contents, page headings, page numbers; it designs covers where needed. It is responsible for the marking of copy with instructions to the printer, reviewing proofs, and checking on the assembly and binding of publications. It also sets up questionnaires and internaluse forms for printing, and the divisions of NSO have been instructed to refer any material to the Publications Division that requires the reproduction of more than 50 copies. The Forms Design Staff of PUB offers advice to the subject divisions during the developmental stages of questionnaires and other forms, and is expected to point out potential deviations from NSO standards. Responsibility for both the content and general design of forms for outside use, however, rests with the subject divisions.

6. PERSONNEL AND PAYROLL

The Personnel Division (PERS) is responsible for maintaining personnel records on all employees of NSO. It is also responsible for interviewing certain applicants for employment and making recommendations to the various NSO divisions as to the qualifications and other matters concerning individual applicants. The latter functions apply only to clerical employees of the central office in Estrella. Professional and other higher-grade employees are interviewed by the supervisory employees in the various NSO Divisions, and the field personnel, such as enumerators, are interviewed by the provincial field offices.

PERS also maintains payroll records and distributes pay checks to NSO employees. The funds for these disbursements originate with the Ministry of Finance, and authority for distribution of pay comes from the Director through his administrative aide.

In the early planning stages, agreement on the division of responsibility among the divisions and other organizational units was reached. FOD agreed that certain types of establishments (establishments of multi-unit companies, single units with 100 or more persons engaged, electricity and gas plants, and government owned-and-operated plants) would be canvassed by the staff of MMD. The major reason for this decision was to simplify data collection. The NSO believed there were less than 500 establishments of these types, and the planners reasoned that the writing of complex instructions to the field staff could well be replaced by oral instructions to a small group of experts plus brief written instructions. Field operations would, however, be under the direct control of the head of FOD in conformity with the policy of NSO on data collection.

As to editing of questionnaires, it was agreed that all long forms would be edited by the MMD staff. The short forms, on the other hand, would be edited by clerks in the DPD in accordance with specifications to be furnished by MMD.

Tables for publication would be hand-posted by clerks in DPD or printed by computer, according to MMD specifications.

Exhibit II-2-1 lists the various functions

to be carried out in the industrial census. The exhibit gives the name of the organizational unit in the Providencia NSO having major responsibility for the function.

9. COORDINATOR AND EXECUTIVE COMMITTEE In recognition of the need for liaison among divisions, the Director appointed a senior staff member of MMD to the position of Industrial Census Coordinator. The primary responsibility of this position is to be aware of the progress of the various phases of work in the census, and to consult with the heads of divisions when delays develop, when staff have too much or too little work, or when disagreements arise among staff members as to their

responsibilities. Although the Coordinator's primary contact is the head of MMD, the function calls for frequent consultation with other division heads as well. most part, the Coordinator's responsibilities are advisory rather than administrative.

For the

All NSO major decisions are made or approved by the Executive Committee. This committee, which consists of the Director and the heads of NSO divisions or offices, meets regularly once a week and at other times as necessary. This regular review of ongoing projects is vital to the success of the census. Regular review not only assures top-level attention to problems but it keeps minor matters from developing into problems.

1. IMPORTANCE AND CHARACTERISTICS
OF CENSUS MAPS

Maps are important and essential tools in the conduct of any census in which personal enumeration is the chief method of collecting data. They are of use in preventing overlapping of jurisdictions and in avoiding omission and duplication of work and of data--all the way from the enumerator's level up to the central office. Maps help to locate places and assist in defining boundaries for general administrative purposes. They can also be used effectively in preparing budget estimates of travel costs, location of field offices, and number of employees required in the field. If maps are not available or if the maps are not very accurate, the progress of the field work can be seriously hampered.

Many maps are constructed for specific purposes such as naval charts, military field maps, urban planning projections, geological maps, sales quota maps, etc. A census agency also has special needs. The maps required will vary, depending on the type of unit to be enumerated, the kind of area involved (particularly, whether the area is urban or rural), and the needs of the individuals working with the maps. In general, large-scale maps are needed by field enumerators and their supervisors, whose basic objective is to canvass a small area completely and without duplication. Intermediate-scale maps are needed by employees whose scale of operations embraces a larger area such as a county, while small-scale maps may provide overviews of a province or the entire country for management personnel. There is not, then, one standard map which can be used effectively by enumerators, provincial

supervisors, and top management. Furthermore, maps that are adequate for a population census or a census of distribution may be inadequate for an industrial census. Maps that are serviceable in rural areas may be completely inadequate for the enumeration of urban areas.

