Front Cover
 Title Page
 Table of Contents
 Additional training materials of...
 Evaluation sheet
 Volume II: Organization of...
 Unit I: What kind of testing to...
 Unit II: What treatments to test,...
 Unit III: How to design trials...
 Unit IV: How to carry out...
 Unit V: How to manage and administer...

Group Title: FSR/E Training Unit: Volume II
Title: Design techniques for on-farm experimentation
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00053816/00001
 Material Information
Title: Design techniques for on-farm experimentation
Series Title: FSR/E Training Unit: Volume II
Physical Description: Book
Language: English
Creator: Farming Systems Support Project
Affiliation: University of Florida -- Farming Systems Support Project -- Institute of Food and Agricultural Sciences
Publisher: Farming Systems Support Project, Institute of Food and Agricultural Sciences, University of Florida
Publication Date: 1987
Subject: Agriculture   ( lcsh )
Farm life   ( lcsh )
Farming   ( lcsh )
University of Florida.   ( lcsh )
Spatial Coverage: North America -- United States of America -- Florida
 Record Information
Bibliographic ID: UF00053816
Volume ID: VID00001
Source Institution: Marston Science Library, George A. Smathers Libraries, University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved, Board of Trustees of the University of Florida
Resource Identifier: oclc - 17763091

Table of Contents
    Front Cover
        Front Cover
    Title Page
        Title Page
    Table of Contents
        Table of Contents
        Page i
        Page ii
        Page iii
        Page iv
    Additional training materials of interest
        Page v
        Page v
        Page vii
        Page viii
    Evaluation sheet
        Page ix
        Page x
    Volume II: Organization of manual
        Page xi
        Page xii
    Unit I: What kind of testing to do
        Page 1
        Linkages between diagnosis and design
            Page 1
            Page 2
            Page 3
            Page 4
            Page 5
            Page 6
            Page 7
            Page 8
            Page 9
            Page 10
        Ways to do on-farm experimentation
            Page 11
            Page 12
            Page 13
            Page 14
            Page 15
            Page 16
            Page 17
            Page 18
        Planning for evaluation of alternative technologies
            Page 19
            Page 20
            Page 21
            Page 22
            Page 23
            Page 24
            Page 25
            Page 26
            Page 27
            Page 28
            Page 29
            Page 30
            Page 31
            Page 32
            Page 33
            Page 34
            Page 35
            Page 36
            Page 37
            Page 38
            Page 39
            Page 40
            Page 41
            Page 42
            Page 43
            Page 44
            Page 45
            Page 46
    Unit II: What treatments to test, and where
        Page 47
        On-station and on-farm linkages
            Page 47
            Page 48
            Page 49
            Page 50
        What kinds of fields are available for testing
            Page 51
            Page 52
            Page 53
            Page 54
            Page 55
            Page 56
            Page 57
            Page 58
            Page 59
            Page 60
        What treatments are possible
            Page 61
            Defining treatment objectives
                Page 61
                Page 62
                Page 63
                Page 64
                Page 65
                Page 66
                Page 67
                Page 68
                Page 69
                Page 70
                Page 71
                Page 72
            What to consider in selecting subsets of experimental treatments
                Page 73
                Page 74
                Page 75
                Page 76
                Page 77
                Page 78
                Page 79
                Page 80
                Page 81
                Page 82
                Page 83
                Page 84
                Page 85
                Page 86
            Statistical techniques for selecting subsets
                Page 87
                Page 88
                Page 89
                Page 90
                Page 91
                Page 92
                Page 93
                Page 94
                Page 95
                Page 96
                Page 97
                Page 98
                Page 99
                Page 100
                Page 101
                Page 102
                Page 103
                Page 104
                Page 105
                Page 106
            Choosing control treatments
                Page 107
                Page 108
                Page 109
                Page 110
                Page 111
                Page 112
                Page 113
                Page 114
                Page 115
                Page 116
        What is a treatment
            Page 117
            Specification of experimental treatments
                Page 117
                Page 118
                Page 119
                Page 120
                Page 121
                Page 122
            Specification of non-experimental variables
                Page 123
                Page 124
                Page 125
                Page 126
                Page 127
                Page 128
                Page 129
                Page 130
                Page 131
                Page 132
            Inputs and calculations
                Page 133
                Page 134
                Page 135
                Page 136
                Page 137
                Page 138
        Examples of treatments for different types of problems
            Page 139
            Page 140
            Page 141
            Page 142
        Looking ahead: What are some tradeoffs between treatments and replications
            Page 143
            Page 144
            Page 145
            Page 146
            Page 147
            Page 148
            Page 149
            Page 150
            Page 151
            Page 152
            Page 153
            Page 154
    Unit III: How to design trials to obtain analyzable data
        Page 155
        How objectives change in the research-extension process
            Page 155
            Page 156
            Page 157
            Page 158
            Page 159
            Page 160
            Page 161
            Page 162
            Page 163
            Page 164
            Page 165
            Page 166
            Page 167
            Page 168
            Page 169
            Page 170
        What designs can do
            Page 171
            Page 172
            Page 173
            Page 174
            Page 175
            Page 176
            Page 177
            Page 178
            Page 179
            Page 180
            Page 181
            Page 182
            Page 183
            Page 184
            Page 185
            Page 186
            Page 187
            Page 188
            Page 189
            Page 190
            Page 191
            Page 192
            Page 193
            Page 194
            Page 195
            Page 196
        What designs are possible
            Page 197
            Ways to replicate treatments within and across farms
                Page 197
                Completely random designs and randomized complete block design
                    Page 197
                    Page 198
                    Page 199
                    Page 200
                    Page 201
                    Page 202
                    Page 203
                    Page 204
                    Page 205
                    Page 206
                    Page 207
                    Page 208
                Incomplete block designs across farms
                    Page 209
                    Page 210
                    Page 211
                    Page 212
                    Page 213
                    Page 214
                    Page 215
                    Page 216
                    Page 217
                    Page 218
                    Page 219
                    Page 220
                    Page 221
                    Page 222
                    Page 223
                    Page 224
                    Page 225
                    Page 226
                    Page 227
                    Page 228
                    Page 229
                    Page 230
                    Page 231
                    Page 232
            Ways to combine treatments within replications
                Page 233
                Page 234
                Page 235
                Page 236
                Page 237
                Page 238
    Unit IV: How to carry out trials
        Page 239
        Ways to involve household members
            Page 239
            Page 240
            Page 241
            Page 242
            Page 243
            Page 244
            Page 245
            Page 246
            Page 247
            Page 248
        How to lay out trials
            Page 249
            Page 250
            Page 251
            Page 252
            Page 253
            Page 254
            Page 255
            Page 256
            Page 257
            Page 258
            Page 259
            Page 260
        How to obtain and handle data from trials
            Page 261
            Page 262
            Page 263
            Page 264
            Page 265
            Page 266
            Page 267
            Page 268
            Page 269
            Page 270
            Page 271
            Page 272
            Page 273
            Page 274
            Page 275
            Page 276
            Page 277
            Page 278
            Page 279
            Page 280
            Page 281
            Page 282
            Page 283
            Page 284
            Page 285
            Page 286
            Page 287
            Page 288
            Page 289
            Page 290
            Page 291
            Page 292
            Page 293
            Page 294
            Page 295
            Page 296
            Page 297
            Page 298
            Page 299
            Page 300
            Page 301
            Page 302
            Page 303
            Page 304
            Page 305
            Page 306
            Page 307
            Page 308
            Page 309
            Page 310
            Page 311
            Page 312
            Page 313
            Page 314
            Page 315
            Page 316
            Page 317
            Page 318
            Page 319
            Page 320
            Page 321
            Page 322
            Page 323
            Page 324
            Page 325
            Page 326
            Page 327
            Page 328
            Page 329
            Page 330
    Unit V: How to manage and administer FSR/E at the field level
        Page 331
        Page 332
        Page 333
        Page 334
        Page 335
        Page 336
        Page 337
        Page 338
        Page 339
        Page 340
        Page 341
        Page 342
        Page 343
        Page 344
        Page 345
        Page 346
        Page 347
        Page 348
        Page 349
        Page 350
        Page 351
        Page 352
        Page 353
        Page 354
        Page 355
        Page 356
Full Text



'01 AW Ilk
I .......... '"Pow
padickwo 1"m now

P 1 .



Prepared By:

Farming Systems Support Project
International Programs
Institute of Food and Agricultural Sciences
University of Florida
Gainesville, Florida 32611

Technical Editor:

John Caldwell, Virginia Polytechnic Institute

Coordinating Editor: Lisette Walecka, University of Florida

2nd edition December, 1987

The Farming Systems Support Project (FSSP) is a cooperative
agreement between the University of Florida and the United States
Agency for International Development, Cooperative Agreement No.
DAN-4099-A-00-2083-000, Project number 936-4099.

~ETEp E~IqLr,1A,


PREFACE ..................................................... ............ i

ACKNOWLEDGEMENT .........................................................

VOLUME II: ORGANIZATION OF MANUAL ...................................... xi


What Kind of Testing To Do
Linkages Between Diagnosis and Design....................... 1
Ways to Do On-Farm Experimentation........... ........11
Planning for Evaluation of Alternative Technologies..........19 -

UNIT II: What Treatments To Test, and Where
(II,A) On-Station and On-Farm Linkages............................. 47
(II,B) What Kinds of Fields are Available
For Testing................................................51
(II,C) What Treatments are Possible
(II,C,1) Defining Treatment Objectives........................61
(II,C,2) What to Consider in Selecting Subsets of
Experimental Treatments..............................73
(II,C,3) Statistical Techniques for Selecting Subsets..........87
(II,C,4) Choosing Control Treatments.........................107
(II,D) What is a Treatment
(II,D,1) Specification of Experimental Treatments .............117
(II,D,2) Specification of Non-experimental
(II,D,3) Inputs and Calculations............................133
(II,E) Examples of Treatments for Different
Types of Problems ...........................................139
(II,F) Looking Ahead:.What are Some Tradeoffs
Between Treatments and Replications .......................143

UNIT III: How to Design Trials to Obtain Analyzable Data
(III,A) How Objectives Change in the
Research-Extension Process.................................155
(III,B) What Designs Can Do........................................171
(III,C) What Designs are Possible
(III,C,1) Ways to Replicate Treatments Within and
Across Farms
(III,C,l,a) Completely Random Designs and Randomized
Complete Block Design......................... 197
(III,C,l,b) Incomplete Block Designs Across Farms..........209
(III,C,2) Ways to combine treatments within

UNIT IV: How to Carry Out Trials
(IV,A) Ways to Involve Household Members.........................239
(IV,B) How to Lay Out Trials.....................................249
(IV,C) How to Obtain and Handle Data from Trials......... ........261
Activities: ............ .............. .................. .........303

UNIT V: How to Manage and Administer FSR/E
at the Field Level.........................................331


One of the major objectives of the Farming Systems Support Project is
to provide training and support for training activities in FSR/E
methodology. This collection of training units has been produced in
response to an absence of available training materials which could be used
in training practitioners in the skills necessary for implementing the
FSR/E approach to agricultural development. The development, testing,
review, and revision process has been rapid due to the demand for training
materials and the limited time remaining for the project. These training
volumes are not error free. We encourage your scrutiny. As you work with
the training manuals and if you have comments, additions., adaptions,
corrections, or suggestions please let us know.

This collection of training units is not a course. Rather, it is a set
of resources which supports FSR/E courses. It is an attempt to provide the
trainer and practitioner trainee with a wide variety resources for teaching
and learning specific content and skills needed for implementing FSR/E

Volume I, Diagnosis in FSR/E, contains nine units for introducing
trainees to various diagnostic steps in the FSR/E approach. It stresses,
but is not limited to, initial diagnosis. Volume One also contains units
which detail on-going, or continuous, diagnosis throughout the FSR/E
process. Links between social and biological science disciplines are
stressed, as are considerations of intra-household and socio-cultural
issues. The final unit focuses on problem identification and
prioritization, a step leading toward appropriate trial design.

Volume II, Design Techniques for On-Farm Experimentation, contains five
units which detail the farm trial design, layout, and management process.
The links between biological and social science disciplines in on-farm
research are considered, and, like Volume One, intra-household and
socio-cultural issues are addressed. This volume begins by focusing on the
importance of establishing clear evaluation criteria before designing a
trial and culminates with a discussion of practical implications of
managing on-farm experimentation.

Volume III, Analysis and Interpretation of On-Farm Experimentation,
establishes a framework for analysis reinforcing the importance of the
establishing evaluation criteria early in the design process. It provides
basic statistical and analytical techniques useful for on-farm
experimentation and ties all volumes together by introducing the concept
of integrated analysis.

A Trainer's Manual accompanies these three volumes, and provides notes
for the trainer which accompany the variety of activities presented in each
volume. One of the objectives of this series of training materials is to
provide participatory activities which will involve the participant
directly in the training in a "hands-on" fashion. It emphasises group
discussions and role play as well as other types of activities.

This is the second edition of the training units, and the revisions
which were made were based on comments from a variety of sources, including

specific reviewers, participants in shortcourses, individual users of the
training manuals, and others. We have tried to address the majority of the
concerns voiced. Major revisions included integration of livestock issues
and expansion of economic analysis material. Emphasis has also been placed
on presenting material in a "how to do" fashion.

The units have not been developed to be exhaustive texts of the the
topics presented. Rather, they have been developed to convey basic
information in a format as complete and concise as possible. It is our
hope that both trainers and trainees will search out more information on
specific topics covered in the training units. The learning objectives and
key points focus on the main essence of the unit or section. A common
glossary gives all the definitions in one place, since many terms are used
in more than one unit or section. Many units are divided into sub-units,
sections and sub-sections, each with its own set of learning objectives,
key points, list of terms, and discussion. Suggested learning activities
accompany the units or sections and each activity has separate instructions
for trainers.

The units are not thought to be the "final word." Rather, they have
been developed as the foundation of developing training units in FSR/E.
Your comments, adaptations, additions, and suggested activities are
welcomed and encouraged. The best measure of the usefulness of a product
is given from those who use the product. The best way to improve a product
is to listen to the users. At the end of this introduction you will find a
one page evaluation sheet. We hope that you will use this form to send us
your comments. This is not meant to limit your comments (and we encourage
detailed-comments) but rather to encourage you, the user, to let us know
what you think and suggest.

Throughout the development process of the FSR/E training units, from
the planning, writing, initial editing, reviewing, testing, revising, to
the final production, and second edition, many individuals have been
involved. FSSP would like to acknowledge their efforts. The individuals
are listed below with their affiliations at the time of their

Technical Editors:

Volume I: Tim Frankenberger University of Arizona
Steve Franzel Development Alternatives, Inc.
Malcolm Odell Synergy International
Marcia Odell Synergy International

Volume II: John Caldwell Virginia Polytechnic Institute

Volume III: John Caldwell Virginia Polytechnic Institute

Volume III, Economic Analysis sections:

Dan Taylor Virginia Polytechnic Institute

Initial Planning:

Emanuel Acquah
Lorna Butler
Steve Franzel
Dan Gait
James Jones
Susan Poats
Federico Poey
Lisette Walecka


Jay Artis
Emanuel Acquah
Kenneth Buhr
Lorna Butler
John Caldwell
Cornelia Flora
Steve Franzel
Dan Galt
Martha Gaudreau
John Hammerton

James Jones
Kenneth McDermott
James Meiman
Malcolm Odell
Ramiro Ortiz
Donald Osburn
Susan Poats
Kenneth Sayre
Jerry Van Sant
Robert Waugh
Peter Wotoweic
Peter Hildebrand
Dan Taylor
Henk Knipsheer
Al Hagan
Don Osburn
Marilynn Prehm
John Lichte
Jim Oxley
Mark Kujawa
John Russell


University of Maryland, Eastern Shore
Washington State Univeristy
Development Alternatives, Inc.
University of Florida, FSSP
University of Florida, FSSP
University of Florida, FSSP
Agricultural Development Consultants, Inc.
University of Florida, FSSP

Michigan State University
University of Maryland Eastern Shore
University of Florida
Washington State University
Virginia Polytechnic Institute
Kansas State University
Development Alternatives, Inc.
University of Florida, FSSP
University of Minnesota
Caribbean Agricultural Research and Develo
University of Florida, FSSP
t University of Florida, FSSP
Colorado State University
Synergy International
Agricultural Development Consultants, Inc.
University of Florida, FSSP
International Agricultural Development Ser
Development Alternatives, Inc.
Colorado State University
University of Florida
University of Florida
Virginia Polytecnic Institute
Winrock International
University of Missouri
U.S. Agency for International Development
Virginia Polytechnic Institute
University of Florida
Colorado State University
Colorado State University
University of Florida

Ron Knapp

Dan Minnick
Robert Tripp
Janis Timberlake
Clive Lightfoot
Ly Tung

Centro Internacional de Mejoramiento de Maiz y Trigo
International Rice Research Institute (IRRI)
Virginia Polytechnic Institute
Cornell University
Visayas State College of Agriculture, Philippines

pment Institute

vice (IADS)

Training Consultants:

Kathy Alison Office of International Cooperation, & Development (OICD)
Peg Hively Office of International Cooperation, & Development (OICD)

The FSSP would like to thank CIMMYT Economics Program and CARDI for
permission to include portions of their work in economic analysis and
on-farm experimental design respectively.


The draft edition of the Volume Two, Techniques for Design and Analysis
of On-Farm Experimentation, was used for the first time in the FSSP/Gambia
Agricultural Diversification workshop on On-Farm Experimentation in May,
1985. Parts of Volume One, Diagnosis in FSR/E, were used for the first time
in the Jamaica Farming Systems Research Workshop, June, 1985. Feedback
received during this initial testing was used, along with other feedback,
in the revising effort.

Richard Bernsten, Michigan State University, presented the FSSP
training units for review at the "Farming Systems Research Socio-Economics
Monitoring Tour/Workshop," held September 16 28, 1985, at IRRI, Los
Banos, Philippines, at the request of Marlin Van Der Veen, IRRI. Comments
from that session, as well as detailed comments by Richard Bernsten, were
very useful in revising both volumes. Susan Almy, Rockefeller Foundation,
also provided very detailed comments. Additional review comments were made
by Peter Hildebrand, University of Florida. Martha Gaudreau, University of
Minnesota, played an important role in the revision of the Diagnostic Unit.
Klaus Hinklemann, Virgina Polytechnic Institute, provided valuable
consultation on some statistical aspects of the units.

