completeness checks are to be identified (see exhibit V-4-1). A critical item (or items) missing is a reason for rejection. The item number is to be circled and referred to as flagged. Since the NSO processing standards do not permit any formula for estimation of entries for critical items in the case of establishments with 10+ TEP (total engaged persons), the respondent must be asked to correct the deficiencies.

In the manual editing, when no total has been reported for an item by the respondent, the procedure calls for deriving that total by adding the reported detail. In the process,

there is a check made as to whether the item's last detail entry is the total of the item's previous detail entries. If the last detail entry is not the total, the sum of all detail should be entered on the required total line. If the last detail, the editor-coder should delete it and transfer that total to the

passed sequence A, which is the test for completeness. The short form questionnaires, however, may have passed only a portion of sequence A.

4.1 Step B-1: Physical location of establishment

The entries in item 2, physical location of establishment, were checked and the appropriate codes entered in boxes 1 through 4 and 10 through 15 by the Control Operations Branch of DPD. Also, the item 6 total (TEP) was checked by the Control Clerk in MMD for the skip-list establishments and in DPD for those not on the skip list to make sure the form qualifies as a long form.

4.2 Step B-2: Form of ownership

If either box has been checked in item 4a, and if the entry is consistent with the name of the establishment in item 1 and Form 123, the numeric code is to be entered in box 16 at the top of the form under the legend F/0. If neither is checked, the name of the establishment in item 1 is to be compared with the name on Form 123. Then check the appropriate box of item 4 and enter the corresponding numeric code in box 16 at the top of the form.

4.3 Step B-3: Register No.

For both the long and short form, boxes

5 through 9 labeled "Register No." should have been completed in the field. The digits should have been entered in the correct position working from right to left, using leading zeros if necessary. For example, if the number is 2001 it should be entered as "02001." If the zeros were omitted, the card punch operator can be instructed to insert them. With a short form, if the boxes are blank or if the entry is suspect, the form is returned to FOD. If the same conditions exist on a long form, the questionnaire is to be returned to MMD for follow-up.

5. EDITING AND CODING PRODUCTS AND MATERIALS

(SEQUENCE C)

The objectives of sequence C in the prepunch processing are two-fold: (a) to realize the maximum publishable information on individual classes of products and materials and (b) to make sure that the assignments of economic activity codes to establishments are done in a systematic rather than a haphazard manner (see exhibit V-4-2).

Probably in no other sequence of the editing and coding operations do the lines. between "editing" and "coding" become so hazy. This is evidenced by the fact that the chief end-product of the products and materials "editing" is a "code." To be sure, this is not a routine code, as in sequence B, but rather is what the literature commonly refers to as a "research" code.

Professional analysis of the reporting of product shipments and materials consumption is a pre-requisite to the proper classification of manufacturing and mining establishments according to primary economic activity. This review work occasionally involves consulting external sources of information, and always demands the exercise of good judgement. Moreover, for some industrial sectors, knowledge of technical terms and the institutional structure is very desirable, if not absolutely essential.

5.1 Step C-1: Product code I long form)

This edit is to insure that for every product listed in item 18, col. (B), lines 1 through 9, there is a code listed in col. (C) and that the code is correct.

Each code is examined in terms of the Product Reference List Booklet, the entries in item 16, col. (B), and the information shown in item 17. Any improper code necessarily must be corrected.

the sum of the values of the uncoded entries equal or exceed the sum of the coded product entries plus any contract work receipts shown on line 10? If not activity-determining, the missing product codes are assinged in accordance with the Product Reference List Booklet. Then the codes assigned are reviewed.

If these uncoded entries are activitydetermining, the form may require field followup and is referred to the commodity specialist. If the specialist determines that the entries in question are uncodable, "C-1 reject" is marked in the upper left corner of the first page and the form is given to the staff handling field follow-up.

5.2 Step C-2: Product code (short form)

The product descriptions entered in col. (B), lines 1 through 5 of item 18 are compared with the definitions in the Product Reference List Booklet and the appropriate code for each product is entered in col. (C).

