TABLE VII-6. Number of collaborators and area planted
 in twin rows of maize interplanted with
 potatoes, Chimaltenango, 1980 (n = 7).

 Area (ha)
 or Number Percent

Total area planted to maize (ha) 3.85 100

Collaborators who used the 4 57
 twin rows

Total area of all maize on
 4 farms with twin rows (ha) 1.4

Total area of maize in
 twin rows on 4 farms 0.23 16

Index of acceptability 9
 (57 x 16) / 100
Source: Samayoa, 1980



Response surfaces derived from modified stability
analysis

 Previous examples of the modified stability analysis
were based on discrete, or non-continuous, variables.
Crossing of the regression lines suggests a need to
partition the recommendation domain because it indicates
that in part of the domain one technology is better, while
in another part of the domain a different technology is
superior.
 Modified stability analysis is also amenable to
analysis of continuous variable data, such as fertilizer
response. With a continuous variable, differential
response between poor and good environments can be
detected and used to partition a recommendation domain
even where cross-over does not occur. Data from a
nitrogen response trial for radishes conducted as part of
a farming systems class at the University of Florida are
used in the following to show continuous data analysis and
the potential for partitioning recommendation domains.