Specific reviewers for the volumes included Hal McArthur (University of
Hawaii), Roque de Pedro (Viscaya State College of Agriculture, The
Philippines), Cornelia Flora (Kansas State University), John Lichte
(University of Florida), Eric Crawford (Michigan State University).
Valuable comments were also offered by Janis Jiggins and Federico Poey

The FSSP acknowledges the above contributions and those of others who
may have been inadvertently omitted. I would like to gratefully
acknowledge the patience, hard work, and general support of the FSSP
secretaries, Lana Bayles, Shirlene Washington, and Jack Weiss throughout
the training unit development process. I would also like to thank Donna
Long, secretary senior at Virginia Polytechnic Institute for her valuable
and patient assistance throughout the revision process.
Lisette Walecka
Coordinating Editor
December, 1987


The statistical interpretations and explanations in volume II of this
series is based on the statistical tables in (Rolf and Sokal, 1969), other
tables may be slightly different. We recommend:

F. James Rolf and Robert R. Sokal, State University of New York at
Stonybrook, 1969, Statistical Tables, W. H. Freemand and Company, San

CARDI, April, 1984, "On-farm Experimentation: A Manual of Suggested
Experimental Procedures.

CIMMYT, revised November, 1985, "Introduction to Economic Analysis of
On-Farm Experiments", Draft Workbook, CIMMYT Economics Program,

FSSP, 1985, "Bibliography of Readings in Farming Systems, volume 1,"

Poey, F. et. al, 1985, "Anatomy of On-Farm Trials: A Case Study From
Paraguay", FSSP.

Hildebrand, P. and F. Poey, 1985, On-Farm Agronomic Trials in Farming
Systems Research and Extension", Lynne Rienner Publishers, Inc., Boulder

The FSSP has developed and initially tested a case study based on
Dominican Republic data from the Las Cuevas region which gives trainees the
opportunity to interview farmers and develop research priorities.

Intra-Household Dynamics and Farming Systems Research and Extension: Case
Studies in Agricultural Development. The Population Council and The Farming
Systems Support Project hacvce developed a set of seven teaching cases which
directly address the relationship between an understanding of intra-household
dynamics and the design and extension of new technologies for improving farm
production. Each case, in two or three sequenced sections, provides trainees
with information drawn from actual project experience with which they can
analyze relationship of gender roles and intra-household dynamics to the
farming system and make decisions about future project activities. The seven
cases, described below, are accompanied by background papers, a conceptual
framework for analyzing the cases, guidelines for studying a case, and
teaching guidelines.

Based-on the work of the Adaptive Research Planning Team in Central Province,
Zambia, the material includes initial diagnostic surveys, labor survey,
on-farm trial protocols and results and special studies on decision making and
female headed households. It is a good beginning case and can be used alone
or with other cases for either short term training or a longer term classroom

Improvement in the production of staple cereals and other crops was the
objective of the Purdue Unversity and the Semi-Arid Food Grain Research and

Development project (SAFGRAD) in three villages of the Mossi plateau of
Burkina Faso. The case includes initial diagnosis, the results of three years
of on-farm trials, and labor studies. This case is particularly suitable for
a longer term training situation and for audiences with technical interests.

This case covers eight months of an on-farm testing project for varietal and
fertilizer technology components conducted by the International Center for
Tropical Agriculture (CIAT) and the International Center for Fertilizer
Development (IFDC) in Pescador, Colombia. The material includes a description
of the composition and objectives of the multidisciplinary research team,
successive stages of information generated to design and evaluate the
experimentation phase, design of on-farm trials and the generation of
additional information regarding women's activities related to production and
consumption in the farming system. This case works well in both short and
long term training and with general and technical audiences. it is
particularly useful for looking at different disciplinary perspectives towards
technology design, innovative approaches in diagnostic research, and the
inclusion of consumption considerations.

This case describes diagnostic surveys and a proposed intervention undertaken
in the Mabouya Valley of St. Lucia by the Caribbean Research and Agricultural
Development Institute (CARDI). The area is dominated by plantation
agriculture on the valley floor and small farms and subsistence farms at
higher elevations. Seasonal and long term migration of males is
characteristic. The three parts the diagnostic surveys, case profiles, and
proposed interventions may be used in several ways in both short and long
term training. It is best used as a second case in a series of cases.

This case describes an agroforestry research and extension project undertaken
by a non-government organization, CARE/Kenya, with assistance from the
International Center for Research in Agroforestry (ICARF) in the Western
Province of Kenya. Diagnostic and extension activities are done with groups
and individual farm households. Material includes initial diagnosis, the
training for and methodology used by field personnel to insure that both women
and men were included, the results of formal trials and further research, and
on-farm design activity. This case is particularly suited for looking at
methodologies for working with groups and for applying benefits analysis to
technology choice. It is suitable for both short term and long term teaching

The primary objective of TROPSOIL'S multidisciplinary team is the development
of techniques for soil management in Sitiung, a transmigration site in Sumatra
which includes migrants from Java as well as indigenous peoples. Thecase
includes technical information on soils and forages, procedures and results of
the initial sondeo, on-farm trials, time allocation studies, nutrition and
income studies, and forage trials. Both ethnic and gender differences
influence farmer preferences and technological possibilities. This case is
particularly rich and is best used in a long term training situation.

This case depicts a project to improve arable production in the Mahalapye
District of Botswana, an area with low and erratic rainfall, an economy
dominated by cattle and a high percentage of female headed households.
Included are a summary of the technical and socio-economic research during the
first three years of the project with increasing specification of household
characteristics and dynamics, the fourth season's trails and farmer
evaluations, and additional diagnostic work targeted on poorer predominantly
female households. This case is best used in a longer training situation.

Written Materials: For use by trainers and trainees or for self study.
Volume I: Case material
Background articles on Gender Roles and Farming Systems and on Farming
Systems Research; Conceptual Framework for analyzing household dynamics
and farming systems; Introduction to the Case Study Method; and Individual
Case Studies
Volume II: Analysis and Teaching Notes
Teaching by the Case Study Method and examples; Best uses for each case;
and Analysis and Teaching Notes for individual cases.
Experienced consultants for training, case writing, or project assistance
One day or two day pre-conference workshops
One week course on Intra-household Dynamics and Agricultural Research and
One week course on developing own case materials
Training of trainers

For more information contact: Hilary Sims Feldstein, Managing Editor
RFD 1 Box 821
Hancock, New Hampshire 03449

Dr. Susan Poats
FSSP/ University of Florida
3028 McCarty Hall
Gainesville, Florida 32611

Your comments are encouraged. Please feel free to write your comments
and send them to the FSSP at the address listed on the back of this form.
Being specific about the unit, sub-unit or sections which you are
discussing will assist us in our efforts to provide quality materials.

(optional) NAME:



1. How did you find the units most/least useful?



2. How was the content most..........



3. Was the level of presentation appropriate?

4. Was the volume organized appropriately?

5. In the future editions what would you want to see......





6. How useful were the existing activities provided in the unit?






This volume presents techniques useful in the design. On-farm
experimentation moves through a series of steps. At each step there are
various choices to make. No one choice is right all the time. Each choice
has advantages and disadvantages. Sometimes one choice is better than
another under one set of farmer conditions and a given problem to solve.
The next time, under different farmer conditions, or for a different
problem to solve, another choice may be better. The objective of this
volume is to help farming systems teams make better choices.

The steps in on-farm experimentation help farming systems teams and
farm households answer several questions. Each unit in this volume
addresses one of these questions.

Unit I. What Kind of Testing To Do?

Initial diagnosis usually reveals more farmer problems than a team can
address at one time. Prioritization of farmer problems is necessary.
Prioritization is the link between initial diagnosis and design of on-farm
experimentation. Planning for evaluation through setting clear evaluation
criteria helps to focus on-farm experimental design and analysis.

Types of trials for priority problems can be classified by three

a. The basis of the production system (yam-based, rice-based, etc.)
b. The cropping or grazing pattern (monoculture, crop association,
pasture rotation, etc.)
c. The subject of the problem (nutrition, disease, spacing, etc.)

On-farm experimentation can take two pathways. One pathway is based on
spontaneous farmer experimentation. Another pathway is based on
researcher-planned experimentation. Researcher planned experimentation
moves through a sequence of trials. Farmer management increases as the
trial sequence progresses.

Unit II. What Treatments to Test, and Where?

Many different treatments may be possible for a priority problem. Some
will be more useful to farmers than others. The amount of land available
on each farm can limit the number of treatments. More replications on each
farm can also limit the number of treatments. Conversely, more treatments
may limit the number of replications possible on each farm. Input from
farmers is important in making choices among experimental treatments,
control treatments, treatment specifications, and replications.

Unit III. How to Design the Trial to Obtain Analyzable Data?

Farmers' fields and livestock are variable even without experimental
treatments. Statistics is a set of techniques for comparing treatment
differences against natural variation. Experimental designs allow a team
to analyze trial data using statistical techniques. Different designs


result in different layouts of treatments within each farm, and from one
farm to another. The different types of designs each have advantages and

Unit IV. How to Carry Out the Trial?

Dialog with farmers is needed to determine where to place treatments in
fields. Timing of planting is critical. Some data gathered from on-farm
trials are different from on-station trials. Which data to collect, how to
collect it, how to sample, and recording and handling all involve different

Unit V. How to Manage and Administer FSR/E at the Field Level?

There are many commonly encountered management and administrative
problems which interfere with the smooth implementation of the FSR/E
approach. Recognizing the major areas of problems encountered by
field-level practitioners will help in actually implementing a FSR/E




1. Ways to Decide Which Problems to Test First for Answers
2. Ways to Select Cooperating Households and Farmers
3. Ways to Classify Trials


Should be familiar with information in Volume I Diagnosis in FSR/E.


Agricultural research assistant
Extension technology verification technician


After completing this unit the participants will be able to:

1. Identify 3 types of information which may affect prioritization of
farmers' problems for design.

2. Identify 3 characterizatics of farm households which a team should
assess in choosing trial cooperators.

3. Explain 3 ways to classify trials, list resulting trial types in each
classification system, and identify key issues to consider in design
and analysis of each resulting trial type..


1. Prioritization of farmers' problems for research is the link between
initial diagnosis and design of on-farm experimentation.

2. A team should choose cooperators who belong to the research demain of
the problem which the trial will focus on, who are responsible for the
specific enterprise targeted, and who are willing to cooperate in the

3. Trials may be classified by:

a. use of product and scale of production chance/comsumption vs.
b. basis of the production system (predominant crop or animal and
cropping pattern or feeding system); or
c. basis of the problem (variety or breed, plant nutrition, plant
protection, cultural practices, or feed resources).

Volume II: I,A
page 1



1. Ways to Decide Which Problems to Test First for Answers
2. Ways to Select Cooperating Households and Farmers
3. Ways to Classify Trials


Should be familiar with information in Volume I Diagnosis in FSR/E.


Agricultural research assistant
Extension technology verification technician


After completing this unit the participants will be able to:

1. Identify 3 types of information which may affect prioritization of
farmers' problems for design.

2. Identify 3 characterizatics of farm households which a team should
assess in choosing trial cooperators.

3. Explain 3 ways to classify trials, list resulting trial types in each
classification system, and identify key issues to consider in design
and analysis of each resulting trial type..


1. Prioritization of farmers' problems for research is the link between
initial diagnosis and design of on-farm experimentation.

2. A team should choose cooperators who belong to the research demain of
the problem which the trial will focus on, who are responsible for the
specific enterprise targeted, and who are willing to cooperate in the

3. Trials may be classified by:

a. use of product and scale of production chance/comsumption vs.
b. basis of the production system (predominant crop or animal and
cropping pattern or feeding system); or
c. basis of the problem (variety or breed, plant nutrition, plant
protection, cultural practices, or feed resources).

Volume II: I,A
page 1

agricultural production system
analysis of variance (ANOVA)
arcsine transformation
border rows
causal agent
complete factorial
control treatment
error mean square (MSE)
interaction effects
logarithmic transformation
main plot error
natural variability
normal distribution
on-farm experimentation
probability distribution
production system
randomized complete block design (RCBD)
research domain
researcher planned experimentation
response curve
split plot arrangement
sub plot error
superimposed trial
test plot



Prioritization of farmers' problems is the link between initial
diagnosis and design of on-farm experimentation. The preliminary result of
the diagnostic process is problem identification. The diagnostic stage has
been discussed in detail in Volume I of the FSSP training units. Unit I:IX
specifically addresses problem identification.

Usually, initial diagnosis reveals more farmer problems than a team can
address at one time. Setting priorities for research problems is necessary
and will determine the types of testing needed to answer questions related
to the priority problems. Research priorities should be a function of the
farmers' situation and available information. The farmers'pre-accept-
ability of a triel, management practices and production system are among
the items to consider which may affect the priority setting decision.

Farmers' pre-acceptability and understanding of the treatments and
objectives of the trial are determining factors in designing meaningful
trials. Farmers should be informed about the variables considered for
testing and consulted on probable degree of acceptance in case some of the

Volume II: I,A
page 2

variables prove experimentally to be good alternatives. Based on their
response, the team is in a better position to define variables and levels,
and to promote the new alternatives if this is justified after testing
them. This is discussed in greater detail in (II:IV,A and B).

Management practices may affect the decision of whether or not to
conduct a given type of trial. These should be carefully investigated
before proposing a trial design. Existing management practices should be
followed as much as possible in the design and management of the trials.
For example, if the trial is designed to evaluate bean varieties and
farmers plant beans as a relay crop after maize, the bean trial should be
planted in a maize field at the same time that farmers plant their
commercial bean crop. Conducting livestock trials on farms maynj e limited
by certain management systems. For example, testing of multiple treatments
on the same farm is impossible when only one or two animals are raised.
The farm setting can also affect the decision to conduct a given type of
trial. One example would be the land tenure of the farm households in the
domain. A soil conservation oriented experiment might not be attractive to
farmers that do not own land and rent land to farm.

The production system of the identified research domain serves as the
basis for grouping components according to the researchable objectives. As
much as possible, experimental designs should combine components in
exploratory trials to study new variables and practices not used in the
production systems. As the research process advances and experimental
alternatives have been reduced in number, they should be handled
independently, in order to obtain socio-economic data and avoid confounded
results that are difficult to interpret. For example, an exploratory trial
could include two levels each of nitrogen fertilizer, planting distance,
and weed killer application in a 2 factorially arranged, randomized
complete block design, while at a later trial, three levels of nitrogen
could be tested in larger plot sizes that would allow a cost/benefit
analysis of treatments. In the first case, plot size would be smaller but
more replications would be recommended.


In selecting trial collaborators, the field team should first choose
households representative of the research domain intended for each trial.
A research domain is a problem focused environmental (agroecological and
socioeconomic) range throughout which it is expected that hypothesized
solutions to a defined problem could have potential applicabilty.

Care must be exercised to ascertain which of the household members are
decision makers. Trial cooperators should be selected from among those who
are responsible for the specific enterprises targeted for trials. It is
always important, however, to involve the whole household in the evaluation
of a new technology. Decisions to adopt a new technology may be made in
consultation with members not directly responsible for the production
activities. For example, in some locations women may not be involved in
producing a certain food crop, but since they process it and prepare it,
their preferences may decide whether new varieties are acceptable or not.

Volume II: I,A
page 3

In some instances, the most representative farmers may not be willing
to cooperate. In such cases, the FSR/E team may choose from among the most
cooperative farmers in an attempt to attain the desired representation.
More detailed information on trial cooperators is given in (II:IV,A and


a. By Use of Product and Scale of Production
The size and complexity of the production system as well as the
intended -ses of the products determines certain aspects of the trials.
These in turn have direct bearing on the type of evaluative analysis that
might be considered. Figure 1 shows a division of production systems based
on size and uses and lists common types of analysis.

Figure II: I,A-1 Consumption and Nutrition in FSR/E Design, Testing, and


Focus of
Design and

of Trial

Consumption/Nutrition- Non-Consumption/
Related Problem Nutrition Problem


Nutritional Impact
Other Household
Economics Impact
(Inputs, Labor)

Tree and

Nutritional Impact
Economic Impact


Tree and

Economic Impact

*(adapted from Caldwell, 11/86)

Volume II: I,A
page 4

b. By Basis of the Production System

Farmers' production systems help make-up the physical setting in which
trials will be set. The term "production system" is an abbreviation for
"agricultural production system". The agricultural production system
consists of the crop and animal production activities of the farming
system. Whereas diagnosis and evaluation consider the entire farming
system, including household, non-farm, and off-farm production and
consumption activities, design and testing of interventions are generally
based on the agricultural production system. Farm households in a given
research domain share the same production system, and have similar problems
and researchable priorities in that production system.

The following classification framework can help the multidisciplinary
team accurately identify the predominant production system and thereby
adapt trials to farmers' existing practices. Two major components of the
production system need to be considered. First is the predominant type of
crop or animal that forms the basis of the production system. (This
collection of units places greater emphasis on crop-based production
systems). In crop-based production systems, the second major component is
the cropping pattern of the predominant crop. A second major component in
an animal based system is the feeding system and feed resources. The
different combinations of the first component, the basis of the production
system (for example, maize), and the second component, the cropping pattern
(for example, intercrop), identify different specific production systems
(for example, maize intercropped with beans).
The FIGURE (II:I,A.2) shows a matrix with examples of how the two
components can be combined to identify five different specific production
systems. First, predominant crop bases include: cereals, legumes, roots
and tubers, fruiting and leafy vegetables, ornamentals, other field crops,
feed crops, pastures, trees, and others. Second, the predominant crop type
in turn may be grown in one of the following different cropping patterns:
sole crop, relay, intercrop, rotation or crop/animal mixture (where the
crop is the main activity). If the animal is the main activity, the mixture
is referred to as an animal/crop mixture.

Volume II: I,A
page 5

Figure II:I,A.2

rPnA "

C =

Matrix Depicting Five Production Systems as Combinations
of Predominant Crop Bases and Cropping Patterns


Rice Legume
Tomato/Sugar Cane

The results of diagnosis have indicated which specific component or
components of the predominant crop and its cropping pattern are the
researchable priorities for trial design. Treatment selection and choice
of experimental design focus on those specific components. Other
components can affect treatment responses and evaluation. The different
types of components to consider include the following:

Plant nutrition (fertilizer)
Plant protection
Cultural practices
a. Spacing
b. Planting time, intercropping, and crop rotation
c. Crop care

Volume II: I,A
page 6

Monocrop Relay Intercrop Rotation Crop/Livestock

Cereals __ A .