1.1 General characteristics

Certain general statements can be made in regard to the requirements of all census maps. The more important items are listed below:

1.11 Distance.--Showing the exact distance between two features is not as important as reflecting the relationship between them. Whether a factory is shown 50 meters or 75 meters from the road is less significant than whether it is shown on the correct side of the road.

1.12 Areas to be tabulated separately. All boundaries which define areas for which data are to be shown separately in the census tabulations must be clearly and accurately shown on the maps. In the industrial census, considerable emphasis was placed on the correct delineation of county boundaries in relationship to industrial establishments and other features of the landscape, because the county was the smallest geographic area for which separate data were published.

1.13 Unnecessary detail.--Maps should be simple and contain little information which is not pertinent to the needs of the user. A detailed and complicated map may only serve to confuse the user, because some temporary employees, particularly enumerators, have had only brief training in map use prior to their work in the census.

This principle was violated to some extent in the industrial census because, as explained below, all of the maps used were adapted from those prepared for the 1970 Census of Population and Housing. They contained some detail which was not needed for the industrial census enumeration. To a considerable extent, this disadvantage was compensated by the fact that many of the enumerators and supervisors in the industrial census had been employed in the 1970 census and were trained in map use.

1.14 Graphic scale.--Map distance should be shown by means of a graphic scale (see the exhibits for this chapter). Maps can readily be enlarged or reduced in size by modern copying methods. Graphic scales, as opposed to those which state a fixed relationship, such as one centimeter equals 500 meters, do not need to be altered if the size is changed.

1.2 Physical and "cultural" features

Features shown on census maps are of two kinds--physical (or natural) and cultural (or man-made). Natural features such as mountains, rivers, marshes, deserts, lakes, etc. are included when they are expected to be of assistance in locating the boundaries of an area or place to be enumerated. Contour maps showing gradual changes in land elevation are generally not useful, but very prominent elevations such as abrupt ridges should be shown.

Man-made features such as highways, roads, canals, and trails should be carefully delineated on industrial census maps, because places to be enumerated are usually located along them. The kind of road (trail, graded dirt road, gravel surface, paved, etc.) should be indicated if possible because it provides a clue to the accessibility of places and is of help in estimating travel requirements. Knowing the kind of road may be particularly

important if the period of enumeration coincides with a rainy or very cold season. For the industrial census, man-made features such as cemeteries, churches, schools, apartment buildings, etc., were of less importance but were allowed to remain on the maps adapted from the population census because they were considered to be useful in identifying roads and streets, which were seldom named on street signs. Features such as industrial establishments, which were added when adapting the map, were also of considerable value.

1.3 Boundaries

Normally, the best boundaries to delimit an area are those which can be seen easily on the ground, such as roads, streams, and railroad tracks. Nevertheless, since it is neces

sary to hold to administrative area limits, imaginary lines must often be used. Sometimes, property lines may be followed in order to define areas; at other times, imaginary lines connecting visible features must be drawn to complete a boundary.

Between 1970 and 1975, there were many changes in the boundaries of wards and counties, but none in the province boundaries. Fairly accurate information on the boundary changes was obtained from the Ministry of Finance, and the census maps were revised accordingly. To allow time for the reproduction and distribution of maps, the Geography Division established a cut-off date of 30 June 1975. Changes received later than that date were not carried through to the maps.

2. KINDS OF MAPS USED IN THE
INDUSTRIAL CENSUS

The maps used in the industrial census, as pointed out above, were adapted from those used in the 1970 Census of Population and Housing. Some of their features were of no particular value in the industrial census, but

the cost of deleting these features from the maps led to a decision to retain them. For example, some maps described below included crew leader area boundaries. Crew leaders were first-level supervisors employed in the population census but not in the industrial

census.

Most of the exhibits show only enough of the entire map to indicate its features. The standard size of the map, in most cases, was 60 by 72 cm. Maps of this size to be carried in the field could conveniently be folded to 10 by 24 cm., a size which would readily fit into pocket or purse.

2.1 Large-scale map for rural areas (exhibit 11-3-1)

These maps were used in the field by enumerators, both for the canvass of "regular" establishments and for the household area sample. Although ED boundaries are shown on the maps, the ward was sometimes treated as the smallest enumerator assignment. Usually, each map covered a complete ward, and two maps were needed for some of the larger wards.

2.2 Large-scale urban map for household sample areas (exhibit 11-3-2)

These maps were used by supervisors to record and control the enumeration of urban areas included in the household sample. The enumerators in urban areas did not carry whole maps but were given portions of maps together with a written description of the boundaries of the areas they were expected to canvass. The estimates of the number of household units in each block, recorded in the 1970 census, provided an approximate idea of the number of interviews to expect from each block.

2.3 Large-scale urban map for general use lexhibit 11-3-3)

These maps were used by enumerators and supervisors in the canvass of "regular"