The Fieldwork Setting: An
Integrated Perspective

Daniel Lerner

Israel Central Bureau of Statistics

Tel. 972-2-655 3202

FAX 972-2-652 2319

General

In order to distribute and then collect close to two million questionnaires, it was
necessary to employ thousands of workers for periods of time ranging from eight
weeks to six months, at one hundred and fifty sites across the country.

The fieldwork integrated means and methods, including a workforce, operational
methods, a material logistics network, and an enumeration system.

Accordingly, basic planning assumptions were determined and possible problems and
constraints were identified. Similarly, it was determined what resources, both
manpower and physical, would be needed, and it was decided to computerize the field
setting.

Basic Assumptions and Constraints at Planning
Stage


1.	Out of some 43,000 candidates, about 10,000 had to be recruited and dispersed over the entire country.
2.	No more than six months were allowed for the fieldwork, from beginning to end.
3.	Limitations on resources and direction coming from the financing organization.
4.	Expectation of numerous objections and refusals on nationalistic, religious, and other over-pious grounds.
5.	The massive immigration of the early 90s would be a source of language difficulties.
6.	Legal problems requiring solution: the Tenders Law, the Statistics Ordinance and urban by-laws.
7.	Telecomputing problems, from the needs/requirements, technological solutions, and project financing points of view.
8.	Only one general rehearsal had been carried out in support of the planning effort.

Logistics

Sites

According to plan, the Bureau had rented 150 sites of various kinds that met
planners' specifications. Due to postponement of the census and the final decision
on the date of the Census Day occurring only shortly before the date itself, we
were forced to reorganize the rentals at the last minute. We had to compromise on
the size, nature, and even the location of the sites rented. Ideally, we had aimed to
use public buildings, so as to minimize inconvenience to the public, but in the end we
were even obliged to use a number of private dwellings. In choosing accommodation,
special attention was paid to preparing the building for census needs.

Means

An electrical and communications (computers, fax, telephones, etc.) network were
installed, and a complete equipment inventory was supplied, according to a pre-set
standard, in every site. This equipment comprised some six hundred items, from
tables to training aids. Equipment planning had taken into account the category of
items required, the source of supply, the issue of tenders, an extensive
procurement effort, quality control and stock and management control of the
purchased goods.

Items of Equipment

Standard office equipment -- training aids, fans, room-heaters, etc. -- were supplied
from central warehouse, which compiled an appropriate set of items and transported
it to every site. Equipment, such as furniture and computers, as well as the census
questionnaires, was supplied direct to each site by the manufacturer and/or agent,

Problem Solving

Needs and problems that could not be solved at sub-regional or regional level were
referred to the Control Center, set up on the grounds of the Central Bureau of
Statistics in Jerusalem. The Control Center was manned 14 hours per day, from 8:
00 in the morning until 10:00 at night, which were the enumeration offices' hours of
work.

The Control Center

Functions


1.	To maintain, with input from Central Headquarters, a continuously updated picture of the state of all field activity (mainly, regarding coverage and enumeration procedures) and to keep Census management apprised of the state of progress.
2.	To mediate all communications from Headquarters to the field and from field to Headquarters.
3.	To monitor fieldwork in order to ensure execution according to plan.
4.	To arrange visits and activities of Headquarters staff and others to field offices.

Tasks

In practice, the Control Center constituted the main connecting link between the field
and the three Headquarters units (Planning, Operations, Information Technology).
Problems that could not be solved at census Headquarters level, chiefly
administrative ones, were referred on to the Bureau management. Every solution to
a problem which arose at one field office was communicated by the Center to all
other offices, so as to avoid the need to find ad-hoc solutions locally and to
maintain uniformity and homogeneity of working methods around the country.

It is worthwhile noting that the Center's performance did not fully meet expectations.
It was set up very late in the process, by which time the staffers best qualified to
run it were no longer available. As a result, the brunt of the burden on the Center
was borne by one person, the Director, a fact that significantly reduced the
proper functioning and efficiency of the Control Center.

Despite this, I am convinced that the decision to set up the Center was a correct
one and its part in the correct implementation of the fieldwork was considerable.

Computerization of the Field Setting

Computerization of the fieldwork was intended to serve both census Headquarters
and field offices in the areas of staff administration, enumeration, payments, and
logistics. For Headquarters, computerization was an instrument of administration,
overview and control of census operations, and a means of ongoing supervision, by
comparing actual performance against planned performance.

For the field setting, computerization served many needs, from human resources
management (recruitment, assessment, training, placement, drawing up contracts of
employment, comparing job slots against worker roster, etc.) through to
administration of enumerator and worker payments system (from overviewing of
individual output to payment of wages) to control and monitoring of the work of
enumerators, supervisors and logistics.

The overall computer system comprised of over five hundred PCs, dispersed over
three types of sites - census Headquarters, regional offices, and sub-regional
offices -- about 150 sites in all. Since numerous difficulties of equipment
operation were anticipated, the system service provider needed to cope with the
system's special features, which were:


1.	Broad geographical scatter -- which posed problems for equipment maintenance.
2.	Unskilled users -- who were recruited especially for the census and were unfamiliar with the type of work and work procedures.
3.	Brief period of operation -- six months from installation to removal of equipment -- which allowed little time for a run-in period.
4.	A high degree of system resiliency was required in order to be able to absorb in real time the inputs of the enumeration work coming in from thousands of enumerators in the field.
5.	Assistance and support, needed to be available nationally, including quick reaction and mobility of resources, as necessary.
6.	Even geographic coverage of all services supplied to the system, in order to avoid the development of uneven progress of enumeration at different offices and to ensure that the installation and removal of the system would occur within the time allotted at all sites.

In order to meet these demands a Help Desk was set up on the supplier's premises
to act as a node of communications between the supplier and users. The duties of
the Help Desk were defined as follows:

Response to telephoned queries from users.

Telephone assistance in system operation and problem solving.

Support, instruction, and identification and repair of system failures, using
the Help Desk capacity to 'override' user computers.

Distribution of new versions of programs and updating of parameter and
table files through telephone lines.

Recording of calls for help and following up the handling of system
failures; recording control visits and results.

Communication of administrative and operational instructions/guidelines to
sites, as necessary.

Activation of supplier of computer hardware, and basic and communications
software.

Activation of telephone company to handle problems with data transfer over
telephone lines.

Activation of the office's development staff to handle system failures and
find solutions to user problems in their sphere of responsibility.

10.

Activation of support staff to handle system failures that could not be
solved via regular communications, i.e. remote computer or by telephone.

11.

Activation of census operations staff to solve administrative and logistical
problems.

12.

Responsibility for maintenance of office applications, including installation,
updating applications programs and parameter tables, and applications
failures.

13.

Conducting follow-up of the transfer of daily data from enumeration sites to
census Headquarters and vice versa.

14.

Dismantling of sites in orderly fashion up on completion of activity, including
backing-up of stored data, deleting data from storage drives, and dismantling
and packaging of all system components and preparation for transport.

Conclusion

The main factors that contributed to the success of the 1995 census were:


1.	Precise planning and pragmatic implementation.
2.	Management by strict pre-determined rules, without deviation from planned operational procedures.
3.	Fieldwork computerization, which permitted close, ongoing supervision and control of workers, procedures, and outputs.
4.	The cooperation of the population, including groups that were expected to express negative attitude to the census.
5.	Above all, the unqualified dedication and pertinacity of all involved in carrying out the census.

The Fieldwork Setting: An Integrated Perspective