Legumes C _

Roots and

Vegetables D _


Other field


Feed crops E


Other _

5. Residual products
6. Feed resources

c. By Basis of the Problem.

The most commonly identified researchable priorities include variety
evaluation, plant nutrition, plant protection, other cultural practices,
and feed resources.

1. Variety or Breed Evaluation

Testing improved genetic material is a common type of research in
farmers' fields. The following five considerations are important in
variety testing:

a. Control treatments should include the recommended variety for
the region as well as one or more local materials used by the farmers.
Comparison of experimental varieties against these controls enables the
team to reach more meaningful conclusions. This is because the team is
interested not only in identifying higher yielding varieties, but also in
identifying other characteristics that farm household members consider in
their assessment of varieties.

b. The farmers' own agronomic practices should be respected. The
main objective of on-farm evaluation of new varieties is to know their
potential under real farm conditions. Therefore special "experiment
station" handling of these trials should be avoided.

c. Experimental varieties selected for testing should include as
many available alternatives with a theoretical potential of excellence as
is possible for the trial to accommodate. This means that the experimental
varieties from national research should be tested along with local
materials, varieties from private research programs, and material from
neighboring nations, regional or international centers, and private seed

d. Randomized complete block design (RCBD) is the experimental
design most often appropriate for these types of trials. With a large
number of varieties, incomplete block designs (IBD) may also be useful.

e. Each experimental unit should be protected from environmental
bias coming from growth habit of neighboring varieties. In maize, for
example, where varieties may differ widely in plant size, extra rows of the
same variety at each side of the experimental unit should be added. Those
border rows are not harvested for experimental purposes. A common practice
in maize is to plant four rows of each variety but only use the inside two
rows for harvest area and yield measurements.

2. Plant Nutrition

Fertilizer response trials are commonly conducted as site-specific
experiments. Information on soil characteristics, previous management, and
soil analysis should be determined before specific placement of the trial.
Generally, at least three levels of each factor should be considered in

Volume II: I,A
page 7

order to estimate a response curve. Experimental designs should allow for
measurement of interaction effects which are common in fertilizer response
trials. Completely factorial arrangements offer a better estimate of
interactions among factors than split-plot arrangements. The reason is
that, in analysis of variance (ANOVA) for a completely factorial
arrangement, the error mean square (MSE) is estimated with more degrees of
freedom. The split-plot arrangement has the same number of degrees of
freedom for interactions as the completely factorial arrangement, but
residual degrees of freedom have to be distributed between the main plot
error and the subplot error.

Special care must be used in field design to avoid fertilizer runoff
effects from adjoining plots. Border rows or ample distance should be
established between experimental units.

Control treatments should reflect local practice. .When the local
practice is not to use fertilizer, the control treatment should reflect
that practice. When farmers' practices include some fertilizer use, the
control treatment should not be a zero check, but should be based on the
level used by farmers.

3. Plant Protection

Evaluation of insect, weed, and disease problems is more difficult than
the other types of trials discussed so far. The main reason is that causal
agents vary in intensity and mode of action, not only from year to year,
but also within the plot area. Therefore, pest management trials require
large experimental units with many replications, repeated for various crop
cycles or seasons. Either complete factorial arrangements or split-plot
arrangements in randomized complete blocks can be used. Frequently, a pest
management trial is established by superimposing it on farmers' fields.

The probability distribution of counts or percentages of pest or
disease damage does not usually follow a normal distribution. Sample data,
therefore, need to be-transformed in order to approximate the normal
distribution, which is a theoretical requisite for the valid use of common
statistical procedures. The most frequent transformations for these kinds
of data are: (1) logarithmic [log(X) for count data that cover a wide range
of values but have no zeros, or log(X + 1) when zero values are present];
(2) square root of x for values consisting of small whole numbers or
percentage values between either 0% and 30%, or 70% and 100%, and square
root of (X + 0.5) for small whole numbers with zeros; and (3) the angular
or arcsine transformation for data with percentage values that overlap the
ranges of 0-30, 30-70, and 70-100%. For percentage data within the range
30-70%, no transformation is normally necessary.

4. Cultural Practices

These include different between-the-row and in-the-row spacing,
planting times and sequences, crop care practices, and water management.
When several typical farmer practices are to be compared, a superimposed
field trial may be appropriate. On the other hand, when some treatments
will be dramatically different from typical practices, conventional field

Volume II: I,A
page 8

trials should be established. A split-plot arrangement can be appropriate
when working with more than one variable. For example, when one variable
requires different row arrangements, or if there is a large border effect
and the experimental unit size is large, that variable can be assigned to
the main plot. The other variables, such as in-the-row spacing, varieties,
or secondary crop alternatives can then be assigned to the subplots.
Precision will be greater for the variables in the subplots because more
degrees of freedom are associated with subplots than with main plot error.
An economic interpretation of these types of trials is mandatory since the
crops involved generally have different market values, making individual
yields alone an inadequate criterion of measurement.

5. Feed Resources

Poor nutritional status of animals is the major cause of unsatisfactory
productivity. The testing and evaluation of new feed resources are often
undertaken when conducting livestock trials. Low numbers of animals
available for the trial and the difficulty in grouping within farm animals
for the evaluation of two or more treatments limits testing to one
treatment per farm. Therefore, many farms are needed to detect treatment
differences. Likewise, more variation usually exists within farm -
animals than between farms. More information about treatment differences
can be obtained when utilizing this type of design. To adequately compare
treatments with local practices or feeds, the control diet must be
quantified over the length of the trial. Many times this becomes difficult
to accomplish. There may be large variations in the basol diet from both
within and between farms due to seasonal and managmnt differences. The
team must monitor the diet's composition frequently, or strictly regulate
the diet ofthe control animals in order to accurately determine treatment

Volume II: I,A
page 9



1. What Alternate Pathways are Possible for On-Farm Experimentation?
2. How Researcher-Planned Trials Change Over Time.
3. Ways to Allocate Resources and Responsibilities


Agricultural research assistant technician
Extension technology verification technician


After completing this section the participants will be able to:

1. Identify 3 conditions which must be met for on-farm experimentation to
be useful.

2. Describe the sequential trend of researcher-planned trials over time.

3. Identify 5 types of activities which should be programmed to assure
successful on farm trials.


1. Two basic pathways in on-farm experimentation are: a) observing and
making inferences from farmer initiated experimentation, and b)
observing and making inferences from researcher planned

2. Researcher-planned trials change over time following a general
sequential trend.

3. Planning activities are necessary both among institutions and within
the field team to assure successful on-farm trials.



The purpose of on-farm experimentation is to test alternative
production practices to solve problems identified by farm households.
Three conditions must be met for on-farm experimentation to be useful:

-Focus on real problems to which farmers want answers.
-Compare alternate practices under real farm conditions, and
-Enable farmers to predict the likelihood that alternate practices will
give improved results.

Volume II: I,B
page 11

The first condition means that the team has understood farmers' needs.
Many techniques allow researchers to do this. These techniques are called
diagnosis. Volume I discusses these techniques in detail. A team always
begins with diagnosis, but diagnosis never ends. As it discovers more
about farmer problems, the team may need to use diagnostic techniques in
the middle of on-farm experimentation. In fact, on-farm experimentation
itself is partly a diagnostic technique. This is because a team improves
its understanding of farmer problems by observing how farmers compare
practices the team suggests with their own traditional practices.

One diagnostic technique is to identify experimentation that farmers do
themselves. For example, a team may find that some farmers are planting
several varieties side by side or intermixed. Or, a team may find some
farmers introducing new crops in an alley cropping pattern. A team should
pay special attention to what better farmers are doing. Often, they will
use one edge or a part of a field to try something new that they have
thought of themselves, or heard about. These farmer-initiated changes can
suggest ideas for more on-farm experimentation.

The second condition means that the team distinguishes between changes
to be tested, and changes not to be tested. If a change is not for
comparison purposes, it should not be made. Otherwise, the changes to be
tested will be tested in the presence of other changes. Those changes will
make test plots or animals different from farmer plots or animals. For
example, if the team wants to test varieties, but farmers do not use
fertilizer, test plots should also not use fertilizer. By-not using
fertilizer on the test plots, the team can compare the new varieties with
farmer varieties even outside the test area. Farmers can see if the new
*varieties are different than their own under the same conditions their own
varieties face.

This condition is likely to be met if farmers themselves initiate the
experimentation. The team has to be careful not to violate this condition
when it plans experimentation.

The third condition means that the team understands variability. There
will always be variability within farmers' fields. For example, consider
two rows or two plots side-by-side, planted at the same time, with the same
varieties, and grown the same way. Each plant will not have exactly the
same height, nor will the yield from the two rows or two plots be exactly
the same. The same would be true for two animals born to the same dam
(mother). It is unlikely that they would weigh the same, for example.

Now consider different farmers' fields of the same crop (or crop
combination), or animals born to different dams of the same breed. The
variability will be even greater. All this is natural variability which is
present even without on-farm experimentation.

In on-farm experimentation, farmers and the team are looking for
differences among alternate practices. Suppose the team does find
differences. The question then is, are the differences among alternate
practices greater than differences just due to natural variation? What
inferences can farmers and the team draw from the experiments? The answer
is not definite.
Volume II: I,B
page 12

The team and farm household members want to know what the chances are
that the observed differences are really due to the alternate practices.
Perhaps there is only a 50:50 chance. Are farm households willing to
gamble on a 50:50 chance of an alternate practice being better? Or are
they unwilling to gamble unless the chance is higher? How much higher does
the chance have to be for them to be willing to gamble, and, furthermore to
plant even one-tenth of a field using an alternate practice? Because of
the high value of livestock, farmers need a higher probability of an
improved response before accepting a new or improved technology. Related
to this is te question of how large a difference are farm households
willing to gamble on. Will they take a 50:50 gamble on doubling yield, but
will gamble on a 25% increase only if the chances are 4:1 of getting that

The cost of the gamble also affects the chances that farm households
are willing to take. Farm households may be willing to gamble on the cost
of a packet of a seed at 10 units of money, even if there is only a 50:50
chance of increasing yield by 25%. This is especially true if they have to
purchase seed anyhow. On the other hand, they may be unwilling to gamble
100 units of money on fertilizer if there is only a 50:50 chance of a 25%
increase. They may want a 4:1 chance of increase before they are willing
to gamble that much scarce cash.

With variable soils, variable management, and variable rainfall, what
is the chance of that 100 units of money gamble paying off? Here is where
on-farm trials on farmers' fields throughout the area can help give
answers. Many farm households cooperate together through the on-farm
trials to spread the risk of determining what those chances are.

The team plays an important role here. First it helps pool all their
results. No one farmer could afford the time and cost to travel all over
the area to pool all the results from many farms.

Second, the team also brings specialized knowledge to help estimate the
chances of various percentages of increase based on the pooled results.
That specialized knowledge is called statistics.

Everyone together bears a little of the cost of the team's time and
expenses in pooling the results. Everyone does this through public funds.
Public funds are from the taxes that everyone pays. Thus, even people in
town and in the city help support the cost of the team.

Statistics is a set of techniques to determine the chance that
differences are real, and not just due to natural variation. Statistics
consists of many techniques. A team needs to choose the best techniques
for each situation.

What are the choices among statistical techniques? First, there are
two basic pathways in on-farm experimentation. Each pathway leads to use
of different types of statistical techniques. The two pathways are:

-Observing and making inferences from farmer-initiated experimentation.
-Observing and making inferences from researcher-planned

Volume II: I,B
page 13

Following the first pathway, the team does not plan the
experimentation. There must be enough farmer-initiated experimentation to
allow meaningful comparisons to be made among several farms. One useful
technique is to group farms based on similar comparisons. Techniques to
allow researchers to do this are currently under development. One such
technique is called cluster analysis.

Following the second pathway, the team plans the experimentation. The
team may add treatments to farmer-initiated experimentation. Or, the team
may design the experimentation and seek farm households willing to
collaborate. The design may involve simply doing some practices
differently in fields farm households have already planted. Such
experimentation generally consists of superimposed trials. For example,
the team may test several different times or methods of weeding. Or, the
design may involve the team and farmers together putting out a test plot
from the beginning of the season.

Statistical techniques for the second pathway are different than for
the first. While techniques for the first pathway involve grouping
observations, those for the second involve testing planned comparisons
against one another. There are many ways to test planned comparisons.
The best technique to use depends on many criteria. The rest of this
volume presents some of the most important criteria and some of the most
widely-used techniques from which practitioners can choose.


a. Trial Function in the Research Extension Process

According to their function in the research/extension process, trial
types follow a general sequential trend. For each type of trial, specific
designs and types of analysis are common. Units III. A and C in.this
volume describe the different designs referred to here in detail. Units
III. A and B in Volume III describe the different analysis procedures
referred to here in detail.

1. Exploratory testing

These are trials conducted when little is known about the domain or
about possible treatment effects in the domain. They can be complementary
to, or part of, the characterization of the domain and usually precede
refinement trials. These trials normally assess qualitative effects of
several factors, rather than quantitative effects. Unit II.A in Volume III
explains these types of effects in more detail. Frequently, two levels of
each factor are included and few replications are used. The most common
designs used include the 2 factorial and "add-on" or "take-off" trials.
This type of trial can sometimes be superimposeE on farmers' fields without
the necessity of special preparation of the experimental area.

2. Refinement testing

Two kinds of trials can be included in this stage: site specific
trials and regional trials.

Volume II: I,B
page 14

Site specific trials are trials done on only one farm. They often
focus on quantitative effects. They are similar in design to conventional
trials, but usually fewer treatments are involved. Perhaps as many as 20
to 25 treatments can be included, although this is not recommended unless a
more complex type of design (e.g., a lattice or Latin square) is used to
keep the experimental error at an acceptable level. Because of the
requirement for intensive researcher's management, only a few of these
trials are normally conducted in a given domain. The most common design is
randomized complete blocks (RCBD) with four replications.

.Regional trials are trials done on more than one farm, but analyzed as
one set of data. They are amenable to both agronomic and
agro-socioeconomic analysis. They are designed to expose the best
treatments from site-specific trials to a much wider range of environments
within a domain. Perhaps six treatments may be included, and five to ten
sites can be utilized. A recommended design is randomized complete blocks
(RCBD) or incomplete blocks (IBD) with two to four replications per site.
ANOVA, regression, or modified stability analysis can be utilized.
Combined analysis with site as a source of variation can be used in ANOVA
to quantify treatment-by-environment interactions.

3. Validation testing

These trials provide the opportunity for the farmers themselves to
manage and the farm households to evaluate the one or two most promising
interventions identified in refinement testing. Large plots with no
replications within farms are used. The purpose of these trials is for the
farmers to compare the interventions with their own practices, so one plot
with existing practices can be included in the design. This individual
farmer control plot serves the researchers more than the farmers, because
the farmers will be able to evaluate results based on their own fields. If
researchers wish to measure results of the farmers' own practices, they can
also sample the farmers' fields. However, agronomic and economic records
of the farmers' practices must be kept to provide the necessary
information. If possible, it is desirable to have at least 30 farmers
conducting these trials in a given domain, although sometimes as few as 10
farmers may be acceptable. The larger numbers improve the precision of the
evaluation of the degree of acceptance by farm households of the new

b. Researcher-Farmer Management Sharing

The relative participation of the multidisciplinary research team and
farmers in conducting trials leads to another classification that will
influence the number of trials of each kind in a given time and resource
situation. There is a close correlation between management type and trial

1. Researcher Planted/Researcher Managed

This category includes those trials that represent a high economic risk
to farm households because of the unpredictable or unknown behavior of
intervention treatments under farmer conditions. Normally these trials
Volume II: I,B
page 15

would either be conducted in the experiment station, or if planted in a
farmers' fields, the total cost of labor and inputs should be covered by
the project. These trials are most common in exploratory and refinement
testing. An example would be testing an array of new weed killers.

2. Farmer Planted/Researcher Managed

This category includes "superimposed" trials where treatments are
placed on fields which have already been planted and are being managed by
the farmers themselves. Treatments are marked by stakes or other means, and
individual treatments are installed either by the researchers or the
farmers. Together, researchers and farmers harvest the crop when it is
mature. The design of a superimposed trial should be simple. Replications
should be used at each location, although data from designs without
replications at each site can be combined for regional analysis and
interpretation. These trials are also most common in exploratory and
refinement testing, for example, fertilizer sidedress application in a
maize field.

3. Farmer Planted/Farmer Managed

Trials completely handled by farmers must include the following
characteristics: a) technology must be simple enough for farmers to
comprehend and manage; b) farmers must use their own resources so they can
understand all implications of the alternatives; and c) design of the trial
must be simple enough that farmers can observe differences in treatments
and/or measure them, with their own means of measurement. These trials are
the most common in validation testing. An example would be testing of a
new cultivar under the farmers' normal planting and cultivating procedures.
The farmers pay all their, usual costs plus the cost of the seed of the new


a. Station and on-farm linkages.

Trials established on the experiment station and trials established on
farmers' fields are not substitutions for one another, but rather
complement each other. II:II,A describes this linkage in more detail.

1. Station

Basic research trials are probably the only type of trial that should
be planted solely on station. Other types of trials can be established on
both on the station and on farms. For example, trials of little known
variables or treatments may be properly handled on either the station or as
a researcher-planted, researcher-managed on-farm trial. Also, in a series
of trials to expose treatment to a wide range of environments, the station
can represent a "good" location to be considered in a combined analysis of
results from various locations.

Volume II: I,B
page 16

2. On-Farm

Homogeneous or uniform experimental areas are the rule rather than the
exception on the experiment station. The opposite is true on farms.
Nevertheless, agronomists can reduce experimental error on farms by
following a few common sense rules. For example, it is never wise to
locate a research area adjacent to a house unless that is the environment
in which the crop in the trial is going to be planted normally. Likewise,
paths, canals, large trees and other conditions which are not a part of the
environment for most of the crop should be avoided. If the crop is usually
planted in these special environments, of course, it is appropriate to
locate the experimental area in them. Likewise, if farmers' fields have
significant areas with poor environments, the team should not avoid those
areas and only use the best parts of the fields. (Unit IV,B) describes
there considerations in more detail.

The number of trials of each type of experiment will be related
inversely to the relative participation of the team in each case. The more
the research team's control, the fewer the number of trials. In cases of
farmer planted, farmer-managed trials in the validation testing, the number
of trials may exceed 50 per research domain, while trials on the
exploratory testing may be only 3-5.

The planning of the activities and the personnel involved in conducting
the trials needs to be well defined, financed and managed. This is
discussed in more detail in Unit V of this volume.

