adaptation-intervention cycles. However, communities in different contexts or
locations may often share similar domains of change. Hence, using predetermined
domains of change should be considered advisable rather than compulsory. Furthermore, changes that have occurred outside predefined domains should also be
collected (i.e. identified and characterized) for learning purposes in order to
improve future adaptation action M&E.
Purposive sampling was then used to collect individual level stories of changes
through interviews. The sample size for individual interviews was kept small for
experimental reasons. Purposive sampling was preferred to random sampling
because the ultimate objective of our adaptation M&E was to learn from stories
of changes, and ultimately move agricultural extension practices more towards
success and away from failure. However, to improve the validity and reliability
of the purposive sampling, discussions were conducted to collect stories of changes
of male and female groups of farmers.
The most significant change technique (Davies and Dart 2005) was used to
collect stories of changes of both individual farmers and gender-based groups.
The technique is not based on predefined performance indicators, but on “fieldbased stories” that give meaning to people’s reality and effects of projects on that
reality. It allows the story tellers (individuals or groups) to describe what has
happened in their lives and practices (particularly, in this project, the way they
farm) in conjunction with the participatory action research adaptation action.
Scientists from the respective countries’ national agricultural research systems
collected the stories of change.
The collected significant stories were subjected to participatory processing, in
which characteristics of behavior changes in the stories were counted, and then the
most significant changes were selected, substantiated and validated. To select the
most significant changes participants read the stories one by one and discussed the
characteristics of changes described by the individuals or gender-based groups. The
substantiation involved field visits and triangulation processes including discussion
with resource persons and groups in the communities to ascertain whether behavioral changes noted in the stories had effectively occurred. Such substantiation has
two objectives: (i) to verify the effectiveness of the occurrence of the change
characteristics with the story tellers, other community members and fieldworkers
who have worked with the selected communities, (ii) to gather additional data to
complement information obtained during the story collection step.
The characteristics of behavior changes were counted by extracting all identified
characteristics in the collected stories, then calculating their frequencies of occurrence, in terms of the percentages of people whose stories included them. This also
allowed the identification of domains of life where changes had been induced in the
selected communities by the participatory action research of the CCAFS program.
In this chapter we have chosen to present frequencies of occurrence of behavioral
change characteristics, but not the selection and substantiation results (which can be
obtained from the authors on request).
260
J. Somda et al.
locations may often share similar domains of change. Hence, using predetermined
domains of change should be considered advisable rather than compulsory. Furthermore, changes that have occurred outside predefined domains should also be
collected (i.e. identified and characterized) for learning purposes in order to
improve future adaptation action M&E.
Purposive sampling was then used to collect individual level stories of changes
through interviews. The sample size for individual interviews was kept small for
experimental reasons. Purposive sampling was preferred to random sampling
because the ultimate objective of our adaptation M&E was to learn from stories
of changes, and ultimately move agricultural extension practices more towards
success and away from failure. However, to improve the validity and reliability
of the purposive sampling, discussions were conducted to collect stories of changes
of male and female groups of farmers.
The most significant change technique (Davies and Dart 2005) was used to
collect stories of changes of both individual farmers and gender-based groups.
The technique is not based on predefined performance indicators, but on “fieldbased stories” that give meaning to people’s reality and effects of projects on that
reality. It allows the story tellers (individuals or groups) to describe what has
happened in their lives and practices (particularly, in this project, the way they
farm) in conjunction with the participatory action research adaptation action.
Scientists from the respective countries’ national agricultural research systems
collected the stories of change.
The collected significant stories were subjected to participatory processing, in
which characteristics of behavior changes in the stories were counted, and then the
most significant changes were selected, substantiated and validated. To select the
most significant changes participants read the stories one by one and discussed the
characteristics of changes described by the individuals or gender-based groups. The
substantiation involved field visits and triangulation processes including discussion
with resource persons and groups in the communities to ascertain whether behavioral changes noted in the stories had effectively occurred. Such substantiation has
two objectives: (i) to verify the effectiveness of the occurrence of the change
characteristics with the story tellers, other community members and fieldworkers
who have worked with the selected communities, (ii) to gather additional data to
complement information obtained during the story collection step.
The characteristics of behavior changes were counted by extracting all identified
characteristics in the collected stories, then calculating their frequencies of occurrence, in terms of the percentages of people whose stories included them. This also
allowed the identification of domains of life where changes had been induced in the
selected communities by the participatory action research of the CCAFS program.
In this chapter we have chosen to present frequencies of occurrence of behavioral
change characteristics, but not the selection and substantiation results (which can be
obtained from the authors on request).
260
J. Somda et al.