishment, religious body, or organization of any nature whatsoever, to answer completely and correctly to the best of his knowledge all SECTION 11.-That the information furnished questions relating to his respective company, busi- under the provisions of this act shall be used only ness, institution, establishment, religious body, or for the statistical purposes for which it is supplied. other organization, or to records or statistics in his No publication shall be made by the Census Office official custody, contained on any census schedule whereby the data furnished by any particular estabprepared by the Director of the Census under the lishment or individual can be identified, nor shall the authority of this act, or of the act to provide for a Director of the Census permit anyone other than the permanent Census Office, approved March 6, 1902, or sworn employees of the Census Office to examine the of acts amendatory thereof or supplemental thereto; individual reports.

[From Public Law 671-80th Congress ; 62 Stat. 478, 13 U.S.C. 121)

AN ACT

To provide for the collection and publication of statistical information by the Bureau of the Census

SECTION 1.-That (a) the Director of the Bureau immediately preceding the taking thereof; Provided, of the Census, hereinafter referred to as the Director That the census of manufacturers shall not be taken and the Bureau, respectively, is authorized and di- in 1949. The censuses herein provided for shall in

, rected to take, compile, and publish the censuses of clude the United States and its Territories and such manufacturers, of mineral industries, and of other possessions as may be determined by the Director businesses, including the distributive trade, service with the approval of the Secretary of Commerce. establishments, and transportation (exclusive of (6) That the Director is further authorized to means of transportation for which statistics are re- make such surveys as are deemed necessary to furquired by law to be filed with a designated regulatory nish annual and other interim current data on the body), in the year 1949 and every fifth year there- subjects covered by the censuses provided for in this after, and each such census shall relate to the year and other Acts.

174

[blocks in formation]

Your

THIS INQUIRY IS AUTHORIZED BY ACT OF CONGRESS (approved June 19, 1948, 62 Stat. 478; 13 U. S. C. 121-123), WHICH REQUIRES THAT YOU FILE A REPORT. Your report is accorded confidential treatment, subject to the provisions of law. Census report cannot be used for purposes of taxation, investiga. tion, or regulation. Please complete and return this form to the Bureau of the Census office shown on the enclosed return envelope. Retain the green copy for your files. Betore answering each question, please read enclosed instructions. Compare each entry with your 1949 report, if any.

[blocks in formation]

MA-100

[blocks in formation]

ITEM

3

(Give date)

Name

4. Sold by you

[ocr errors]

Street and number

(Give date)

INVENTORIES: Report the value of all inventories wherever located, for the account of this establishment. Include goods owned by you, but held by others. Do not include goods held by you, but owned by others.

[blocks in formation]
[blocks in formation]

ITEM

(Specify date above)

(Omit cents)

(Specify date above) (Omit cents)

A. Inventories of finished products.

B. Inventories of materials, supplies, work in process, fuel, etc...

C. Total inventories (Sum of A and B).

CAPITAL EXPENDITURES AT THIS ESTABLISHMENT DURING 1980 (Exclude land):

A. New structures and additions to plant.. $.

B. New machinery and equipment.

C. Expenditures for used plant and used equipment acquired from others........

D. Total expenditures (Sum of A, B, and C). $.

A. SALES of products bought and resold without further manufacture (or processing) in this establishment....

B. COST of products bought and resold without further manufacture (or processing) in this establishment...

Total, 1980

(PLEASE CONTINUE ON REVERSE SIDE)

EXPENDITURES (Omit cents)

(Omit cents)

(Omit cents)

175

ITEM 10

PRODUCTS OF THIS ESTABLISHMENT SHIPPED (Including interplant transfers): Examine the enclosed reference list to determine in which class (or classes) the products you made in 1950 belong. For each such class, enter in column B the code number given on the reference list and in column C the total 1950 value of shipments for all products of this establishment falling within the product class. Then on the same line, in column A, describe in your own words the products you included in that class. If some of your products do not seem to fall within any of the classes of prod ucts given, list each such product separately in column A and

DESCRIPTION OF PRODUCTS
(A)

enter in column C the value of shipments. If more space is needed, use the "Remarks" section or additional sheets.

The values reported should represent values f. o. b. plant after discounts and allowances, and should not include freight charges or excise taxes. Transfers of products to other plants of the same company should be included at approximate mar ket value in the appropriate class of products.

The figures reported in this item should cover all shipments or receipts of this establishment in 1950, except for products resold in the same condition as when purchased, which should be included in item 9A.

[blocks in formation]
[blocks in formation]

ITEM METALS CONSUMED IN 1950: Report the consumption during 1950 of the specified metals for all purposes (i. e. include 11 consumption for production of fabricated products, for maintenance, repair, and operating purposes, and for construction and/or alteration of plant facilities). Include as "mill shapes" all rolled, drawn, and extruded products such as bars, sheet, strip, wire, plates, pipe, tubing, structural shapes, tinplate, extruded shapes, etc.; include as "castings" all rough (not machined) castings as received from foundry.

[blocks in formation]
« SebelumnyaLanjutkan »