2. Planning Activities at the Inter-institutional and Field Team

Five types of planning activities should be programmed among and within
the institutions involved to assure successful on-farm trials.

At the inter-institutional level; 2 activities are essential each year;
Work plan session
Results presentation sessions

At these sessions, multidisciplinary and inter-institutional
representatives should be present in addition to the field team members who
carry on the main responsibilities. Specific working sessions with
research program personnel, extensionists and other interested persons
should be programmed within the period of activities. By institutional
planning, specialists from different disciplines and institutions are led
to interact. If this interaction does not have a committed framework, by
strictly voluntary participation the motivation and continuous level of
participation tends to weaken.

At the field team's level 3 activities are essential. First a detailed
schedule of activities should be prepared for the specific assignment of
each member, giving a timetable indicating the beginning and completion
date of each activity. See Unit VI or more details.

Volume II: I,B
page 17

The administration of personnel and the availability of in ts required
to conduct the trials programmed should also be indicated in te overall
schedule of activities. Although ideally the multidisciplinary field teams
should include personnel from the various biological sciences (agronomy
and/or horticulture, plant protection, animal production, etc.) and social
sciences (economics, sociology, anthropology, geography, etc.), this is
seldom possible. Therefore, an approximation to the ideal situation
becomes the logical alternative. This implies that the actual members of
the field team need to include activities in the missing disciplines) to
the best of their abilities.

Volume II: I,B
page 18



1. The Ultimate Success: Acceptability by Farmers
2. The Importance of Planning Ahead
3. Key Factors to Consider When Determining Evaluation Criteria
4. Identifying Relevant Evaluation Criteria
5. Weighing the Importance of Different Criteria


III:I Analytical Framework
II: IV,C How to Design Trials to Obtain Analyzable Data
Hildebrand, P. and Poey, F., On-farm Agronomic Trials in Farming
Systems Research and Extension, 1985, pp. 9-12.


Extension technology verification technician
Agricultural research assistant


After completing this section participants will be able to:

1. Explain why it is important to develop an understanding of
farmer perspectives and circumstances when identifying evaluation,
criteria for analysis of technological alternatives.

2. Explain why it is important to examine and understand the farm
production goals, roles, and strategies of individuals as well as

3. Discuss important factors which influence the farm production
incentives, goals, and strategies of individuals and households.

4. Identify evaluation criteria for analysis of on-farm research
which are relevant to different stakeholders and situations in farm


1. Developing an evaluation plan before trial design helps to identify the
types of data which need to be collected in trials over time.

2. Selecting evaluation crit ria which are relevant to different
farm-household and individual goals and perspectives is critical.
Poorly chosen evaluation criteria lead to wrong conclusions about the
viability of the alternatives being compared.

Volume II: I,C
page 19

3. Any farm enterprise is the outcome of the efforts of various
individuals who participate in differing ways or have a stake in the
outcome (decision-makers, investors, beneficiaries, stakeholders).

4. Different stakeholders in farm production judge proposed changes in
farming practices by different evaluation criteria.

5. Important factors to consider in understanding farm production goals
and strategies include individual and household consumption
preferences, resources, other farm and non-farm production activities,
and the social, cultural, economic, and institutional environment
outside the household.

6. Risk, or the chance that returns might fall below some minimum
acceptable level, is faced by every farmer. Likewise, risk must be
considered by every researcher or extensionist who evaluates
alternative technologies and who makes recommendations to farmers. In
a true subsistence situation risk of starvation may outweigh all other
decision criteria.

evaluation criteria
ex- ante evaluation
ex-post evaluation



The purpose of FSR/E is to develop new agricultural technologies that
address identified and selected priority problems of farm households
(Volume I). The ultimate measure of the success of a new technology is the
acceptance, adoption, and sustained .use by farmers. FSR/E teams use
on-farm experimentation involving active participation of farm household
members to test alternative technologies. The analysis and interpretation
of the results of on-farm trials (see Volume III) allow the FSR/E team to
evaluate the potential success of new farming technologies and to make
specific recommendations to farmers.

Evaluating new farming technologies in FSR/E must extend beyond
determining the biological viability of the enterprise in question. FSR/E
teams must couch the evaluation of new technologies in the context of the
technology's acceptability to farmers and farm households. FSR/E teams
must consider a range of factors which may influence the farmers ultimate
acceptance and adoption of an alternative technology and establish
appropriate criteria for evaluation.

The purpose of this section is to help identify the wide range of
factors to consider when determine how to evaluate the success of a trial
emphasizing the critical involvement of the farmer in the evaluation
process. FSR/E depends on farmer participation in the initial diagnosis

Volume II: I,C
page 20

through design and continued characterization and ultimate in the
evaluation of the technologies.

Establishing appropriate evaluation criteria based on identified
problems in a farming system will help the FSR/E team to better evaluate
alternative technologies with a farmer's perspective. (One analysis
technique for the evaluation of acceptability by farmers is discussed in
Volume III:).

An ample understanding of farm household goals, incentives, farming and
non-farming activities, available resources and constraints is the
foundation upon which to build the design and analysis of on-farm research.
Much of the knowledge will be generated during the diagnosis phase of FSR/E
(see Volume I). Given this understanding, and full farmer participation,
appropriate evaluation criteria and procedures for analysis can be
selected. Although rarely is this insight easily achieved, the usefulness
of analysis depends upon doing so. Poorly chosen evaluation criteria lead
to wrong conclusions about the viability of the alternatives being

a. Seven Key Questions in the Farmer's Evaluation

Rhoades suggests that the following seven basic questions can help
FSR/E practitioners to come closer to their clients by trying to "think
like a farmer." -A fuller explanation of the implications of these
questions can be found in Rhoades' excellent article "Understanding
Small-Scale Farmers," Journal of Agronomic Education 13:64-68, 1984. Some
discussion of these questions is presented below.

1. Is the Problem to be Solved Important to Farmers?

Scientists sometimes project their values or preferences too much into
the farmer's circumstances. What may be scientifically important may not
be important to farmers. Farmers may cooperate with on-farm
experimentation because they feel they have to, and not because they view
the problem as a high priority. If this is the case, the experiment may be
successful, but no one will adopt the solution. On-farm trials are an
excellent way to test whether a problem really is a problem for farmers..

2. Do Farmers Understand the Trials?

This question raises others. Was the trial clearly explained? Was the
number of experimental variables too large? Were there too many
replications? Was the technology too complicated or sophisticated? Did
farmers understand the utility of the new technology? "Technologies which
build on existing, traditional practices will probably stand the best
chance of being understood."

3. Do Farmers Have Time, Inputs, and Labor Required by the Improved

Practioners carrying out on-farm trials must always consider the
logistics of the trials from the farmer viewpoint. Farmers do not have
research stations or projects to supply inputs, additional labor, or

Volume II: I,C
page 21

vehicles to carry supplies to the field or purchase them in town. Farmers
weigh each new technology by the resources they control or to which they
have access. These resources include land, labor and capital. Within the
household, there may be competition for these resources. Farmers may have
sufficient land, but lack the capital or credit for the inputs needed for
new technology. More subtle is the question of labor and time. Farmers
may have to weigh allocating family labor to the new technology against the
time already needed to collect fuel and water, herd animals, collect feed
for livestock, scare birds from fields, weeding or many other tasks which
must be completed. FSR/E practitioners should not assume that because a
household member is not engaged in agricultural tasks, that their time is
free, because many other tasks must be completed to sustain the household.

FSR/E practitioners must also remember that households are not
egalitarian units. Resources and benefits are not always distributed
equally. Differences in access and control over resources and benefits
often exist in terms of age and gender. Children, unmarried adults and
elderly members of the household, even if they farm their own fields, often
do not have the same access to farm inputs such as seed, fertilizers or
animal traction, as the head of the household and his or her spouses) do.
More striking and important in the design and potential adoption of new
technology.is the fact that gender often distinguishes the access or
control of a farmer over the resources needed to farm or the benefits
gained from farming. Women are often denied access to credit and
membership in cooperatives which supply seeds, fertilizers or machinery.
Women are often overlooked by extension services even when they are the
primary farmers of a household. Women may have great potential to gain
from new technology but are denied the ability to use it. On the other
hand, women's needs for technology may differ from men's and therefore new
technology may be inappropriate or fail to address their production

The key point to remember here is when asking the question whether
farmers have time, inputs and labor required by the improved technology,
practitioners should always ask further if all farmers (young and old, male
or female) have the needed requirements. IT~not, should adjustments be
made in the proposed technology?

4. Does the Proposed Technology Make Sense Within the Present Farming

A change in one part of a system, caused by the introduction of a new
technology, will cause changes in the rest of the system. Will the new
technology or proposed changes fit within the system? Will it cause
changes in other areas which will have a negative impact on farmers? Will
it negatively impact certain farmers (women, children, elderly)? Will it
fit within the existing rhythms of production, such as the time needed for
harvesting major cash crops or times when family labor is moved to a
different agricultural zone to work for wages? Understanding whether a
proposed technology will make sense requires close examination, probing
within the system, observation and talking with farmers. Again, a
technology may make perfect sense to the scientist who conducts an
evaluation at field level, but it may make no sense at all in terms of the
whole farming system.
Volume II: I,C
page 22

5. Is the Mood Favorable for Investing in New Technologies or Crops
in a Region?

Rhoades points out that "this question suggests understanding farmers'
orientations toward investment or innovation in crop production brought
about by broader economic conditions. If trials are conducted when prices
have hit rock bottom and have stayed there for two or three seasons,
promoting changes could be a losing battle. Even if farmers believe a
change may be beneficial, they may respond with general pessimism."

6. Is the Proposed Change Compatible with Local Preferences,
Beliefs, or Community Sanctions?

FSR/E practitioners should remember that taste or color preferences of
foods, superstitutions and ceremonies are as important to farmers as they
are to everyone else. Rather than viewing these as quaint or as obstacles,
practitioners should see where they fit in the farming system as a whole.
As Rhoades points out, "planting days tied to religious festivals may be an
ingenious way of guaranteeing that work is done by a certain day."
Technologies designed to take these aspects into account are more likely to
be acceptable, thus facilitating FSR/E work rather than hindering it.
Preferences for food color, shape, size and taste must always be considered
in the design of agricultural technology and can be critical in determine
the evaluation of technology. The social science perspective in the
evaluation may uncover that though the technology increased.yields by 200%,
the color of the new variety was unacceptable to farmers and consumers and
therefore no one was interested in adoption. Finally, farming systems are
linked to larger community and government systems. The evaluation of new
technology for the farming system must always consider the potential impact
on the community at large and whether government may pose restrictions on
the utility of the technology. Can local marketing boards handle increased
production? Can government suppliers of credit and inputs handle increased
demand? .Will the new technology create an advantage for only some farmers
while creating a disadvantage for others?

7. Do Farmers Believe the Technology will Hold Up Over the Long Term?

Rhoades points out that "a farmer's view is normally based on long-term
needs, not on a couple of seasons, and sometimes on generations of
experience with the crop and land." Researchers and extension agents may
view a technology favorably based on the results of three or four years of
testing, but unless the records for the area are exceptionally good or they
have lived and farmed themselves in the area for a long time, they cannot
estimate how the technology will respond to the longer tests of time.
Farmers who have farmed in the area for a long time, or who have learned
farming skills from generations of farmers in the area, can evaluate
technological performance using many more criteria than researchers.
Creating opportunities for farmers, even those not directly involved in a
trial, to assess technology being tested today in the field can provide an
evaluation based on generations of agro-ecological, economic and social
criteria. Standing up against the rigors of such a test will often yield
technology far more likely to be readily adopted by farmers.

Volume II: I,C
page 23

Posing these seven questions will greatly assist FSR/E practitioners in
assuring that a social science perspective and a farmer perspective are
incorporated in the testing and evaluation of new technology. In the same
article, Rhoades also reminds us of a very important fact in evaluating

"In the end, the acceptability of a technology depends on what the
farmers actually do. This can only be discovered in a final stage of farmer
testing where farmers themselves take over the new technology and incur all
risks, costs and benefits. Until this final step is taken, all other
evaluations remain only suggestive of the technology's potential."


Planning ahead for evaluation before trial design is useful in several

a. It helps link diagnosis and design, by organizing information on
production, the roles of different household members, and their goals into
a matrix that is based on treatments proposed for the trial.

b. It helps identify which data are needed to assess the success or
lack of success of a trial. Unit II:IV, C presents guidelines for
collecting different types of crop, animal, economic, and consumption data.
An evaluation plan can help select which data in these guidelines to

To make it easier to compare on-farm trials done in different
countries, a minimum data set has been developed. This indicates data
which all on-farm trials should .collect (II: IV, C).

c. It helps assess trade-offs among different farm household members.
This can help the team make better decisions for refinement and validation
trials about which treatments to continue to test. It can also help the
team make better decisions about recommending a new practice at the end of
validation testing. Better information on the different trade-offs among
household members can also suggest recommendations for policy support and
programs in related areas (for example, reduction of the workload for
obtaining water by female farmers).

d. It helps link testing back to design. Since the treatments are
assessed for their impact on different farm household members, the
assessment of costs and benefits can be compared with the goals of the
household members. These goals as first identified in the diagnostic
informal survey were the basis for initial design. The comparison of trial
results can lead to better re-design of the next series of trials. This
linkage from testing back to design is an essential part of FSR/E.

e. It helps document long-term progress towards an acceptable solution
to farm household priority problems. The first year's trials may appear to
be unsuccessful, as treatments found not acceptable are eliminated.
However, over time, these results can be seen as a necessary step in
identifying acceptable technology. By using an evaluation framework
systematically, at pre-determined intervals for a series of trials carried
Volume II: I,C
page 24

out over a number of seasons, the value of earlier trials for later trials
can be documented.


goals of the household and of individual stakeholders
scarce resources (time, labor, cash, land, animals, etc.)
probability of returns being less than a minimum acceptable level
control and distribution of inputs and benefits
possible effects on other enterprises and on overall household
production, consumption, and welfare
wage employment opportunities
access to credit, supplies, information
community goals
cultural and social factors
government policy

Clearly, more than a single evaluation criterion may be required.

a. Recognizing Individual Roles in Farm Procedure

Any farm production enterprise is rarely the outcome of the efforts of
a single person. Household members and others participate in differing
ways or have a stake in the outcome. These stakeholders can be grouped by
the roles they fulfill, even though each often has more than a single role.
Decision-makers use management expertise and/or authority to decide what to
produce, and when and how to produce it. Investors provide resources such
as time, labor, land, capital and animal traction. Beneficiaries receive
benefit from the production activity. Examples of benefits might include a
portion of the harvest, part of the sale proceeds, or time freed from
production. It is usually assumed that beneficiaries gain some positive
outcome but frequently negative benefits are also the case.

Level of involvement in each of these production roles is often
associated with age, gender and/or position in the houseFold or community.
Since individuals in different roles have differing goals and incentives,
it is useful to consider stakeholders by their production role and also by
the socioeconomic categories of age, gender and position.

To ensure adequate consideration of individual perspectives and
circumstances farming systems practitioners must constantly ask "who?" Who
participates in the decision to produce? Who provides what resources? Who
participates in the production enterprise? Who receives the benefits of
production? This questioning helps to ensure awareness and consideration
of the needs and roles of different household members throughout the
farming systems sequence.

One way to help FSR7E practitioners to incorporate intra-household and
gender sensitivity in their evaluation of on-farm experiments will be to
acquire new analytical skills by working through the "Case Studies on
Gender and Intra-Household Dynamics in Farming Systems Research and
Extension" which form a part of the overall training package which includes
this manual. Feldstein and Poats (1985) developed a conceptual framework

Volume II: I,C
page 25

for the case studies to provide a guidelines by which information on gender
and the intra- and inter-household aspects of farming systems may be
gathered, analysed, and applied to the design of improved technologies for
agricultural and livestock systems. It covers the information necessary to
model a farming system and the process by which farmers (men and women) are
included in the research and extension activities in a given area. Some of
the key issues and questions provided in the conceptual framework are
summarized here regarding the evaluation of on-farm trials.
First of all, what are intra- and inter-household dynamics and
variables? What do they contribute to the analysis and evaluation of
on-farm experiments?

The basic notion underlying these terms is that a 'household' is not an
undifferentiated grouping of people with a common production and
consumption function, i.e. with shared and equal access to resources for
and benefits from production. Rather, individual members of households or
families share some goals, benefits, and resources; are independent on
some; and in--onflict on others. Individuals are also members of other
groups through which they may gain access to productive resources or
benefits and to which they may have obligations. Poor rural households
often depend on a number of activities, on and off farm, and alliances for
survival. Farm management decisions on any enterprise are affected by the
interplay of the roles and resources of the individuals connected with that
enterprise as investors, laborers, and beneficiaries. Thus, there are
patterns of activity within the household and between households which
relate to the ways in which members make choices and carry out activities.

What we face is complexity, not homogeneity. In a particular farming
system or a single enterprise within that system, the pattern of resources
and incentives must be discovered, not assumed. The conceptual framework
is designed to assist in this discovery.

The way the conceptual framework operates is to examine the four areas
of knowledge important to FSR/E to which a consideration of intra-household
dynamics can make a contribution: labor, non-labor resources, incentives,
and the process by which farmers are included in FSR/E. These areas are
considered for each stage of FSR/E (diagnosis, design, on-farm
experimentation and evaluation, and recommendations) by asking a series of
questions. We will consider here only those appropriate to experimentation
and evaluation activities.

Volume II: I,C
page 26

1. Labor

What changes in labor allocation, in time or task, are actually
associated with on-farm experiments? Do these contribute to or detract
from increases in productivity or income for this enterprise? Do changes
in labor allocation impact on other enterprises including household
production? Do they fit what was predicted in the design?

2. Access and Control of Non-Labor Resources

How and to whom have new resources been supplied? Who has/has not used
them? What networks of relationship or exchange have been used to garner
any additional resources needed? Can further constraints in access to
resources by particular groups be identified as result of the testing?

3. Incentives

What motivates people's decisions about the allocation of labor and
other resources to farm production, home production, and alternative uses?
What incentives/disincentives are there for farmers (men and women) to
modify practices concerning the enterprise in question? What
incentives/disincentives are associated with the particular modifications
being tested? Are there incentives or disincentives associated with being
a cooperating farmer? How do the technologies being tested affect
individual income streams?

