should guide the FSR/E activities within each region.
 Factors which influence the choice of research
regions where work is to begin or to be concentrated can
include: the number of small, limited-resource farms;
rural population density; the importance of the region for
food and cash crop production; the agricultural potential
of the region; and the potential to extrapolate research
findings to similar regions. As a practical matter, the
choice of research regions may be to support political
objectives, such as counteracting rural unemployment or low
income levels or enhancing a national development plan.
In determining the number of research regions to which
on-farm research teams are to be assigned, the number of
separate research subjects, number of trials, and the like
should be considered. The main criteria will be those
that reach an equilibrium between priority needs and what
is possible, given available resources. The importance of
program planning and budgeting cannot be overstated. Once
research is underway in several regions, interregional
coordination must assure that information is exchanged and
resources are allocated to each region and client group on
an equitable and efficient basis.

EVALUATING ON-FARM RESEARCH

 As research moves off station, some support and
operation costs will be incurred. The logistics of the
support and budgeting for on-farm research is an integral
part of the planning process. These costs will be more
than offset by the benefits of conducting more of the
research under the physical, ecological, and socioeconomic
conditions faced by farmers. Further, as leaders of
commodity and disciplinary groups are able to carry out
more trials on-farm, they are also better able to target
on-station research toward the needs of client groups.
Three principal groups of clients use the results of
agricultural sector research. These are: farm families;
agricultural sector support services which provide inputs
and/or marketing services to farms; and public planning
and policy clients, such as legislatures, planning
departments, or financial ministries.
 As an FSR/E program matures, some mechanism is needed
by which both on-farm and on-station research activities
can be evaluated, along with their contribution toward