5.3 Step C-3: Product quantity (long form)

There should be an entry in col. (E) of item 18 for the product listed that has the highest value entered in col. (F) and for all products with an entry of $100,000 or more. For each product that has an entry in col. (E), the value in col. (F) is divided by this entry. The resulting implicit unit value is compared with the price range information available in the editing and coding manual. If the implicit unit value is beyond the acceptable range, the form is referred to the commodity specialist. If the value is within range, the sequence continues with step C-4.

If the product with the highest value, or any other product with an entry of $100,000

or more in col. (F), does not have an entry in col. (E), an entry is imputed using the price range information available in the editing and coding manual. For each product, the median of the known price range is calculated and the resulting value is divided into the entry in col. (F) to get an entry for col. (E). If the information is not available in the manual, "No basis for estimating" is noted for this product.

5.4 Step C-4: Cost of materials (long form) Is the total reported for cost of materials in col. (F) line 7 of item 16 the same as the figure entered in 13a?

If it is, the Materials Coding Booklet and information given in item 16 col. (B) is used as the basis for correcting or entering the appropriate code in col. (C) for each material listed. The detail on lines 1 through 5 is then added and line 6, col. (F) is used for any balancing entry needed to bring the

total on line 7 and the sum of the detail into agreement.

If the total is not the same, the item 16 total is replaced with the item 13a entry. Then it is made certain that the entry on line 6 of item 16 effects a sum balance against the new total for item 16.

6. ROUNDING

The purpose of rounding is to have all entries expressed in terms of common units of measure before they are sent for tabulation, and to make certain that published figures are significant to the smallest digit called for in publication plans.

Providencia has carefully considered the advantages and disadvantages of different methods of rounding prior to punching. The rounding rules may be summarized as calling

for rounding the three digits to the left of the decimal point so that all value entries except weekly payroll will be expressed in terms of the nearest thousand. For example, a figure of $18,329 is rounded to "18" while Where

a figure of $18,589 is rounded to "19." reporting is in terms of physical units, the full figures are punched and tabulated.

7. VERIFICATION OF EDITING-CODING All questionnaires reaching this final stage in the pre-punching processing stream should be visually inspected. This inspection does not need to be on a 100-percent basis. In Providencia's case, questionnaires were assembled in work units by type of questionnaire. This favored the use of systematic sampling to accomplish the inspection of reports before data processing.

The sampling rates cannot be foretold in advance. However, after a 100-percent inspection of the first work unit for any particular activity group, equally reliable results are achieved by visual scanning of progressively smaller samples of succeeding work units for that type of questionnaire (activity group). The initial sampling rates vary according to: (a) the proportion each work unit bears to the universe, or number of questionnaires expected for the activity group in question, and (b) the quality of editing and coding, or fraction defective. The quality is measured by the rate of errors that are reported by the verifier upon complete inspection of the first work unit(s) for the particular type of questionnaire.

8. WORKLOAD CONTROLS

The number of questionnaires by type determines the workload for editing and coding. The following estimates were prepared by the NSO.

The procedures in this unit provide instructions for (a) transcribing data from the census questionnaires to punched cards (key punching) and (b) verifying the accuracy of the information punched in the cards by comparing with the original information on the questionnaires. The receipt, check-in, and transmittal of questionnaires, the posting and checking of control totals, the maintenance of required control forms and records, and the preparation of administrative and other reports were done by the Control Clerk in the DPD Control Operations Branch. Card punching, card verification, error-card correction, reverification of corrected cards, and the card counting operations were done by the Key Punching Branch.

Card punch and verification procedures are presented in two types of instructions: (a) general and (b) detailed. The general instructions are applicable to the punching and verification of all cards required for the censuses. The detailed instructions are presented for the Census of Manufactures only. Similar types of detailed instructions would be prepared for the Minerals and the Electricity and Gas forms. Detailed instructions

should be considered as a supplement to the general instructions.

(2)

Receipt of documents