4. Inclusion

Are women as well as men included as cooperating farmers in on-farm
research? For particular enterprises? Fields? In the management of
trials? Are they included in interviews evaluating the trials? Are there
factors which inhibit the participation of particular categories of

This framework is flexible and can be used to describe a farming
system or the variables affecting a particular enterprise. People are
often overwhelmed when confronted with a new list of questions to consider
as they analyze and evaluate a situation. The questions presented in this
section on social science and farmer perspectives are not designed to
burden FSR/E practitioners with interesting but irrelevant detail.
Instead, the purpose is to provide practitioners with the tools and skills
to better understand the nature and processes of farming systems in order
to identify better solutions to the problems confronting all farmers today.

b. Understanding Farm Production Incentives, Goals and Strategies.

The general end sought by the farm household can be considered to be
improving or maintaining the overall welfare and security of its members.
However, underlying this overall end -s a complex of individual and
household goals. Some goals such as obtaining food for mutual consumption
are common to the household, while others like increasing individually
controll-e funds may be held by individual members and even may conflict
with goals common to the entire household. Strategies are the methods
which the household uses in an attempt to achieve its goals.

Volume II: I,C
page 27

Households and individuals, considered as farm production units, are
commonly placed into two categories. Those producing for home consumption
are classified as subsistence. Those producing for sale or exchange are
considered as market or commercial. However, most farm producers actually
follow strategies which are both subsistence and commercial in nature. On
most farms, crops are grown for direct home use and market. Likewise,
livestock are raised to produce products for household consumption, and
some livestock and/or livestock products are sold.

For primarily subsistence crop or animal enterprises, the strategy
followed by families is to produce in order to meet home consumption needs.
Producing at least a minimum subsistence level of outputs is of greater
concern than gaining high yields. Common strategies for lessening the risk
of failure to meet minimum needs include intercropping, farming parcels
located in different ecological zones and micro-environments, and
maintaining mixed herds of differing aged animals. Production arrangements
frequently substitute farm produced resources, such as household labor,
fodder, manure, seed from previous crops, and so on, for off-farm resources
requiring cash purchase, such as hired labor, commercial feed and fodder,
chemical fertilizer and hybrid seed. However, there usually is a need for
a minimal cash return to ensure the purchase of essential consumption items
and some farming supplies not produced on the farm.

For primarily commercial or market crops and livestock, the strategy
followed by producers is to gain maximum returns on resources invested,
usually in the form of profit or net income. This may be done by increasing
yields, improving product quality, or changing the amount of inputs used
until maximum return per unit of land or other relevant resource is
reached. Often, commercial plantings and livestock herds are managed as
businesses, somewhat apart from household concerns. Therefore, concern for
minimizing risk is less intense than under a subsistence strategy where
failure means hunger in the family.

Household factors such as consumption preferences, resources like time,
labor and cash, and activities such as food preparation and processing are
of consequence in setting farm production goals and strategies. For
example, a decision to purchase materials for a new roof on the home might
limit the cash available for buying fertilizer. In another case, a
consumption preference for a local type of chicken leads to a decision
against raising other breeds which might provide more eggs and meat. Maize
yields might be limited by less than timely weeding, but a recommendation
requiring more time spent at weeding might not be accepted if that time is
needed in collecting fuelwood or for carrying water for the household or if
the weeds are needed for animal fodder after harvest.

Household commitment to farming is affected by other non-farm
production activities and wage employment opportunities, and the
distribution of costs and benefits within the household. Some households
farm only as a secondary activity, while deriving primary income from home
food processing activities like making tortillas or beer for sale. Others
may depend on the wages of one or more members, working part-time or
full-time either locally or as migrants. Successful evaluation of proposed
farming improvements is undertaken with full consideration of possible
effects on non-farm production activities.
Volume II: I,C
page 28

Community social and cultural factors outside the household are also of
concern in understanding farming decisions and strategies. A young man
might not be considered a full-fledged adult by the community until he
first owns land and farms it. The social prestige and standing of a family
in a community might depend upon the number of cattle it owns. In areas of
communal land holdings, a household might continue to farm a depleted field
simply to maintain control of it.

Macroeconomic factors including government policy, prices, and access
to credit, information, supplies and markets also influence farming
decisions. Obviously, when market prices for beans at harvest a
traditionally low and farmers are unable to store their harvest, few are
likely to plant beans commercially.

(The following (a-c) is adapted from Hildebrand and Poey pp. 74-78).

Identifying appropriate evaluation criteria for analyzing the results
of on-farm research is a critical step. Evaluation criteria are
biological, economic or social measures which are used to assess the
acceptability of two or more alternatives. Appropriate criteria which are
relevant to farmers must be identified. These criteria provide a basis for
comparing farmer practices with proposed alternatives and for evaluating
the results of each.

Careful farming systems practitioners begin to identify criteria by
considering each stakeholder's perspective and priorities within the
overall framework of the household.

a. Land as a Scarce Resource

The most common evaluation criterion used by agronomists is yield per
unit of land area, frequently kg/ha. The use of this criterion implies
that land is the most limiting resource on the farm and therefore that
productivity of the land is the most important evaluation criterion. This
is not always the case. On many small farms, even though there is little
land, land is not the most limiting constraint. Nor is the same constraint
necessarily the most limiting for different production activities.

For example, small farmers in Narino, in the south of Colombia,
traditionally plant their scarce potato seed by spacing it widely to
maximize the productivity of each potato seed. The amount of seed
determines the size of the potato field. Hence, land is not the most
limiting resource with respect to potato production on these small farms.
However, the rest of the land on these farms is planted into grain crops.
For grain, land is a limiting resource. For this reason, in the case of
potatoes, technological changes which increase the productivity per unit of
land area but decrease the productivity per unit of seed will not be
attractive to these farmers. On the other hand, the same kind of
technology for grain crops could be acceptable. The importance of using a
relevant criterion in evaluating on-farm trials is obvious in this case.

Volume II: I,C
page 29

b. Labor as a Scarce Resource

In some areas of Africa, land is not a limiting resource. Farmers can
plant as much land as they are able to manage. However, in these same
areas, rainfall is scarce so weeding the crops to reduce competition for
the limited soil moisture becomes a critical factor. These farmers tend to
plant the amount of land they can effectively weed because planting more
land is a waste of effort if it cannot be weeded. In this case, labor for
weeding becomes an important evaluation criterion and changes in crop
production practices must also be evaluated against this factor.

In some areas, such as eastern Guatemala, crops must be planted as soon
as possible after the initiation of the rains. Delayed planting reduces
yield heavily because of a mid-season dry spell, increased pest problems,
or because the crop does not mature before the rains terminate. In this
case, labor available for planting becomes a very important criterion.

c. Cash as a Scarce Resource

In commercialized agriculture, cash can effectively substitute for most
other inputs. If more seed is needed, it is purchased with cash (or credit
which is another form of cash). If more labor is needed, it is also
purchased with cash. However, in many small limited resource farm
situations, nearly all resources used in the production process come from
the farm. Only a few inputs are purchased. On farms where farmers are
unaccustomed to making purchases with cash, great care must be taken to
evaluate the return to the additional amount of cash required for
alternative technologies, whether even a limited amount of additional cash
is available, and if it is not available, where it will come from.

On fully commercial farms, where cash is basically not a limiting
factor, the criterion of profit maximization may be relevant. Profit
maximization is achieved when the value of the product obtained from the
last unit of input is just equal to the cost of that additional unit.
Commercial farmers will often have objectives other than profit
maximization and other constraints which will limit the fulfilment of the
profit maximizing criteria.

Farmers with very limited amounts of cash will not usually be
interested in using as much cash in a single enterprise as is required to
maximize profit. Rather they will be looking for ways to achieve the
highest return per unit of cash invested. In this situation, the amount of
product per unit of cash is a relevant evaluation criterion. But even
commercial farmers will often have other objectives than just profit
maximization, and other constraints, which will limit the fulfillment of
the profit maximizing criterion.

Because cash can be converted into many different kinds of inputs, it
is more critical to look at alternative uses for it, especially on small
farms where family necessities compete directly for limited cash resources.
If researchers consider only return to cash investment in the commodity in
which they are interested, they may well find that what appears to be a
"good" technology is not acceptable to farm families who would rather use
the cash for a wedding or to repair the house.
Volume II: I,C
page 30

d. Considerations Related to Risk

Often the measures used in field research are based on averages. It is
common, for example, to consider the difference between mean yields of two
or more treatments from a trial or experiment. Techniques in biological
analysis (see II:V,B,1) including analysis of variance are used to
determine if the mean yields of two or more treatments are really
different. In Figure II:I,C.1, "Examples of Evaluation Criteria at the
Enterprise Level", criteria listed under "Returns To Scarce Resources" are
also averages. If only means or averages are used and risk is not taken
into account, it is easy to assign a single value to, for example, net
income per hectare as a measure of return to land. Average yield of a
treatment is multiplied by a single estimate of price and this is
considered as gross income. Too often this is done even when special care
has been taken to use replicated trial designs to provide estimates of
variance for crop yield. Even when an attempt is made to record all inputs
and other costs of production, only a single value is used for the prices
or costs of different inputs. This provides a single estimate of the cost
of production which is then subtracted from the single estimate of gross
income. The difference between the gross income and the cost of production
provides a single estimate of net income per hectare.

But, with everything that can go wrong in crop and livestock
production, a farmer might easily obtain a net income that is much lower
than the single, average value calculated above. If FSR/E field teams do
not consider all sources of variation and attempt to assess the risk
farmers would face in using a proposed practice, they are conducting an
incomplete evaluation. Even worse, the team may be misleading farmers if
they recommend a proposed practice or technology based on such an
incomplete analysis.

Means or averages are useful beginnings, but do not tell the whole
story. Farmers also want to know what the chances are that their yield or
income may fall below some minimum acceptable level if they adopt an
alternative to their present practice. In other words, how risky is it?

In focusing on evaluation of technological alternatives in this unit,
"risk" can be considered as the probability of returns from a farm
production activity falling below some minimum level acceptable to farmers.
Risk, as defined here, is evaluated by all farmers within the scope of
their individual farm settings. For some farmers, the possibility of
starvation may be the most important risk factor which they-face. FSR/E
field teams must consider aspects of risk for farmers as a group within
recommendation domains, as well as risks associated with individual farms.

When the probability of returns following below the minimum level can
not be calculated, there are alternative methods available to evaluate risk
such as sensitivity analysis (see III:III,A,5 and III:III,A,6) and the
construction of confidence intervals (see III:III,B,2).

Risk, as considered by individual farmers, arises from variability and
change they face which are related to their individual farm setting.

Volume II: I,C
page 31

Specifically, some facets of variability considered by farmers when they
make their estimates of riskiness include the following;

1. changes in yield or product quality caused by such factors as
variations in weather which happen over time even when farming
practices do not change;

2. changes in farming practices over time
a. changing input quality
b. changing rates or times of application
c. changing cultivars

3. changes in the prices of inputs
a. seasonal price fluctuations
b. long term price trends due to inflation or various cycles
c. other factors such as government policy changes

4. changes in prices received for products
a. seasonal price fluctuations
b. long term price trends due to inflation or various cycles
c. other factors such as government policy changes

Changes related to (1) come about because of bioclimatical effects that
differ from year to year. These are beyond the control of farmers. But,
with their years of local experience, farmers have a feel for the extent of
these effects. Changes related to (2) are a result of differences in
management, the human factor. Farmers usually have a good idea of the
expected results of changing their practices before they do so. However,
there remains the possibility that they were mistaken and that the changes
might produce negative outcomes. Changes related to (3) and (4) result
from economic conditions mostly or completely outside the control of
farmers. However, they are aware of previous trends in costs and prices
and use this awareness to estimate risk. FSR/E field teams must consider
all these factors contributing to variation within a single farm setting
when assessing alternatives.

At the same time, the field team must include variation among farmers
in a recommendation domain in its considerations. Different farmers often
use very different practices in growing the same crop. Costs of inputs
vary greatly among farmers depending upon their distance from a source of
supply, transportation available and what balance of farm produced versus
purchased inputs they use. Prices received by different farmers vary
according to factors such as product quality, time of marketing, and
distance from market.

Figure II:I,C.1 lists some evaluation criteria which might be used
under various circumstances to compare the economic benefits of alternative
technologies and farmer practices. This listing is not complete. It is
limited primarily to the enterprise level of analysisTrom a short-term
change perspective. As might be inferred from Figure II:I,C.1, economic
benefits can result from yield increases, reduction in cost, decreased use
of other scarce resources, reduced drudgery of tasks, enhanced product
quality, stability in production and/or improved distribution of outputs.

Volume II: I,C
page 32


Assessment of acceptability by each type of stakeholder
verbally by each
adoption by relevant stakeholders


-Returns/unit of labor at planting, weeding, harvest, ect.
Returns/unit of land
Returns/unit of seed
Returns/unit of cash
Returns/unit of long term capital investment
Returns/animal/day for traction
Returns/unit of land/unit of time

Probability of receiving less than a certain minimum level of
return acceptable to farmers. Measured using any of the
relevant criteria under the "Returns to Scarce Resources"
heading above.

disaggregatedd by age, gender, position, and/or household type)

Labor inputs
Management time inputs
Variable cash inputs
Variable noncash inputs (seed, manure, etc.)
Returns (benefits) received
Indirect benefits (by products, access to resources)
Net Benefits
Gross Margin

(related to market price and/or household utility)

Consumption preferences (taste, size, color, shape, etc.)
Susceptibility to.preharvest pest damage
Storability and susceptibility to postharvest pest damage
Processing quality
Cooking quality
Nutritional quality Usefulness of by products (legume hay)
Length or quality of stalks

Note: some of these criteria may be more suited to evaluation by monitoring
activities over time rather than by analysis of trial data.

Volume II: I,C
page 33

e. Considerations Related to Other Farm-Household Activities

Often secondary effects from introducing alternative technologies occur
in other enterprises on the farm which are not directly involved in the
change. For example, fruit production from an orchard might be increased
by controlling weeds, but those weeds would then be unavailable for
livestock grazing. Increasing the planting density of one crop in an
intercropping situation might decrease the yield of a second crop.

The suitability of changing a farm practice is often seen in a
different light when viewed in respect to the overall production,
consumption and welfare of the household. If the amount of a resource used
in a farm enterprise is to be increased by a proposed alternative, where
will that increase come from? How will that affect the activity whe-rit
is presently used? For example, a recommendation to use additional manure
in cropping to gain better yields might conflict with the need for manure
as fuel for cooking.

If use of a resource in a farm enterprise is decreased by a proposed
change, where and how will that freed resource be used? How will that
increased input affect the activity where it will be used? A
recommendation to increase the planting density of a grain crop might
decrease the amount of land needed to obtain a given yield. If the freed
land remains unused because time is not available to manage a new
enterprise on it, the change may have been for naught. If the freed land
is used to increase plantings of another crop, how will the new plantings
affect the overall costs and benefits to the household? Who among the
household members will have to invest additional management time, labor and
capital, and who will receive the various returns-from the new crop?


With such a variety of potential economic criteria, how can a farming
systems team identify those which are most crucial to the evaluation? One
consideration in weighing and ranking criteria is significance to each
stakeholder (See Section 1). Continuing dialogue with principal
stakeholders is essential. Observing roles and questioning each relevant
type of stakeholder; male farmer, female farmer, head of household,
homemaker, older adult, youth and so on, will provide feedback on the
importance and suitability of specific criteria to each of them. Directed
questioning about proposed changes in farming practices also assists in
pinpointing possible effects on other enterprises and overall household
welfare. The generalized farming systems model introduced in I:II is
helpful in considering interactions among the crop, livestock, household
and off-farm components of the system and in gauging the possible effects
of changes in farm practices.

Economic evaluation criteria do not stand alone in farming systems
analysis. Non-economic criteria from the biological and social realms are
also essential pieces of the technology evaluation puzzle. Economic

Volume II: I,C
page 34

questions and issues cannot be examined apart from biological and social
concerns. Suitability to local climate and soils, compatibility with local
culture and social arrangements, yield sustainability, pest factors and
ecological sustainability are a few of many areas where biological or
social evaluation criteria are essential.

Volume II: I,C
page 35






Volume II: I,C
page 37


OBJECTIVE: After completing this activity you will be able to:

Recognize the importance of gaining a close understanding of farmer
perspectives and circumstances when identifying evaluation criteria for
analysis of on-farm research.


Trainee handout II:I,C Activity One #1 "Conversation with a Manataro Valley
Potato Farmer"

Trainee handout II:I,C Activity One #2 "Farming Systems Potato Research
Team: First year Summary of
Mantaro Valley Activities"


1. The trainer will distribute one of two possible handouts to each group.
If your group receives the handout entitled "Conversation with a
Mantaro Valley Potato Farmer", you are to assume the role of small
potato farmers. Or, if your group receives the handout "Farming
Systems Potato Research Team: First Year Summary Of Mantaro Valley
Activities", you are to assume the roles of members of the local
farming systems team. In either case, read.the handout you receive and
discuss the case among yourselves.

2. In your allowed roles as either researcher or farmer, identify
important evaluation criteria for analyzing on-farm potato trials which
are designed to compare existing small farmer fertilization and insect
control practices with improved recommendations. Make a list of the
possible criteria which your group agrees upon. You may wish to refer
to figure II:I.1 for a partial list of evaluation criteria.

3. Briefly, in the context of your role as farmer or researcher, explain
to the other groups your reasons for selecting each evaluation
criterion. List the criteria so the others can see.

4. The trainer will guide a discussion among the class concerning the
findings of each group.

Volume II: I,C
page 39



We who live on the middle slopes of the Mantaro Valley lead a difficult
life, it is sure. The rich farmers lower down along the river can grow
many more crops than we. The cold, the frost and the drought keep us from
growing maize and other crops from the warm lands below. I am thankful
that we can grow potatoes and barley at least. Feeding a family by farming
is risky and dangerous for us who live on the slopes of the Andes
Mountains. When God wills, our crops are lost to the hail, the frosts and
the drought....

We cannot grow many crops well like those below us, but our sheep and
other animals do better here and so are important to us. Yes, we do sell
potatoes, barley, and sometimes vegetables, wool and mutton in the market
when God has provided a good year. But you see, one never knows when the
crop will fail....

No, we do not plant large fields. Like our neighbors, we have small
parcels of land all around where we raise our crops and graze our animals.
On our better land, we watch our crops closer. On our poorer land, we
plant, but weed less and do not put fertilizer. For us, it is more
important to be very sure that our mouths are full than to keep our pockets
full! Senor, we are poor and humble farmers....

You see, we are farmers first, but we do many other things to live.
Some men leave our town to work in the mines for months during the year.
Others work by the day in fields around here. A few have a small store in
their home. I for one am also a carpenter. Our mothers, wives and
daughters work in the fields too, especially when the men are away. They
too work hard at growing the potatoes. They also dry the potatoes and
prepare them to eat. Many times the women take the crops we sell to the
market and spend the day selling. Our little ones begin to help in the
house and field almost when they first walk. That is the truth, senor....

Si, senor, as did our fathers, we too cultivate potatoes. We plant our
small potato field each year.. Like our neighbors we almost never plant
more than one hectare of potatoes. Some of our harvest we store in our
house and we dry some when we can. During the year we have food to eat and
to sell when it is needed. In my family, we sometimes eat potatoes for all
three meals. Near planting time we are careful to eat less so potatoes are
left. Those which are left when time to plant again are put in the soil
for the new crop....

Senor, the worst problem we have is the cold, the hail, the frost and
the drought. I believe only God can solve these. The worms that bore into
the potato and the insects that eat the leaves are also big problems. The
diseases of the field and of the house sometimes take many potatoes. The
poisons to kill the insects cost us much of our money. Senor, we are poor

Volume II: I,C
page 41

people with few coins in our pockets....

Si senor, we have met the extension agronomo from the city. The
recommendations he makes seem very good. No senor, very few of us have
tried them in our potato fields....

For various reasons, senor. You see, he tells us to plant the new
varieties. They will give more potatoes. Listen, when we plant the new
varieties, the plants grow slow with little fertilizer from the bag. We
see more insects and the potatoes rot. When we take the new potatoes to our
home to eat, our families want the old ones that taste better....

Si, of course, we could put on more chemical fertilizer from the bag
and spray more often against the insects. But potatoes given fertilizer
from the bag taste bad. And you know like I do that growing potatoes is
risky. What would one do if the crop were lost after spending much on new
seed, on chemical fertilizer and insect poison? This I ask you, senor. We
work hard for our wages so our families can eat and be clothed. Potatoes
are not the only crop we raise. One must also have money for supplies for
the other crops and the animals. It cannot be that a man's family will
have empty mouths! ....

He tells us to plant closer, to spray the poisons more often, to weed
better! He does not know. You see the extensionist is not married and has
never grown potatoes. If he had he would know that the things he tells us
to do mean we must stay longer in each field. Well, maybe that is all.
right for me, I work here in this town. But what about those families
where the man must leave right after planting the potatoes so he can work
in the mines? The little ones and the women of the house must then care for
the potatoes. How can they spend even more time in the field like the new
practices need? No senor, he does not know....

With much pleasure, senor. God go with you also....

Volume II: I,C
page 42


In the Intermediate Zone of the Mantaro Valley (3450-3950 meters above
sea level), fewer crops are grown than along the valley floor because of
cold and increasing frosts. Farming is a risky business on the slopes of
the Andes Mountains. Crops are often lost to hail, frosts or drought.

Potatoes, barley and other small grains are the major crops of the
small farms found here. Sheep and other livestock are especially important
at the higher altitudes in the zone. Although some crops and livestock are
raised primarily for sale, potatoes are grown largely for home consumption.
However, some families sell surplus potatoes to middlemen for sale in the
urban markets. Nearly every family cultivates a hectare or less of
potatoes. In most small farm-households at least one member works
off-farm, so some limited cash is available for home use and for purchasing
inputs such as chemical fertilizers, pesticides and labor for raising
potatoes. A few larger growers specialize in commercial production of
potatoes using high levels of inputs. By investing more capital in
production and more time in management they usually achieve much higher
yields than the typical small potato farmer.

The farming systems team began its potato research work here almost a
year ago with the intent of testing and adapting alternative agronomic and
post-harvest potato practices on small farms. Much has been accomplished
during those hectic first two years. Surveys of the area have shed light on
the ecology and agriculture of the valley and provided a partial
understanding of constraints to potato production, processing and storage.
Farmers' practices have been studied and records of the costs and returns
from potato production have been compiled. Observations and dialogues with
the men, women and children involved in potato growing have provided useful
insights about their different production goals, roles and situations.

Small farmers who were interviewed by the team were found to face
several constraints to raising their yields and profit from potatoes.
Their use of traditional low yielding varieties and practices keeps their
yields less than half those of commercial farmers. Most small farmers
store their seed from season to season. This results in increased virus
infection of their seed stock. Many also eat the larger tubers, thus
planting the smaller ones. Less vigorous plants are likely to result from
planting smaller tubers.

Although many use chemical fertilizers and insecticides, few follow
recommendations for dosages and timing of applications. Some supplement
their meager applications of chemical fertilizer with manure but overall
the soils in small farmer potato parcels seem to be generally infertile.
In most cases, the team found farmers' insect control practices to be
inadequate. It appears that a number of small farmers have not adopted

Volume II: I,C
page 43

recommended practices because of a lack of information or due to their
traditional resistance to change.

These initial observations led the farming systems team to the
following priority list of researchable problems as a basis for planning
and designing on-farm trials for the coming season;
poor quality seed
inadequate fertilization
inadequate insect control
use of poor yielding unimproved varieties

Although considered as major constraints, climate-related factors are
not considered researchable at the present because solutions are not
readily available. Several tentative solutions are being studied by the
team and various on-farm trials are currently being designed for the coming
potato growing season.

Volume II: I,C
page 44


OBJECTIVE: After completing this activity you will be able to:

Select economic evaluation criteria for analysis of on-farm research which
are relevant to the different stakeholders and situations in farm


1. Read trainee handout II:I,C Activity One #1 "Conversation With A
Mantaro Valley Potato

2. In this role playing activity, the trainer will ask your group to
assume one of the following roles and situations in potato production
in the Mantaro Valley;

a. The male head of a household who leaves, after potato planting, to
work in the mines for several months to provide income for household
needs and for purchasing potato fertilizer and insecticides.

b. The wife/mother of a subsistence household who manages and cares for
the potato crop after planting because her husband is away working
in the mines.

c. The wife/mother of a farm-household whose doesn't work in potato
production but processes and cooks part of the harvest for home use.

d. A young, landless farmer who grows potatoes commercially on rented
land and hires a team of oxen for preparing the land.

e. A farmer raising 25 hectares of hybrid potatoes for sale in the

-f. Members of a household that derives income from raising and storing
seed potatoes for sale to other farmers at planting time.

3. From the perspective of your allowed role, identify important
evaluation criteria for analyzing on-farm potato trials. Make a list
of the possible criteria which your group agrees upon.

4. Take a few minutes to explain to the other groups your reasons for
selecting each evaluation criterion given your assumed role or
situation. List the criteria so the others can see.

5. The trainer will guide a discussion among the class concerning the
findings of each group.

Volume II: I,C
page 45

6. Farming systems teams are notoriously overworked and have limited
resources with which to accomplish their tasks. Obviously not every
analysis nor every possible evaluation criterion can be fully explored.
Which of the proposed criteria do you think are important enough to use
in analyzing an on-farm potato trial which compares existing farmer
fertilization and insect control with recommended practices?

Volume II: I,C
page 46





Researchers, technical assistants, extensionists, and administrators


After completing this sub-unit the participants will be able to:

1. Describe how FSR/E contributes to station research, and how station
research contributes to FSR/E, in terms of the FSR/E sequence of
diagnosis, design, testing, and extension.

2. Explain under what conditions station research may be necessary prior
to on-farm experimentation.

3. Explain the reasons why it is necessary to consult with farmers in
planning station, commodity, and disciplinary research.


1. FSR/E, through on-farm research, can greatly assist research stations
in setting priorities for on-station research and experimentation,
based on farmer needs and priority problems.

2. The results of on-station research can provide FSR/E with potential
solutions (technologies) which may be able to solve farmer problems.

3. Assessment of what is known about potential solutions to farmers
problems identified in diagnosis is the key linkage point between
on-farm and on-station research.


FSR/E is an approach which research and extension institutions can use
to increase their effectiveness in planning and implementing programs
appropriate to farmers. Some people have feared that FSR/E is presented as
a substitute for station based research. However, this is most certainly
not the case. FSR/E is a complement to, and not a substitute for, station
research. There is no necessary separation of station research and FSR/E.
In fact, good FSR/E depends on effective and efficient station research.
The same people maybe and often are involved in both. Many field teams
employing FSR/E methods are based at research stations and carry out
experiments on-station while simultaneously conducting on-farm trials.
Rather than being viewed as a separate effort, on-farm research should be

Volume II: II,A
page 47





Researchers, technical assistants, extensionists, and administrators


After completing this sub-unit the participants will be able to:

1. Describe how FSR/E contributes to station research, and how station
research contributes to FSR/E, in terms of the FSR/E sequence of
diagnosis, design, testing, and extension.

2. Explain under what conditions station research may be necessary prior
to on-farm experimentation.

3. Explain the reasons why it is necessary to consult with farmers in
planning station, commodity, and disciplinary research.


1. FSR/E, through on-farm research, can greatly assist research stations
in setting priorities for on-station research and experimentation,
based on farmer needs and priority problems.

2. The results of on-station research can provide FSR/E with potential
solutions (technologies) which may be able to solve farmer problems.

3. Assessment of what is known about potential solutions to farmers
problems identified in diagnosis is the key linkage point between
on-farm and on-station research.


FSR/E is an approach which research and extension institutions can use
to increase their effectiveness in planning and implementing programs
appropriate to farmers. Some people have feared that FSR/E is presented as
a substitute for station based research. However, this is most certainly
not the case. FSR/E is a complement to, and not a substitute for, station
research. There is no necessary separation of station research and FSR/E.
In fact, good FSR/E depends on effective and efficient station research.
The same people maybe and often are involved in both. Many field teams
employing FSR/E methods are based at research stations and carry out
experiments on-station while simultaneously conducting on-farm trials.
Rather than being viewed as a separate effort, on-farm research should be

Volume II: II,A
page 47

part of the entire research and extension continuum. The diagram, shown in
Figure II:II,A.1, and example demonstrate how these linkages operate.

Figure II:II,A.1 shows the interactions between station-based technical
research and on-farm research (Collinson 1982). Franzel provides an
example of how the diagram works under actual conditions (Franzel 1984).

The diagram depicts the interactions between station research and
on-farm research as taking place in a series of stages. Stage One is a
diagnostic survey of farmers in an area where maize-bean intercropping
prevails. In Stage Two, low soil fertility is identified as a constraint.
During this stage, many potential solutions are proposed. Some, such as
compound fertilizer, are rejected as not appropriate (no economic benefit
to be gained, farmers lack cash purchase fertilizer, fertilizer not
available even if farmers have case, etc.). Other solutions, such as the
use of animal manure, are judged as potentially acceptable to farmers and
feasible. Since work has been done on-station and much is known about
manure composition and use, it is decided that experiments can be conducted
on farmers' fields to test both the crop response to the manure and the
different methods of application under farmer circumstances.

At this point, the team moves in two directions. First of all,
concerning the potential solution of utilizing animal manure, the design of
appropriate on-farm experiments begins (Stage Three). These will be
conducted with farmer cooperators within the target group. If experiments
are successful, recommendations are formulated and extended to farmers.

At the same time, the team decides that not enough is known about the
other potential technology solutions to be able to proceed directly to
on-farm trials. Rather, the team decides to send their proposals to the
research station for further research and testing (Stage Four).

Taking one of these research proposals (the introduction of a green
manure crop as a relay crop into standing maize following the bean harvest)
as an example, the decision to continue on-station research before moving
to on-farm trials does not eliminate the need to consult with farmers.
Researchers determine that the crop would have to be plowed under before
planting season. Since this would require additional labor and use of
animal traction, researchers need to consult with farmers about the
potential suitability of such shifts before investing in agronomic and
biological study (Stage Five). Researchers discuss the proposal with
farmers and find that farmers are very enthusiastic, despite the need for
shifts in labor and animal use required by the technology. Researchers
decide to test this innovation on-station in order to establish which
possible forage crops appear most appropriate (i.e., add the most nitrogen,
interfere least with the standing maize, etc.) The green manure crops are
tested in Stage Six and those which respond best and seem most appropriate
become part of the body of materials potentially suitable for farmers in
the area.

These results then become part of the potential solutions which the
team can consider in designing the next series of on-farm experiments.
When once again at Stage Two, researchers decide to take these green manure
crops from Stage Six and test them in Stage Three.

Volume II: II,A
page 48

This example demonstrates that the linkages between on-station and
on-farm research are very important. In Stage Four, the FSR/E team
provides station researchers with ideas for experimentation based on their
identification of farmer problems. Where research funds are scarce and
research stations must focus on priority problems of farmers in a
particular region, the referral of problems from a FSR/E team can greatly
help prioritize problems for the allocation of these scarce funds. In
Stage Six, the station supplies the FSR/E team with potential solutions to
farmers' problems. The example demonstrates the interdependence of on-farm
and on-station research. Neither can function effectively without input
from the other.

It is important to remember that extension staff play important roles
in all of these stages (see also VI, and IV,B). Simply conducting on-farm
experiments does not necessarily mean one is doing FSR/E. An important
part of FSR/E is that practitioners (researchers and extensionists) are
interacting with farmers and are using their holistic understanding of
farmers circumstances to plan and modify technology experimentation.

Volume II: II,A
page 49

Figure II,A.1

Interactions Between Station-Based Technical Research and
On-Farm Adaptive Research.

Target Group Farmers
of a Recommendation
Domain in a Region

Survey diagnosis of
farmer priorities,
(1) resource and environment
problems and development


Experiments on apparently
relevant materials (
and techniques under
farmers' conditions


Identification & Evaluation .
of materials and techniques
.offering potential for problem
solution and the exploitation
of opportunities.


Unsolved technical
problems and possible
(4) new practices and
materials relevant
to farmers' development


Body of Knowledge of
materials and techniques (6)
suitable for the climate
and soils of the Region

Commodity and Disciplinary
research, solving priority
technical problems and
investigating possible
new materials and practices

Figure 1. Interactions between Station-based Technical Research and
On-Farm Adaptive research.

Source: Collinson, M. Farming Systems Research in
Eastern Africa: The Experience of CIMMYT and Some
Agricultural Research Services, 1976, 1981, MSU
International Development Papers, No.3, 1982, page 5.

Volume II: II,A
page 50



1. Differences Within Farms
2. Differences Across Farms
3. Summary


II:I,A What Kind of Testing to Do


Agricultural research assistant
Extension technology verification technician


After completing this sub-unit, participants will be able to:

1. Identify sources of differences within and among fields.

2. Distinguish between site-specific and regional trials.

3. Identify different possible combinations of block size and number
within and across farms.


1. Soil, topography, land fragmentation, farm size, and household ability
to take risk all affect block size.

2. Block size may vary from one farm to another.

3. Unequal block sizes across farms require either more complex designs or
reduction of the number of treatments.


contiguous blocks (plots)
experimental design

Volume II: II,B
page 51



Fields differ in many ways. Here are some examples:

Way a Field May Differ Example

1. Soil texture
2. Soil depth
3. Slope
4. Previous crops

5. Previous management

Way Animals May Differ

One end is sandier than the other end.
Even a light plowing turns up more clay at one end.
Low spot in the center.
Half left fallow but the other half planted in
Ran out of fertilizer and only side dressed the
first five rows.


1. Age Flock consists of many old animals and few young.
2. Sex One male, many females.
3. Lactation status Milking animals calved/lambed at different times;
some have yet to give birth and are dry.
4. Breeds Goats were purchased at the market.
5. Live weights Lambs born at different times.
6. Reproduction Some cows are infertile.
7. Management Bullocks are fed feed supplements.

Farmers know their fields well. They would not expect plants in the
sandy part of a field to be the same as plants in a more silty part. The
team should identify where to divide fields so that each part has no
obvious differences. Parts with no obvious differences are uniform and

Sometimes, two parts of a field may be similar even though they are
separated. For example, there may be a path in the middle of a field, but
both sides are level and sandy.

A block is a uniform area with different treatments. Each block is
subdivided into plots. A plot (experimental unit) is an area with only one
treatment. A treatment is something the team wants to test. For example,
a treatment may be a new variety, a high rate of fertilizer, or
intercropping one row of groundnuts between each row of corn. A treatment
might also be the farmer variety, a low rate of fertilizer, or corn planted
at random among groundnuts, or a new type of farm implement.

Each part of the field that has no more obvious differences can be a
block. Two parts of the same field separated by a path can also be one
block. This would be a block with non-contiguous (separated) plots. In
both contiguous and non-contiguous blocks, plots within the block will be
about the same before the team applies the treatments. Obviously the plots
in a block will be very different after the team applies the treatments.
But they will not be very different in any other way. This is an important
rule: blocks are uniform except for treatment differences.

Volume II: II,B
page 52

Within each block, plots may not be exactly the same size or shape.
One question is: how much difference in size or shape is acceptable? Some
teams have found 20% to be a useful rule of thumb for maximum variability
within blocks.

A block cannot be larger than a uniform part of a field. However, it
can be smaller. A large uniform part of a field may be divided into two or
three blocks. This means the same treatment can occur more than once.
When a treatment occurs more than once, it is replicated. A replication is
a complete set of treatments. More treatments means more plots in a
replication. More plots in a replication means larger blocks. Larger
blocks may mean a uniform part of the field can only be divided into two
blocks, rather than into three blocks. That will reduce replications from
two to three. Or, even larger blocks may mean the uniform part of the
field cannot be divided at all. Then, the farm will have only one
replication. Other replications of treatments will have to occur on other
farms. (I) and (III,B) discuss why replication is important.

Within farm replication of treatments are not generally possible when
conducting animal trials. Not enough animals are raised on one farm that
are physiologically similar. When the experimental unit is a single
animal, it may be difficult to find another animal that is the same age,
sex, or weight. If a group of animals is the experimental unit,
limitations exist in the farm's ability to divide a flock into two and
manage them as replicate treatments. Therefore, a common practice is to
use the herd/flock as the experimental unit, test one treatment per farm,
'and replicate treatments across different farms.

If more replication within each farm is important, the team may need to
reduce the number of treatments. When is this necessary? That depends on
the objectives of the trial and the treatments. (II:I,C,1) and (II:III,A)
discuss how to make decisions about determining objectives.

If a team needs to reduce the number of treatments, how does it do so?
(II:II,C,2) and (II:II,C,3) give some ways to reduce the number of


Farms can differ in the same way as fields. Sometimes part of a field
on one farm will be similar to part of a field on another farm. For
example, both fields may have one end where even light plowing turns up
clay. Those two ends can be paired in incomplete block designs.

Differences in animal management and feed resources can be large
between farms. Some farms may have resources to purchase supplements,
while others graze their animals solely on noncultivated land. However,
the productivity of herds/flocks on different farms can be more similar
than animals on the same farm. The composition of the herds/flocks could
be quite uniform between farms, in terms of age, sex, ect. The degree of
interfarm variability differs between factors. For example, there is more
variability between nursing lambs and kids compared to weaned animals.
Suckled animals are more sensitive to diet differences between and within
farms. With increasing levels of variation, a greater number of replicates

Volume II: II,B
page 53

will be needed to detect significant differences.

Some farms may have many small parcels of land. Other farms may have
only one or two larger pieces of land. The small parcels may all be
different, but the larger pieces of land may be more uniform within each
larger piece. If so, larger blocks will be possible on the farms with the
larger pieces of land.

The amount of land which the farm households will allocate for
experimentation differs from farm to farm. Farms with more land often are
willing to experiment on a larger field, or on a greater number of small
parcels. Farmers with less land will probably not be willing to experiment
on as much of their land. Some teams have found that a useful rule of
thumb for deciding the maximum area to use for experimentation on each farm
is to use no more than 10 % of the farm's total area.

Some farm households can take even less risk. Differences among farm
households that can affect the ability to take risk include age and gender
of household members, and off-farm sources of income. For example, farm
households with only a woman parent and no off-farm income will depend
entirely on their land, crops, and animals. They may be able to put only
one small parcel in an experiment. Farm households with two or three
generations of adults, both men and women, may have many off-farm sources
of income. They may be willing to take more risk and put a larger
percentage of land in an experiment.

Where should a team place trials when farms differ? There are several
decisions to make. The first is:

1. Should one kind of trial gb on one kind of farm?

Whether the answer to this question is yes or no depends on the
objectives of the trial. (II:III,A) will help to determine the answer to
this question. Whether the answer is yes or no affects design and analysis

la: YES: The trial is a site-specific trial (see Hildebrand and Poey,
1985: 45-55W. More than one replication is needed on the farm. This
usually means a fairly large amount of land is needed on one farm. Designs
for site-specific trials are similar to station trials. In fact, a station
trial is simply a specialized site-specific trial, where the station is the
team's "farm."

In station site-specific trials, blocks are usually touching, or
contiguous. In on-farm site-specific trials, however, blocks might not
touch. This means they are non-contiguous. Figure II:II,B.1 compares
contiguous and non-contiguous blocks. Both have four blocks, I, II, III,
and IV. Each block has four treatments, a, b, c, and d. In the
non-contiguous example a path separates blocks I and II from III and IV.

Completely random or randomized complete block designs are most common
(II:III,C,1). Analysis is by Analysis of Variance (ANOVA). The type of
ANOVA used depends on design and treatments. (II:II,C and II:II,E) present
treatment choices. (III:III,A,1 and III:III,A,2) explain ANOVA procedures

Volume II: II,B
page 54

appropriate for different designs and treatment choices.

Figure II:II,B.1 Contiguous and Non-Contiguous Blocks in Site-Specific



ad a
I d I a I



I I | |
| |

I c bI d a I

II Ic l a -

III a c i d I b\

IV -bY- a c T
I I I I ___

lb: NO: The trial is a regional trial (see Hildebrand and Poey, 1985:7).
The questions that follow can help the team decide where to place regional

Regional trials mean the team places treatments across many farms. Where
should the team place treatments when farms differ? There are several more
questions to answer:

2. How much do farms differ in the amount of land available for trials?

2a: Small differences, this means blocks can be the same size on each

2b. Large differences, this means block sizes may be different among
farms. Some farms may have enough land for several blocks, but other farms
may have enough land for only one block.

3. How many treatments does the team want to test?

3a. Small number (equal to smallest block size among the farms), this
means every block can have a complete set of treatments.

3b. Large number (greater than the smallest block size among the farms),
this means some blocks cannot have all treatments. The team has two

(1) Use more complex designs.
(2) Reduce the number of treatments to equal the smallest block size and
use simple designs.

What combinations of block size and number are possible in regional
trials? What designs and analysis procedures are appropriate for different
combinations? The next series of questions looks at several different

Volume II: II,B
page 55

I I c

II V c
iv |BI





possible combinations of block size and number across farms:

4. How many blocks can go on each farm?

4a. Only one block on every farm.

4b. More than one block on every farm with the same number of blocks
on all farms.

4c. Unequal block numbers on different farms.

5. Are blocks the same size on each farm?

5a. Yes, farms have equal block sizes.

5b. No, farms have unequal block sizes.

For each answer to question 4, the team needs to ask question 5. There
are six possible combinations. Which is the best combination? The best
combination is different for each situation. The answer depends on the
objectives of the trial and the treatments. (II:II,C,1) and (II:III,A)
help to determine how to make decisions about objectives. The number of
treatments also affects which combination the team will choose. (II:II,F)
can help identify trade-offs among treatments and replications.

Which combination the team chooses in turn affects design and analysis
procedures. A team needs to consider the design and analysis procedures in
deciding which combination is best in each situation. Figure (II:II,B.2)
shows an example for each combination:

4a. Only one block on every farm.

5a. Equal block sizes, each block is a replication. This means each
farm is also a replication. For example, farm A is also replication I.
Each farm has both treatments, a and b. Randomized complete block design
is appropriate (II:III,C,l,b). In this example with only two treatments, a
and b might be farmer variety versus a new variety. With more than two
treatments on each farm, analysis can be by modified stability analysis.

5b. Unequal block sizes, each farm may not be a replication. For
example, farms C and D have larger blocks and could have a third treatment
c in addition to treatments a and b. Treatment c, for example, might be a
second new variety. Farms A and B would not have the second new variety.
This combination requires use of an incomplete block design (II:III,C,b).
The alternate is to eliminate the extra treatments (in this example,
treatment c) so that all farms have equal block sizes (combination 4a +

4b. More than 1 block on every farm, with the same number of blocks on all

Volume II: II,B
page 56

5a. Equal block sizes, each farm has the same number of blocks. Each
block has the same number of treatments, so each block can be a
replication. For example, farm A has two replications, I and II. Farms B,
C, and D similarly each have two replications. Each replication has all
four treatments, a,b,c, and d. Randomized complete block design
(II:III,C,1) is appropriate. Analysis of variance with combined analysis
(III:III,A,1 and III,A,2) can be used to test whether treatments perform
the same on all farms or not. Modified stability analysis (III:III,B,2)
may also be used.

5b. Unequal block sizes, each farm has the same number of blocks, but
some farms have smaller block sizes. For example, block III on farm B can
take all four treatments, but replication IV can only take three
treatments. Farm C is like farm B. On farm D, both replications VII and
VIII can take only three treatments. This situation may arise when some
farms have smaller amounts of land available for trials. This combination
requires use of an incomplete block design (Unit II:III,C,l,b). The
alternative is to reduce the number of treatments to equal the smallest
block size. In this example, this would reduce treatment number from four
to three. This change would then allow a randomized complete block design
to be used, as in combination 4b + 4a.

Figure II:II,B.2 Examples of Combinations of Block Number and
Block Size in Regional Trials Across Farms

asanuams ax io. aWxsza (A, ........-fa)
(Irn,....... -bl)odk)
(a,b ........ in ploes)

8 + 58 Qay 1 n kqal A B
ay I I b b-a --" 1, ilbI ml aI
4* + 5b OQal m T n II .l C D
aw7y fam. Itbj I r. f IV a
4b + 5a Me thm 1 Eqtal rA C D

M ffft7 fans
4b + 5b m tb 1 lm qul A B C D

m-r ,,--I d,-
bodk.in I.= |Hc nin d VcFd VIE T& |cb
equal ramhe

4c + 5s Uaqul %I**odc Eql A B C D

fams 1 11 -- I1--II--

4c + 5b da*pqul hlod1 Uaeal A B C

Volume II: II,B
page 57

4c. Unequal block numbers on different farms

5a. Equal block sizes, each block has the same number of treatments,
so each block can be a replication, however, some farms have more blocks
than others. This may happen if some farms have less land available for
trials. For example, farms a and b each have two blocks, but farms C and D
have only one block each. Randomized complete block design (II:III,C,l,a)
can be used. Analysis of variance with combined analysis (III:III,A,2) can
be used but analysis is more complex.

5b. Unequal block sizes, both number of blocks and size of blocks
differ among farms. For example, farm A has two blocks, but only block one
can take all four treatments. Block II can take only three treatments.
Farm B also has two blocks, but each can take only three treatments also.
Farm C has only one block, which can take all four treatments. Farm D has
only one block, and it can take only three treatments. This situation may
occur with greater diversity among farms in land available for trials.
However, this combination requires an incomplete block design
(II:III,C,l,b). The alternative is to reduce the number of treatments to
the smallest block size (in this example, 3). This change would allow a
randomized complete block design to be used, as in combination 4c + 5a (but
with 3 plots per replication rather than 4).

Note: In the examples shown in figure II:II,B.1, blocks are not touching
each other. These are non-contiguous-blocks. Non-contiguous blocks are
the rule in regional on-farm experimentation. This is different from
station experimentation.


Farms can differ in many ways. Uniform parts of fields are called
blocks. Blocks are divided into plots. Each plot has a treatment.

Farms can differ in the amount of land available for trials. They can
also differ in size and number of blocks.

Teams can place trials on only one farm or across farms. Trials on any
one farm are site-specific trials. They are similar to station trials.
Trials placed across farms are regional trials. Blocks are located on many
farms. Block size and number may be equal or unequal among the farms.
Equal block size.and number allow simple designs and analysis procedures.
Unequal block size and number require either more complex designs or
reduction of treatment number.


II:IIC,l Defining Treatment Objectives
II:II,C,2 What to Consider in Selecting Subsets of Treatments
II:II,C,3 Statistical Techniques for Selecting Subsets of Treatments
II:II,E Examples of Treatments for Different Types of Problems
II:II,F Looking Ahead: What Are Some Trade-offs Between Treatments
and Replications
Volume II: II,B
page 58


How Objectives Change in the Research-Extension Process
What Designs Can Do

Volume II: II,B
page 59



1. Focusing on Priority Problems
2. Developing a Treatment Objectives Statement
3. Using Previous Research to Provide Clues
4. Checking with Farm Households


II:I What Kind of Testing to Do


Agricultural research assistant
Extension technology verification technician


After completing this section, participants will be able to:

1. List steps useful in developing a treatment objectives statement.

2. Identify differences in treatment objectives common in exploratory,
refinement, and validation testing.

3. Write a treatment objectives statement.


1. A treatment objectives statement provides a set of criteria for
selecting individual treatments.

2. Biological criteria are relatively more important in treatment
objectives statements for exploratory and refinement testing than for
validation testing.

3. Social and economic criteria including food consumption are relatively
more important in treatment objectives statements for refinement and
validation testing than for exploratory testing.


ex ante analysis
exploratory trials (testing)
farming system
refinement trials (testing)

Volume II: II,C,1
page 61



1. Focusing on Priority Problems
2. Developing a Treatment Objectives Statement
3. Using Previous Research to Provide Clues
4. Checking with Farm Households


II:I What Kind of Testing to Do


Agricultural research assistant
Extension technology verification technician


After completing this section, participants will be able to:

1. List steps useful in developing a treatment objectives statement.

2. Identify differences in treatment objectives common in exploratory,
refinement, and validation testing.

3. Write a treatment objectives statement.


1. A treatment objectives statement provides a set of criteria for
selecting individual treatments.

2. Biological criteria are relatively more important in treatment
objectives statements for exploratory and refinement testing than for
validation testing.

3. Social and economic criteria including food consumption are relatively
more important in treatment objectives statements for refinement and
validation testing than for exploratory testing.


ex ante analysis
exploratory trials (testing)
farming system
refinement trials (testing)

Volume II: II,C,1
page 61

response Curve
validation trials (testing)



In researcher-planned experimentation, the team focuses on a priority
problem of farm households. For example, the priority problem may be land
preparation techniques to better conserve water from early rains. The team
wants to test different ways of land preparation. How does a team decide
what to test for the priority problem? How many different ways of land
preparation should it test? What other associated practices should it test
at the same time? For example, should it also test fertilizer placement
methods? Perhaps fertilizer placement methods should change with different
land preparation methods. There are many decisions a team needs to make.

Each different type of land preparation could be a treatment. For
example, flat cultivation without ridges would be one treatment. Making
ridges with hoes at one month would be a second treatment. Making ridges
with donkeys at one month would be a third. Changing the time of making
ridges would result in still more treatments. Other treatments could
include different fertilizer placement methods. Many treatments are

Some treatments are related. For example, flat cultivation, hoe
ridging, and donkey ridging are all land preparation methods. Farm
households would do only one of these. Fertilizer placement in hills or in
bands are also two related treatments. Again, farm households would do
either one or the other.

A set of related treatments is called a factor. The land preparation
methods are one factor, and the fertilizer placement methods are another.
Many combinations of the different treatments in two factors (sets of
treatments) are possible. For example, flat cultivation can be done with
fertilizer placed either in hills or in bands.

Another example of two factors would be kinds of fertilizer elements.
Nitrogen would be one factor, and phosphorous another. Many different
nitrogen rates would be possible at each phosphorus rate.

Combining two factors is often necessary if one factor affects the
other. For example, increasing nitrogen may have little effect at a low
phosphorus rate, but have a large effect at a high phosphorus rate. This
is called an interaction between two factors.

Usually, all the operations are done the same way over large areas.
For example, farmers would usually not use four different land preparation
methods on the same field. In experimentation, however, fields are divided
into different parts. These parts are called blocks and plots. Each plot
gets one treatment or treatments. This way, the team and farm household
members can compare the different treatments. (II:II,B) discusses ways to

Volume II: II,C,1
page 62

choose fields and divide them for treatments. (II:III,C,1 and 2) describe
different ways to place treatments in fields within and across farms, in
order to obtain analyzable data.

The first step in deciding what to test for a priority problem is to
develop a treatment objectives statement. This statement outlines what the
team and the farm households want to learn from the trial. The statement
provides a set of criteria for choosing among many treatment possibilities.


How does a team develop a treatment objectives statement? Four steps
are useful:

1. Reviewing what the team has learned in diagnosis about the priority
2. Reviewing what the team knows from previous station and on-farm
research on the priority problem.
3. Writing a preliminary statement.
4. Checking the preliminary statement with farm household members.

The priority problem is only one part of everything farm households do.
Everything that farm households do is called the farming system. For
example, land preparation is only part of growing sorghum. Sorghum is only
one crop that farm household members grow. Besides all the crops, there
may be donkeys, oxen, or perhaps other animals. There is also milling,
working, selling of the sorghum, and so on.

In reviewing what the team has learned in diagnosis about the priority
problem, the team needs to consider both the problem itself, and all the
other activities related to the problem. Some questions to consider are:

1. What are the current management practices (for example, flat
cultivation or donkey plowing)?
2. What are current levels of input use (for example, renting of donkeys,
fertilizer use, etc.)?
3. Where in the farming system are the primary crop-livestock
4. What other activities compete for labor and inputs (for example, yams
planted at the same time)?
5. How is the product used (for example, what mix of home consumption,
informal barter, sharing, or sale)?
6. Who receives the benefits of the product (for example, is money from
sales retained by one individual or is it redistributed throughout the
7. How are the farm households organized (for example, ages and gender of
different members, etc.)?
8. How available are inputs and markets (for example, credit, distance,
9. Who in the household has access and control over inputs?

Volume II: II,C,1
page 63


How much does the team already know about what can happen with
different possible treatments? Here is where reviewing of previous station
and on-farm research can provide clues. These clues can save time. Here
are some useful questions:

1. What is known about the biological principles underlying the problem
(for example, soil characteristics)?
2. What is known about biological effects of different kinds of inputs
(for example, growth differences with different fertilizer placement
methods, or growth responses to different crop resources?
3. What is known about biological effects of different levels of inputs
(for example, stalk versus grain yield over a range of nitrogen rates,
lodging over a range of nitrogen rates, etc.)?
4. What is known about biological effects of different factors varied at
the same time (for example, nitrogen and phosphorus, weeding and plant
spacing, different rotations or annual age and feed supplementation)?
5. What is known about economic and social effects of different kinds of
in ts (for example, use of traction freeing household labor for more
cash crop vegetable production)?
6. What is known about economic and social effects of different levels of
inputs (for example, yield increases paying for one but not two
wee wings ?
7. What is known about the economic and social effects of different
factors varied at the same time (for example, one plant spacing
reducing labor for weeding, and making it possible for the household to
plant a larger area)?
8. What is known about farm household acceptability of different

When less is known about biological principles or effects, treatment
objectives may be to clarify these first. For example, four fertilizer
rates may be needed to make a good response curve. With a good response
curve, it is possible to find the highest rate. In Figure II:II,C,l.l(a),
with only three points, the curve is still rising. Perhaps a higher rate
will still increase yield. Perhaps it will not. The experiment does not
give a clear answer. In Figure II:II,C,l.l(b) with four points, the curve
has flattened. Now it is clear that the third level is high enough. There
is no gain at higher rates. The higher rate is nevertheless useful, to
probe the response.

Clarifying biological principles and effects is most common in
exploratory testing. The objective is to explore the biological effects.
Sometimes the objective is just to explore which factors have any effect or
response at all. At other times, the objective is to explore principles or
responses for one or two most promising factors.

In refinement testing, clarifying biological effects is often still
important. This is especially true if exploratory testing has eliminated
some factors, but one or two others had a biological effect. The objective
may be to find the cut off point of the biological effect. For example,
the exploratory testing may simply have shown that 100 kg increases yield
over no fertilizer. Another trial is necessary to construct a curve like

Volume II: II,C,1
page 64

Figure II:II,C,l.l(b).

Clarifying economic, social, and consumption effects becomes more
important in refinement testing. For example, in Figure II:II,C,l.l(b) the
100 kg gives more yield than the 50 kg, but the increase is not as great as
from 0 to 50 kg. Yet each increase of 50 kg will cost the same. When does
it stop paying to add 50 kg, at 50 or 100 kg? Or, perhaps the additional
labor needed to apply even 50 kg is better spent on yam production. Who in
the household provides the cash for the fertilizer, who provides the
additional labor, and who benefits from the increased yield? What
qualities of proposed varieties are most acceptable to households?

The team can ask many questions to help clarify possible economic,
social, and consumption effects before trying treatments. Sometimes the
answers to these questions will eliminate a treatment before testing. This
type of analysis is called ex ante analysis. Ex ante means "before". That
is, before actually trying Ue treatment. Here are some useful questions
for ex ante economic, social, and consumption analysis:

1. Could farmers adopt and use a recommendation or technology based on
these treatments?
2. Could farm households afford the money to buy the necessary inputs?
3. Are the required inputs, ifany, locally available? If not, are they
likely to become available? Does the farmer have access to these
4. Does the farm family have the labor resources necessary to adopt a
technology recommendation arising from this experiment? Is hired labor
5. Is there anything in the treatments that could pose cultural problems
to farmers, or to some farmers among the target group?
6. Are the expected benefits from the adoption of the recommendations
expected to come out of the experiment likely to be attractive to farm
households? Are yield increases, labor savings, etc., sufficient to
attract farm households to adopt the expected recommendation?
7. If the treatment is related to subsistence crops, will the proposed
treatment be rejected by households based on unacceptable consumption
characteristics in spite of favorable production characteristics?
(Consumption characteristics may include processing, storage, working
qualities, flavor, etc.)

Volume II: II,C,1
page 65

Figure II,C,1.1 Determining Responses


0 50 100 150

fertilizer (kg/ha)

la: Incomplete

0 50 100 150

fertilizer (kg/ha)

1b: Complete
with probe
at 150 kg/ha


0 50 100 150

fertilizer (kg/ha)

1c: Exploratory

Volume II: II,C,1
page 66


All of the above examples and questions can help a team write a good
treatment objectives statement. There is still one more step before
testing. Even the best ex ante analysis may not reveal possible problems
with treatments. A treatment objectives statement is only preliminary
until the team checks it with farm household members. Here are some points
to check:

1. Can farm household members understand the treatment objectives

la: If yes, then check point 2.

lb: If no, the team needs to reword the statement. Rewording in turn
helps the team explain experimentation in farmers language. This promotes
better communication.

2. Do farm household members agree with the treatment objectives?

2a: If yes, then check point 3.

2b: If no, the team needs to determine why not. This can help the
team understand farm households objectives and constraints better. The
team may need to change treatment objectives.

3. How much risk from the treatments can farm households accept?
There are two kinds of risks to consider.

3a: Ordinary risk, associated with normal farmer practices (due to
weather, diseases, pests, uncertain input supplies, unstable market prices,

3b: Experimental risk, associated with new practices (due to uncertain

Farm households understand ordinary risk. They are willing to (or
forced to) accept it as a normal part of farming. This kind of risk should
not affect treatment objectives and treatment choices. Farmers will
understand losses due to normal risk and would not expect compensation
(although they would probably accept it if offered). In order to avoid
paternalism in the research process, it is better not to plan for
compensation for these cases. Farm households sometimes also take some
experimental risk. On-farm experimentation allows them to take more of
this kind of risk, because the research and extension organizations share
the experimental risk with the farm households.

A large amount of risk is borne by the farmer when animals are
committed to an experiment. Because animal numbers are few, using one
animal for the trial is analogus to committing half of the farm to an
agronomic trial. Likewise, the value of an animal or herd/flock is
relatively much higher than a small plot. The amount of risk to the farm
is related to the value of the animal, the type of experiment, and the
amount of management change needed to participate in the trial. Many times

Volume II: II,C,1
page 67

the risk is large enough to greatly reduce the number of possible
participants. Therefore, compensation is an important consideration, and
is used often when conducting animal trials to reduce experimental risk.

How much experimental risk should farm households take, and how much
should the team take? The answer will depend on each situation. As a
general principle, though, farm households should take enough experimental
risk to feel that the trial is theirs. Farm households should not have to
take so much experimental risk that their well-being is jeapordized. The
team should be prepared to bear input costs that would place undue strain
on farm households. The team should also be prepared to provide
compensation or additional support for the costs of treatments that are
obviously not adapted. If the team cannot afford the compensation or
additional support for such treatments, then it needs to reconsider the
treatment objectives.

4. How sustainable are the proposed treatments to farm households?
Should the treatment be successful what kinds of changes would farmers need
to make to maintain the treatments. The team needs to evaluate the long
term implications and problems and opportunity costs of farm households.

The preliminary treatment objectives statement is the first set of
criteria for choosing treatments. Checking the preliminary treatment
objectives statement with farm households helps finalize those criteria.
The team prepares the criteria with the many treatment possibilities. How
does the team reduce the number of treatment possibilities to match the
criteria? (II:II,C,2) and (II:II,C,3) discuss methods for selecting
subsets of treatments.




What Kinds of Fields are Available for Testing
What to Consider in Selecting Subsets of Treatments
(Optional) Statistical Techniques for Selecting
Subsets of Treatments
Ways to Replicate Treatments Within and Across

Volume II: II,C,1
page 68




Material useful for practical application exercises is
available in the FSSP Paraguay Case Study Training
Document, practicum 1 and 2.

Volume II: II,C,1
page 69



After completing this activity you will be better able to:

1. Develop a treatment objectives statement for an exploratory experiment.


1. Read the background information given below and write a short treatment
objectives statement.

Background Information: Most farmers growing yams plant on mounds, up
to three feet high spaced between 3 x 3 feet and 5 x 5 feet apart.
Present practice is to apply a small amount of compound fertilizer, 1
to 2 ounces per mound, either at planting or about 1 month after vine
emergence. When applied after emergence, it is left on the surface,
and heavy rains may cause wash off. Field station research on the same
variety suggested that yields increased linearly with up to 6 ounces
per mound. There is a potential export market for any increase in

2. Be prepared to discuss your statement.

Volume II: II,C,1
page 71



1. Why Selecting Subsets of Treatments is Important
2. Agronomic Criteria
3. Economic Criteria
4. Social Criteria
5. Consumption Criteria


II:I,A What Kind of Testing to Do
II:II,C,1. Defining Treatment Objectives


Agricultural research assistant
Extension technology verification technician


After completing this section, participants will be able to:

1. Give reasons why it is often necessary to select a subset of

2. Write a complete list of treatment options.

3. Identify agronomic, economic, social, and consumption criteria for
selecting subset of treatments.


1. Writing a complete list of possible treatment options is the first step
in selecting the best subset of treatments.

2. Economic criteria affect how agronomic criteria are applied for
selecting the best subset of treatments.

3. Social criteria include not only taboos, but also effects within the
household (or compound), effects across households (in the community or
village as a whole), and preferences for specific practices,
characteristics, and uses.

4. Consumption criteria include cultural preferences for particular foods
and characteristics such as processing and storage and the anticipated
effects of treatments upon availability of food to the household.

Volume II: II,C,2
page 73


ex ante analysis
exploratory trials (testing)
factorial array
farming system
partial budgeting
production system
refinement trials (testing)
sensitivity analysis



There are often a large number of possible treatments or treatment
options that could be included in an experiment. For example, we might
list ten herbicides that could be used for pre-emergence weed control in
corn. Each could be tested at two sites, and we might want more than one
control treatment. This could give us 22 or more treatments. We could
combine such a trial with varieties, or fertilizer level, etc.

STesting all the treatment options at once is usually not practical for
one or moreof the following reasons. Too many treatments mean a large
trial. Large trials require either more land on each farm, or more complex
designs to spread the treatments across many farms. Large trials on each
farm make the trial difficult for the farm household to understand and
manage. The team may then try to compensate by increasing its own role in
management. This requires more time and more money for travel. It also
reduces farmer management and makes the trial more artificial. That
defeats an important objective of on-farm experimentation: testing
treatments under real farm conditions.

The task is then to select the best possible subset of treatments. How
does a team do this? Three steps are useful:

a. Write a complete list of possible treatment options.
b. Develop agronomic, economic, social, and consumption criteria based on
the treatment objectives statement (II:II,C,1).
c. Use the agonomic, economic, social, and consumption criteria to decide
which treatments to eliminate, and which treatments to include.

Let's look at some examples of criteria for selecting treatment
subsets, and consider their use.


Some of the agronomic criteria a team might use for selecting treatment
subsets include:

Volume II: II,C,2
page 74

a. Adaptability of the variety, practice, or input level to the
b. Acceptability of the characteristics of the crop or animal product for
local processing or consumption, or in the local market, or in an
export market;
c. Availability of seed and other inputs locally, or the probability that
these could become available;
d. Dependency on purchased inputs for acceptable performance;
e. Compatibility with the existing production system and its cropping and
livestock patterns.

The complete list of treatment options might be, for example:

a. A list of varieties (cultivars) with some desirable characteristics,
such as resistance to a specific disease or increased yield potential;
b. A full factorial array of fertilizer treatments for N, P,
and K.
c. A list of pesticides with the capacity to control a specific pest or
group of pests.

Example 1: Varieties

Let's apply these criteria to the first of these examples, the list of
varieties. For example, suppose it is a list of tomato varieties:

a. Which of these varieties are known to be tolerant of our targeted
environment? That is, tolerant of high temperatures, or wet soil
conditions? This may eliminate many.
b. If for local use what consumption characteristics are most important
and which varieties have the preferred characteristics? If for market
sale are the tomato fruits acceptable on the local market or on a
specified and available export market in terms of color, shape and
size? If for an export market, which varieties handle the best?
c. Would seed of these be available locally? Is seed available overseas
in commerical quantities? This may eliminate some "breeding lines".
d. Which of these varieties has a high requirement for other inputs
(II,D,3)? Are some susceptible to the diseases common in the domain
environment, thus requiring heavy inputs of pesticides?
e. Would these varieties fit into the production system we have targeted
and its cropping pattern? Is there evidence that they could not be
intercropped, for instance, if intercropping is the current practice?
f. Are there other useful characteristics we could use to select a subset
of varieties? Resistance to one or more diseases or pests, extended
bearing season, or ability to stand without staking are some possible

Many of these agronomic criteria depend on economic criteria. For
example, market factors would determine acceptable fruit color for criteria
b. Input costs and availability affect criteria c and d. Other examples
are also possible. This illustrates an important concept in farming
systems research: interactions among parts of the system. Each part (for
example, tomato production) depends on other parts (for example, markets
and inputs).

Volume II: II,C,2
page 75

Example 2: Fertilizer

For a fertilizer trial, the complete list of treatments might be 12

1. Three rates of nitrogen (for example, average farmer rate, a higher
rate recommended by the extension service, and an even higher rate as a
2. Two rates of phosphorus (for example, with and without).
3. Two rates of potassium (for example, also with and without).

Combining the three nitrogen treatments, the two phosphorus treatments,
and the two potassium treatments gives 3 x 2 x 2 = 12 total treatment
combinations. This is called a full, or complete, factorial array

Agronomic selection criteria might lead to the following questions:

a. Which fertilizer component is most critical: nitrogen, phosphorus, or
potassium? Which is least critical? Could we eliminate the one least
critical component?
b. What fertilizers or fertilizer carriers are available locally? Are
'straights' or single element fertilizers (like ammonium sulfate or
triple superphosphate) available locally? If only compound fertilizers
(like 10-10-10) are available locally, testing the factorial array will
only be useful for understanding biological response to the three
elements. Would it be more useful to test two or more rates of the
compound carriers locally available instead? The answer will depend on
the treatment objectives statement. In exploratory testing, the
objective may be to determine the biological response under farm
conditions to the individual elements. In that case, testing
treatments for individual elements (nitrogen, phosphorus, and
potassium) may be appropriate. In refinement testing, however, the
objective may be to select the best carrier and rate combination.
c. How might the natural environment affect increased fertilizer use? For
example, is high rainfall likely to leach out increased basal inorganic
nitrogen? Would testing different frequencies of sidedressing be more
d. Would increased fertilizer use require higher levels of other inputs?
For example, is increased use of nematicides necessary for uptake of
increased level of fertilizer?
e. Is increased fertilizer use likely to affect other crops or animals in
the production system? Will any such effects be beneficial or harmful?
For example, will fertilizer residues prove beneficial to following
crops, or increase the feed value of a pasture after the fertilized

Example 2 shows, as did example 1, how agronomic criteria often depend on
economic criteria.

Volume II: II,C,2
page 76


Most farm households aim to make money from at least part of their
farming, although subsistence farming may be an important part of their
farming systems. Subsistence production also can have an economic basis.
Supplies of a staple crop may be uncertain, or, prices may be highly
variable, especially in "hungry" periods. Economic criteria include
linkages between commercial farming activities and subsistence farming
activities. Some of these criteria include:

a. Labor needs for subsistence food crops at the same time as proposed
b. Cash needs for food or for inputs for food crop production at the same
time as proposed treatments.
c. Effects of proposed treatments on land area available for food crops or
d. Effects of proposed treatments on wild crops, animals, or fish used as
food. For example, will increased pesticides kill off fish in
irrigation canals, or make them unsafe to eat?
e. Effects of increased yield from proposed treatments on supply and
prices in the community as a whole. For example, will local merchants
hoard food stocks more, and thereby drive up prices?

A team can use various methods of economic analysis before beginning
the trial to select treatment subsets. Such analysis before testing is
called ex ante analysis. Some methods include:

a. Partial budgeting to compare anticipated increased costs and increased
benefits. Partial budgeting is discussed in III:III,A,5. This leads
b. Projected benefit/cost ratios: does any one treatment show a
substantially higher ratio than others?
c. Sensitivity analysis to examine possible changes in the ranking of
treatments depending on changes in costs of inputs and prices of
d. Input availability.
e. Market opportunities and marketing systems: increased production may
increase marketing problems or force down prices. It might be better
to aim at increased productivity of land, labor or inputs at the same
yield level.
f. Communications, especially roads as they affect access to markets, may
hamper agricultural production.
g. Consumption characteristics of different treatment subsets may pose
economic or technical trade-offs. For example, decreased storage
capability of a new cultivar in relation to traditional varieties may
require increased economic as well as technical input should the
cultivar be the most acceptable based on other criteria.

Because of linkages between commercial and subsistence farming
activities, a team may need to use these methods not only for the trial
crop or animal, but also for other crops or animals that the trial
treatments may affect. For example, if prices of a food crop increase,
farm household members may have to reduce purchases of the food crop. They

Volume II: II,C,2
page 77

may then increase land and labor used for the food crop. That can reduce
land and labor available for the trial crop.

A good reference on how to do partial budgeting, calculate projected
benefit/cost ratios, and do sensitivity analysis is the following: Perrin,
R., et. al., From Agronomic Data to Farmer Recommendations An Economics
Training Manual, CIMMYT.


Farm household members belong to many different groups at the same
time. Some of these groups may also overlap. For example, the farm
household and the community (or village) may overlap. Women may work on
compound fields as a member of their household. They may also work on
private fields of other women as a member of a group that shares labor.
Social criteria include both effects within the household or compound,
(intra-household) and across households in the community or village as a
wEole (inter-household).

Some social criteria within the household (intra-household) include:

a. Effects of changes in production of trial crops or animals on other
household activities. For example, would more frequent weeding re-uce
time available for child care?
b. Effects of changes in production of trial crops or animals on other
crop and animal production sub-systems. For example, would the
introduction of irrigated rice and hybrid varieties to male farmers in
a household take away from female farmers the rights and access to farm
swamp and mangrove rice on the same land? How would this affect
household subsistence?

Other social criteria involve the community as a whole
(inter-household). Some of these involve changes in the community that
might result from proposed treatments. Some of these criteria include:

a. Effects of proposed treatments on labor-sharing or food-sharing
b. Effects of proposed treatments on differences in power and status among
different people in the community. For example, will increased labor
needs at planting or harvest time provide more work opportunities for
landless laborers but make them more dependent on tenant farm

Other community-based social criteria involve preferences and taboos.
These community preferences and taboos can affect individual decisions
about the acceptability of proposed treatments. Some of these criteria

a. Cultural or religious preference for specific crop or animal management
practices (for example, planting mungbean during the first week of
August because of religious significance).

Volume II: II,C,2
page 78

b. Cultural or religious taboos against specific crop management practices
(for example, prohibitions against removing volunteer corn plants in
the field with a succeeding crop, against thinning extra corn plants in
a hill, because of a belief that they are "children of God" or against
castration or ear tagging of animals.

Social criteria can also affect economic and agronomic criteria. A
preference for green-skinned tomatoes may mean lower prices for red-skinned
tomatoes. The team may decide to eliminate red-skinned lines. Or if
resistance to a major disease (for example, bacterial wilt) is found only
in red-skinned tomatoes, a breeding program to transfer the resistance from
red-skinned to green-skinned tomatoes may be needed first, before moving to
on-farm experimentation. Here, social criteria lead to a new linkage
between station research and on-farm research.


Consumption criteria can be useful in evaluating the acceptability of
different trials as well as in assessing the potential impact of trials on
how much food might be available through subsistence production or market

A. Acceptability of Products

The researchers can compare--the consumption characteristics of the
proposed treatments with-current practices and preferences. For
example, is white corn just as acceptable as yellow corn (color,
flavor, ease of grinding, ease in making tortillas, etc.)

Although storage information for particular crops or varieties may
not be available prior to field trials the question of length of
storage, conditions etc. could be raised.

Cultural or religious preferences for specific products uses or
unacceptable uses may influence the acceptability of particular
treatments. For example, preference for using soybean as a coffee
substitute rather than as a vegetable eaten with rice; preference
for rice over sweet potato as a staple and use of sweet potato
only as a snack food. Taboos against specific product uses for
example, prohibitions against pigs and pig meat for Muslims or
Seventh Day Adventists.

B. Effect on Household Food Availability and Accessability

The overall availability and accessibility of households to food
as well as intra-household distribution may be altered by the
proposed trial treatments.

For example, shifts from food crops to non-food crops (cash) may
lower the amount of food available if the generated income is
insufficient or unavailable at the necessary time to purchase

Volume II: II,C,2
page 79

Shifts from production activities managed by one household member
to another may effect consumption. For example, would increased
production and income from a women's crop improve children's
diets? Other effects of changes in production of trial crops or
animals on food preparation may reduce cooking time and
subsequently reduce labor needed for gatherthing fuelwood.
Changes in harvest time may increase the availability of food
during lean times. Another question to consider is how the
treatments might.effect food uses? For example, does pesticide
used on corn mean that corn leaves can no longer safely be used in
food preparation or for animal feed? Do changes in labor
requirements free household members for other activities or only
add to already heavy labor demands at peak times throughout the


Volume III:

Complete Factorials
Analysis and Interpretation of On-Farm Experimentation.

Volume II: II,C,2
page 80






Volume II: II,C,2
page 81

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs